conceptual framework in literature review

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• CBE Life Sci Educ
• v.21(3); Fall 2022

Literature Reviews, Theoretical Frameworks, and Conceptual Frameworks: An Introduction for New Biology Education Researchers

Julie a. luft.

† Department of Mathematics, Social Studies, and Science Education, Mary Frances Early College of Education, University of Georgia, Athens, GA 30602-7124

Sophia Jeong

‡ Department of Teaching & Learning, College of Education & Human Ecology, Ohio State University, Columbus, OH 43210

Robert Idsardi

§ Department of Biology, Eastern Washington University, Cheney, WA 99004

Grant Gardner

∥ Department of Biology, Middle Tennessee State University, Murfreesboro, TN 37132

Associated Data

To frame their work, biology education researchers need to consider the role of literature reviews, theoretical frameworks, and conceptual frameworks as critical elements of the research and writing process. However, these elements can be confusing for scholars new to education research. This Research Methods article is designed to provide an overview of each of these elements and delineate the purpose of each in the educational research process. We describe what biology education researchers should consider as they conduct literature reviews, identify theoretical frameworks, and construct conceptual frameworks. Clarifying these different components of educational research studies can be helpful to new biology education researchers and the biology education research community at large in situating their work in the broader scholarly literature.

INTRODUCTION

Discipline-based education research (DBER) involves the purposeful and situated study of teaching and learning in specific disciplinary areas ( Singer et al. , 2012 ). Studies in DBER are guided by research questions that reflect disciplines’ priorities and worldviews. Researchers can use quantitative data, qualitative data, or both to answer these research questions through a variety of methodological traditions. Across all methodologies, there are different methods associated with planning and conducting educational research studies that include the use of surveys, interviews, observations, artifacts, or instruments. Ensuring the coherence of these elements to the discipline’s perspective also involves situating the work in the broader scholarly literature. The tools for doing this include literature reviews, theoretical frameworks, and conceptual frameworks. However, the purpose and function of each of these elements is often confusing to new education researchers. The goal of this article is to introduce new biology education researchers to these three important elements important in DBER scholarship and the broader educational literature.

The first element we discuss is a review of research (literature reviews), which highlights the need for a specific research question, study problem, or topic of investigation. Literature reviews situate the relevance of the study within a topic and a field. The process may seem familiar to science researchers entering DBER fields, but new researchers may still struggle in conducting the review. Booth et al. (2016b) highlight some of the challenges novice education researchers face when conducting a review of literature. They point out that novice researchers struggle in deciding how to focus the review, determining the scope of articles needed in the review, and knowing how to be critical of the articles in the review. Overcoming these challenges (and others) can help novice researchers construct a sound literature review that can inform the design of the study and help ensure the work makes a contribution to the field.

The second and third highlighted elements are theoretical and conceptual frameworks. These guide biology education research (BER) studies, and may be less familiar to science researchers. These elements are important in shaping the construction of new knowledge. Theoretical frameworks offer a way to explain and interpret the studied phenomenon, while conceptual frameworks clarify assumptions about the studied phenomenon. Despite the importance of these constructs in educational research, biology educational researchers have noted the limited use of theoretical or conceptual frameworks in published work ( DeHaan, 2011 ; Dirks, 2011 ; Lo et al. , 2019 ). In reviewing articles published in CBE—Life Sciences Education ( LSE ) between 2015 and 2019, we found that fewer than 25% of the research articles had a theoretical or conceptual framework (see the Supplemental Information), and at times there was an inconsistent use of theoretical and conceptual frameworks. Clearly, these frameworks are challenging for published biology education researchers, which suggests the importance of providing some initial guidance to new biology education researchers.

Fortunately, educational researchers have increased their explicit use of these frameworks over time, and this is influencing educational research in science, technology, engineering, and mathematics (STEM) fields. For instance, a quick search for theoretical or conceptual frameworks in the abstracts of articles in Educational Research Complete (a common database for educational research) in STEM fields demonstrates a dramatic change over the last 20 years: from only 778 articles published between 2000 and 2010 to 5703 articles published between 2010 and 2020, a more than sevenfold increase. Greater recognition of the importance of these frameworks is contributing to DBER authors being more explicit about such frameworks in their studies.

Collectively, literature reviews, theoretical frameworks, and conceptual frameworks work to guide methodological decisions and the elucidation of important findings. Each offers a different perspective on the problem of study and is an essential element in all forms of educational research. As new researchers seek to learn about these elements, they will find different resources, a variety of perspectives, and many suggestions about the construction and use of these elements. The wide range of available information can overwhelm the new researcher who just wants to learn the distinction between these elements or how to craft them adequately.

Our goal in writing this paper is not to offer specific advice about how to write these sections in scholarly work. Instead, we wanted to introduce these elements to those who are new to BER and who are interested in better distinguishing one from the other. In this paper, we share the purpose of each element in BER scholarship, along with important points on its construction. We also provide references for additional resources that may be beneficial to better understanding each element. Table 1 summarizes the key distinctions among these elements.

Comparison of literature reviews, theoretical frameworks, and conceptual reviews

This article is written for the new biology education researcher who is just learning about these different elements or for scientists looking to become more involved in BER. It is a result of our own work as science education and biology education researchers, whether as graduate students and postdoctoral scholars or newly hired and established faculty members. This is the article we wish had been available as we started to learn about these elements or discussed them with new educational researchers in biology.

LITERATURE REVIEWS

Purpose of a literature review.

A literature review is foundational to any research study in education or science. In education, a well-conceptualized and well-executed review provides a summary of the research that has already been done on a specific topic and identifies questions that remain to be answered, thus illustrating the current research project’s potential contribution to the field and the reasoning behind the methodological approach selected for the study ( Maxwell, 2012 ). BER is an evolving disciplinary area that is redefining areas of conceptual emphasis as well as orientations toward teaching and learning (e.g., Labov et al. , 2010 ; American Association for the Advancement of Science, 2011 ; Nehm, 2019 ). As a result, building comprehensive, critical, purposeful, and concise literature reviews can be a challenge for new biology education researchers.

Building Literature Reviews

There are different ways to approach and construct a literature review. Booth et al. (2016a) provide an overview that includes, for example, scoping reviews, which are focused only on notable studies and use a basic method of analysis, and integrative reviews, which are the result of exhaustive literature searches across different genres. Underlying each of these different review processes are attention to the s earch process, a ppraisa l of articles, s ynthesis of the literature, and a nalysis: SALSA ( Booth et al. , 2016a ). This useful acronym can help the researcher focus on the process while building a specific type of review.

However, new educational researchers often have questions about literature reviews that are foundational to SALSA or other approaches. Common questions concern determining which literature pertains to the topic of study or the role of the literature review in the design of the study. This section addresses such questions broadly while providing general guidance for writing a narrative literature review that evaluates the most pertinent studies.

The literature review process should begin before the research is conducted. As Boote and Beile (2005 , p. 3) suggested, researchers should be “scholars before researchers.” They point out that having a good working knowledge of the proposed topic helps illuminate avenues of study. Some subject areas have a deep body of work to read and reflect upon, providing a strong foundation for developing the research question(s). For instance, the teaching and learning of evolution is an area of long-standing interest in the BER community, generating many studies (e.g., Perry et al. , 2008 ; Barnes and Brownell, 2016 ) and reviews of research (e.g., Sickel and Friedrichsen, 2013 ; Ziadie and Andrews, 2018 ). Emerging areas of BER include the affective domain, issues of transfer, and metacognition ( Singer et al. , 2012 ). Many studies in these areas are transdisciplinary and not always specific to biology education (e.g., Rodrigo-Peiris et al. , 2018 ; Kolpikova et al. , 2019 ). These newer areas may require reading outside BER; fortunately, summaries of some of these topics can be found in the Current Insights section of the LSE website.

In focusing on a specific problem within a broader research strand, a new researcher will likely need to examine research outside BER. Depending upon the area of study, the expanded reading list might involve a mix of BER, DBER, and educational research studies. Determining the scope of the reading is not always straightforward. A simple way to focus one’s reading is to create a “summary phrase” or “research nugget,” which is a very brief descriptive statement about the study. It should focus on the essence of the study, for example, “first-year nonmajor students’ understanding of evolution,” “metacognitive prompts to enhance learning during biochemistry,” or “instructors’ inquiry-based instructional practices after professional development programming.” This type of phrase should help a new researcher identify two or more areas to review that pertain to the study. Focusing on recent research in the last 5 years is a good first step. Additional studies can be identified by reading relevant works referenced in those articles. It is also important to read seminal studies that are more than 5 years old. Reading a range of studies should give the researcher the necessary command of the subject in order to suggest a research question.

Given that the research question(s) arise from the literature review, the review should also substantiate the selected methodological approach. The review and research question(s) guide the researcher in determining how to collect and analyze data. Often the methodological approach used in a study is selected to contribute knowledge that expands upon what has been published previously about the topic (see Institute of Education Sciences and National Science Foundation, 2013 ). An emerging topic of study may need an exploratory approach that allows for a description of the phenomenon and development of a potential theory. This could, but not necessarily, require a methodological approach that uses interviews, observations, surveys, or other instruments. An extensively studied topic may call for the additional understanding of specific factors or variables; this type of study would be well suited to a verification or a causal research design. These could entail a methodological approach that uses valid and reliable instruments, observations, or interviews to determine an effect in the studied event. In either of these examples, the researcher(s) may use a qualitative, quantitative, or mixed methods methodological approach.

Even with a good research question, there is still more reading to be done. The complexity and focus of the research question dictates the depth and breadth of the literature to be examined. Questions that connect multiple topics can require broad literature reviews. For instance, a study that explores the impact of a biology faculty learning community on the inquiry instruction of faculty could have the following review areas: learning communities among biology faculty, inquiry instruction among biology faculty, and inquiry instruction among biology faculty as a result of professional learning. Biology education researchers need to consider whether their literature review requires studies from different disciplines within or outside DBER. For the example given, it would be fruitful to look at research focused on learning communities with faculty in STEM fields or in general education fields that result in instructional change. It is important not to be too narrow or too broad when reading. When the conclusions of articles start to sound similar or no new insights are gained, the researcher likely has a good foundation for a literature review. This level of reading should allow the researcher to demonstrate a mastery in understanding the researched topic, explain the suitability of the proposed research approach, and point to the need for the refined research question(s).

The literature review should include the researcher’s evaluation and critique of the selected studies. A researcher may have a large collection of studies, but not all of the studies will follow standards important in the reporting of empirical work in the social sciences. The American Educational Research Association ( Duran et al. , 2006 ), for example, offers a general discussion about standards for such work: an adequate review of research informing the study, the existence of sound and appropriate data collection and analysis methods, and appropriate conclusions that do not overstep or underexplore the analyzed data. The Institute of Education Sciences and National Science Foundation (2013) also offer Common Guidelines for Education Research and Development that can be used to evaluate collected studies.

Because not all journals adhere to such standards, it is important that a researcher review each study to determine the quality of published research, per the guidelines suggested earlier. In some instances, the research may be fatally flawed. Examples of such flaws include data that do not pertain to the question, a lack of discussion about the data collection, poorly constructed instruments, or an inadequate analysis. These types of errors result in studies that are incomplete, error-laden, or inaccurate and should be excluded from the review. Most studies have limitations, and the author(s) often make them explicit. For instance, there may be an instructor effect, recognized bias in the analysis, or issues with the sample population. Limitations are usually addressed by the research team in some way to ensure a sound and acceptable research process. Occasionally, the limitations associated with the study can be significant and not addressed adequately, which leaves a consequential decision in the hands of the researcher. Providing critiques of studies in the literature review process gives the reader confidence that the researcher has carefully examined relevant work in preparation for the study and, ultimately, the manuscript.

A solid literature review clearly anchors the proposed study in the field and connects the research question(s), the methodological approach, and the discussion. Reviewing extant research leads to research questions that will contribute to what is known in the field. By summarizing what is known, the literature review points to what needs to be known, which in turn guides decisions about methodology. Finally, notable findings of the new study are discussed in reference to those described in the literature review.

Within published BER studies, literature reviews can be placed in different locations in an article. When included in the introductory section of the study, the first few paragraphs of the manuscript set the stage, with the literature review following the opening paragraphs. Cooper et al. (2019) illustrate this approach in their study of course-based undergraduate research experiences (CUREs). An introduction discussing the potential of CURES is followed by an analysis of the existing literature relevant to the design of CUREs that allows for novel student discoveries. Within this review, the authors point out contradictory findings among research on novel student discoveries. This clarifies the need for their study, which is described and highlighted through specific research aims.

A literature reviews can also make up a separate section in a paper. For example, the introduction to Todd et al. (2019) illustrates the need for their research topic by highlighting the potential of learning progressions (LPs) and suggesting that LPs may help mitigate learning loss in genetics. At the end of the introduction, the authors state their specific research questions. The review of literature following this opening section comprises two subsections. One focuses on learning loss in general and examines a variety of studies and meta-analyses from the disciplines of medical education, mathematics, and reading. The second section focuses specifically on LPs in genetics and highlights student learning in the midst of LPs. These separate reviews provide insights into the stated research question.

Suggestions and Advice

A well-conceptualized, comprehensive, and critical literature review reveals the understanding of the topic that the researcher brings to the study. Literature reviews should not be so big that there is no clear area of focus; nor should they be so narrow that no real research question arises. The task for a researcher is to craft an efficient literature review that offers a critical analysis of published work, articulates the need for the study, guides the methodological approach to the topic of study, and provides an adequate foundation for the discussion of the findings.

In our own writing of literature reviews, there are often many drafts. An early draft may seem well suited to the study because the need for and approach to the study are well described. However, as the results of the study are analyzed and findings begin to emerge, the existing literature review may be inadequate and need revision. The need for an expanded discussion about the research area can result in the inclusion of new studies that support the explanation of a potential finding. The literature review may also prove to be too broad. Refocusing on a specific area allows for more contemplation of a finding.

It should be noted that there are different types of literature reviews, and many books and articles have been written about the different ways to embark on these types of reviews. Among these different resources, the following may be helpful in considering how to refine the review process for scholarly journals:

• Booth, A., Sutton, A., & Papaioannou, D. (2016a). Systemic approaches to a successful literature review (2nd ed.). Los Angeles, CA: Sage. This book addresses different types of literature reviews and offers important suggestions pertaining to defining the scope of the literature review and assessing extant studies.
• Booth, W. C., Colomb, G. G., Williams, J. M., Bizup, J., & Fitzgerald, W. T. (2016b). The craft of research (4th ed.). Chicago: University of Chicago Press. This book can help the novice consider how to make the case for an area of study. While this book is not specifically about literature reviews, it offers suggestions about making the case for your study.
• Galvan, J. L., & Galvan, M. C. (2017). Writing literature reviews: A guide for students of the social and behavioral sciences (7th ed.). Routledge. This book offers guidance on writing different types of literature reviews. For the novice researcher, there are useful suggestions for creating coherent literature reviews.

THEORETICAL FRAMEWORKS

Purpose of theoretical frameworks.

As new education researchers may be less familiar with theoretical frameworks than with literature reviews, this discussion begins with an analogy. Envision a biologist, chemist, and physicist examining together the dramatic effect of a fog tsunami over the ocean. A biologist gazing at this phenomenon may be concerned with the effect of fog on various species. A chemist may be interested in the chemical composition of the fog as water vapor condenses around bits of salt. A physicist may be focused on the refraction of light to make fog appear to be “sitting” above the ocean. While observing the same “objective event,” the scientists are operating under different theoretical frameworks that provide a particular perspective or “lens” for the interpretation of the phenomenon. Each of these scientists brings specialized knowledge, experiences, and values to this phenomenon, and these influence the interpretation of the phenomenon. The scientists’ theoretical frameworks influence how they design and carry out their studies and interpret their data.

Within an educational study, a theoretical framework helps to explain a phenomenon through a particular lens and challenges and extends existing knowledge within the limitations of that lens. Theoretical frameworks are explicitly stated by an educational researcher in the paper’s framework, theory, or relevant literature section. The framework shapes the types of questions asked, guides the method by which data are collected and analyzed, and informs the discussion of the results of the study. It also reveals the researcher’s subjectivities, for example, values, social experience, and viewpoint ( Allen, 2017 ). It is essential that a novice researcher learn to explicitly state a theoretical framework, because all research questions are being asked from the researcher’s implicit or explicit assumptions of a phenomenon of interest ( Schwandt, 2000 ).

Selecting Theoretical Frameworks

Theoretical frameworks are one of the most contemplated elements in our work in educational research. In this section, we share three important considerations for new scholars selecting a theoretical framework.

The first step in identifying a theoretical framework involves reflecting on the phenomenon within the study and the assumptions aligned with the phenomenon. The phenomenon involves the studied event. There are many possibilities, for example, student learning, instructional approach, or group organization. A researcher holds assumptions about how the phenomenon will be effected, influenced, changed, or portrayed. It is ultimately the researcher’s assumption(s) about the phenomenon that aligns with a theoretical framework. An example can help illustrate how a researcher’s reflection on the phenomenon and acknowledgment of assumptions can result in the identification of a theoretical framework.

In our example, a biology education researcher may be interested in exploring how students’ learning of difficult biological concepts can be supported by the interactions of group members. The phenomenon of interest is the interactions among the peers, and the researcher assumes that more knowledgeable students are important in supporting the learning of the group. As a result, the researcher may draw on Vygotsky’s (1978) sociocultural theory of learning and development that is focused on the phenomenon of student learning in a social setting. This theory posits the critical nature of interactions among students and between students and teachers in the process of building knowledge. A researcher drawing upon this framework holds the assumption that learning is a dynamic social process involving questions and explanations among students in the classroom and that more knowledgeable peers play an important part in the process of building conceptual knowledge.

It is important to state at this point that there are many different theoretical frameworks. Some frameworks focus on learning and knowing, while other theoretical frameworks focus on equity, empowerment, or discourse. Some frameworks are well articulated, and others are still being refined. For a new researcher, it can be challenging to find a theoretical framework. Two of the best ways to look for theoretical frameworks is through published works that highlight different frameworks.

When a theoretical framework is selected, it should clearly connect to all parts of the study. The framework should augment the study by adding a perspective that provides greater insights into the phenomenon. It should clearly align with the studies described in the literature review. For instance, a framework focused on learning would correspond to research that reported different learning outcomes for similar studies. The methods for data collection and analysis should also correspond to the framework. For instance, a study about instructional interventions could use a theoretical framework concerned with learning and could collect data about the effect of the intervention on what is learned. When the data are analyzed, the theoretical framework should provide added meaning to the findings, and the findings should align with the theoretical framework.

A study by Jensen and Lawson (2011) provides an example of how a theoretical framework connects different parts of the study. They compared undergraduate biology students in heterogeneous and homogeneous groups over the course of a semester. Jensen and Lawson (2011) assumed that learning involved collaboration and more knowledgeable peers, which made Vygotsky’s (1978) theory a good fit for their study. They predicted that students in heterogeneous groups would experience greater improvement in their reasoning abilities and science achievements with much of the learning guided by the more knowledgeable peers.

In the enactment of the study, they collected data about the instruction in traditional and inquiry-oriented classes, while the students worked in homogeneous or heterogeneous groups. To determine the effect of working in groups, the authors also measured students’ reasoning abilities and achievement. Each data-collection and analysis decision connected to understanding the influence of collaborative work.

Their findings highlighted aspects of Vygotsky’s (1978) theory of learning. One finding, for instance, posited that inquiry instruction, as a whole, resulted in reasoning and achievement gains. This links to Vygotsky (1978) , because inquiry instruction involves interactions among group members. A more nuanced finding was that group composition had a conditional effect. Heterogeneous groups performed better with more traditional and didactic instruction, regardless of the reasoning ability of the group members. Homogeneous groups worked better during interaction-rich activities for students with low reasoning ability. The authors attributed the variation to the different types of helping behaviors of students. High-performing students provided the answers, while students with low reasoning ability had to work collectively through the material. In terms of Vygotsky (1978) , this finding provided new insights into the learning context in which productive interactions can occur for students.

Another consideration in the selection and use of a theoretical framework pertains to its orientation to the study. This can result in the theoretical framework prioritizing individuals, institutions, and/or policies ( Anfara and Mertz, 2014 ). Frameworks that connect to individuals, for instance, could contribute to understanding their actions, learning, or knowledge. Institutional frameworks, on the other hand, offer insights into how institutions, organizations, or groups can influence individuals or materials. Policy theories provide ways to understand how national or local policies can dictate an emphasis on outcomes or instructional design. These different types of frameworks highlight different aspects in an educational setting, which influences the design of the study and the collection of data. In addition, these different frameworks offer a way to make sense of the data. Aligning the data collection and analysis with the framework ensures that a study is coherent and can contribute to the field.

New understandings emerge when different theoretical frameworks are used. For instance, Ebert-May et al. (2015) prioritized the individual level within conceptual change theory (see Posner et al. , 1982 ). In this theory, an individual’s knowledge changes when it no longer fits the phenomenon. Ebert-May et al. (2015) designed a professional development program challenging biology postdoctoral scholars’ existing conceptions of teaching. The authors reported that the biology postdoctoral scholars’ teaching practices became more student-centered as they were challenged to explain their instructional decision making. According to the theory, the biology postdoctoral scholars’ dissatisfaction in their descriptions of teaching and learning initiated change in their knowledge and instruction. These results reveal how conceptual change theory can explain the learning of participants and guide the design of professional development programming.

The communities of practice (CoP) theoretical framework ( Lave, 1988 ; Wenger, 1998 ) prioritizes the institutional level , suggesting that learning occurs when individuals learn from and contribute to the communities in which they reside. Grounded in the assumption of community learning, the literature on CoP suggests that, as individuals interact regularly with the other members of their group, they learn about the rules, roles, and goals of the community ( Allee, 2000 ). A study conducted by Gehrke and Kezar (2017) used the CoP framework to understand organizational change by examining the involvement of individual faculty engaged in a cross-institutional CoP focused on changing the instructional practice of faculty at each institution. In the CoP, faculty members were involved in enhancing instructional materials within their department, which aligned with an overarching goal of instituting instruction that embraced active learning. Not surprisingly, Gehrke and Kezar (2017) revealed that faculty who perceived the community culture as important in their work cultivated institutional change. Furthermore, they found that institutional change was sustained when key leaders served as mentors and provided support for faculty, and as faculty themselves developed into leaders. This study reveals the complexity of individual roles in a COP in order to support institutional instructional change.

It is important to explicitly state the theoretical framework used in a study, but elucidating a theoretical framework can be challenging for a new educational researcher. The literature review can help to identify an applicable theoretical framework. Focal areas of the review or central terms often connect to assumptions and assertions associated with the framework that pertain to the phenomenon of interest. Another way to identify a theoretical framework is self-reflection by the researcher on personal beliefs and understandings about the nature of knowledge the researcher brings to the study ( Lysaght, 2011 ). In stating one’s beliefs and understandings related to the study (e.g., students construct their knowledge, instructional materials support learning), an orientation becomes evident that will suggest a particular theoretical framework. Theoretical frameworks are not arbitrary , but purposefully selected.

With experience, a researcher may find expanded roles for theoretical frameworks. Researchers may revise an existing framework that has limited explanatory power, or they may decide there is a need to develop a new theoretical framework. These frameworks can emerge from a current study or the need to explain a phenomenon in a new way. Researchers may also find that multiple theoretical frameworks are necessary to frame and explore a problem, as different frameworks can provide different insights into a problem.

Finally, it is important to recognize that choosing “x” theoretical framework does not necessarily mean a researcher chooses “y” methodology and so on, nor is there a clear-cut, linear process in selecting a theoretical framework for one’s study. In part, the nonlinear process of identifying a theoretical framework is what makes understanding and using theoretical frameworks challenging. For the novice scholar, contemplating and understanding theoretical frameworks is essential. Fortunately, there are articles and books that can help:

• Creswell, J. W. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Los Angeles, CA: Sage. This book provides an overview of theoretical frameworks in general educational research.
• Ding, L. (2019). Theoretical perspectives of quantitative physics education research. Physical Review Physics Education Research , 15 (2), 020101-1–020101-13. This paper illustrates how a DBER field can use theoretical frameworks.
• Nehm, R. (2019). Biology education research: Building integrative frameworks for teaching and learning about living systems. Disciplinary and Interdisciplinary Science Education Research , 1 , ar15. https://doi.org/10.1186/s43031-019-0017-6 . This paper articulates the need for studies in BER to explicitly state theoretical frameworks and provides examples of potential studies.
• Patton, M. Q. (2015). Qualitative research & evaluation methods: Integrating theory and practice . Sage. This book also provides an overview of theoretical frameworks, but for both research and evaluation.

CONCEPTUAL FRAMEWORKS

Purpose of a conceptual framework.

A conceptual framework is a description of the way a researcher understands the factors and/or variables that are involved in the study and their relationships to one another. The purpose of a conceptual framework is to articulate the concepts under study using relevant literature ( Rocco and Plakhotnik, 2009 ) and to clarify the presumed relationships among those concepts ( Rocco and Plakhotnik, 2009 ; Anfara and Mertz, 2014 ). Conceptual frameworks are different from theoretical frameworks in both their breadth and grounding in established findings. Whereas a theoretical framework articulates the lens through which a researcher views the work, the conceptual framework is often more mechanistic and malleable.

Conceptual frameworks are broader, encompassing both established theories (i.e., theoretical frameworks) and the researchers’ own emergent ideas. Emergent ideas, for example, may be rooted in informal and/or unpublished observations from experience. These emergent ideas would not be considered a “theory” if they are not yet tested, supported by systematically collected evidence, and peer reviewed. However, they do still play an important role in the way researchers approach their studies. The conceptual framework allows authors to clearly describe their emergent ideas so that connections among ideas in the study and the significance of the study are apparent to readers.

Constructing Conceptual Frameworks

Including a conceptual framework in a research study is important, but researchers often opt to include either a conceptual or a theoretical framework. Either may be adequate, but both provide greater insight into the research approach. For instance, a research team plans to test a novel component of an existing theory. In their study, they describe the existing theoretical framework that informs their work and then present their own conceptual framework. Within this conceptual framework, specific topics portray emergent ideas that are related to the theory. Describing both frameworks allows readers to better understand the researchers’ assumptions, orientations, and understanding of concepts being investigated. For example, Connolly et al. (2018) included a conceptual framework that described how they applied a theoretical framework of social cognitive career theory (SCCT) to their study on teaching programs for doctoral students. In their conceptual framework, the authors described SCCT, explained how it applied to the investigation, and drew upon results from previous studies to justify the proposed connections between the theory and their emergent ideas.

In some cases, authors may be able to sufficiently describe their conceptualization of the phenomenon under study in an introduction alone, without a separate conceptual framework section. However, incomplete descriptions of how the researchers conceptualize the components of the study may limit the significance of the study by making the research less intelligible to readers. This is especially problematic when studying topics in which researchers use the same terms for different constructs or different terms for similar and overlapping constructs (e.g., inquiry, teacher beliefs, pedagogical content knowledge, or active learning). Authors must describe their conceptualization of a construct if the research is to be understandable and useful.

There are some key areas to consider regarding the inclusion of a conceptual framework in a study. To begin with, it is important to recognize that conceptual frameworks are constructed by the researchers conducting the study ( Rocco and Plakhotnik, 2009 ; Maxwell, 2012 ). This is different from theoretical frameworks that are often taken from established literature. Researchers should bring together ideas from the literature, but they may be influenced by their own experiences as a student and/or instructor, the shared experiences of others, or thought experiments as they construct a description, model, or representation of their understanding of the phenomenon under study. This is an exercise in intellectual organization and clarity that often considers what is learned, known, and experienced. The conceptual framework makes these constructs explicitly visible to readers, who may have different understandings of the phenomenon based on their prior knowledge and experience. There is no single method to go about this intellectual work.

Reeves et al. (2016) is an example of an article that proposed a conceptual framework about graduate teaching assistant professional development evaluation and research. The authors used existing literature to create a novel framework that filled a gap in current research and practice related to the training of graduate teaching assistants. This conceptual framework can guide the systematic collection of data by other researchers because the framework describes the relationships among various factors that influence teaching and learning. The Reeves et al. (2016) conceptual framework may be modified as additional data are collected and analyzed by other researchers. This is not uncommon, as conceptual frameworks can serve as catalysts for concerted research efforts that systematically explore a phenomenon (e.g., Reynolds et al. , 2012 ; Brownell and Kloser, 2015 ).

Sabel et al. (2017) used a conceptual framework in their exploration of how scaffolds, an external factor, interact with internal factors to support student learning. Their conceptual framework integrated principles from two theoretical frameworks, self-regulated learning and metacognition, to illustrate how the research team conceptualized students’ use of scaffolds in their learning ( Figure 1 ). Sabel et al. (2017) created this model using their interpretations of these two frameworks in the context of their teaching.

Conceptual framework from Sabel et al. (2017) .

A conceptual framework should describe the relationship among components of the investigation ( Anfara and Mertz, 2014 ). These relationships should guide the researcher’s methods of approaching the study ( Miles et al. , 2014 ) and inform both the data to be collected and how those data should be analyzed. Explicitly describing the connections among the ideas allows the researcher to justify the importance of the study and the rigor of the research design. Just as importantly, these frameworks help readers understand why certain components of a system were not explored in the study. This is a challenge in education research, which is rooted in complex environments with many variables that are difficult to control.

For example, Sabel et al. (2017) stated: “Scaffolds, such as enhanced answer keys and reflection questions, can help students and instructors bridge the external and internal factors and support learning” (p. 3). They connected the scaffolds in the study to the three dimensions of metacognition and the eventual transformation of existing ideas into new or revised ideas. Their framework provides a rationale for focusing on how students use two different scaffolds, and not on other factors that may influence a student’s success (self-efficacy, use of active learning, exam format, etc.).

In constructing conceptual frameworks, researchers should address needed areas of study and/or contradictions discovered in literature reviews. By attending to these areas, researchers can strengthen their arguments for the importance of a study. For instance, conceptual frameworks can address how the current study will fill gaps in the research, resolve contradictions in existing literature, or suggest a new area of study. While a literature review describes what is known and not known about the phenomenon, the conceptual framework leverages these gaps in describing the current study ( Maxwell, 2012 ). In the example of Sabel et al. (2017) , the authors indicated there was a gap in the literature regarding how scaffolds engage students in metacognition to promote learning in large classes. Their study helps fill that gap by describing how scaffolds can support students in the three dimensions of metacognition: intelligibility, plausibility, and wide applicability. In another example, Lane (2016) integrated research from science identity, the ethic of care, the sense of belonging, and an expertise model of student success to form a conceptual framework that addressed the critiques of other frameworks. In a more recent example, Sbeglia et al. (2021) illustrated how a conceptual framework influences the methodological choices and inferences in studies by educational researchers.

Sometimes researchers draw upon the conceptual frameworks of other researchers. When a researcher’s conceptual framework closely aligns with an existing framework, the discussion may be brief. For example, Ghee et al. (2016) referred to portions of SCCT as their conceptual framework to explain the significance of their work on students’ self-efficacy and career interests. Because the authors’ conceptualization of this phenomenon aligned with a previously described framework, they briefly mentioned the conceptual framework and provided additional citations that provided more detail for the readers.

Within both the BER and the broader DBER communities, conceptual frameworks have been used to describe different constructs. For example, some researchers have used the term “conceptual framework” to describe students’ conceptual understandings of a biological phenomenon. This is distinct from a researcher’s conceptual framework of the educational phenomenon under investigation, which may also need to be explicitly described in the article. Other studies have presented a research logic model or flowchart of the research design as a conceptual framework. These constructions can be quite valuable in helping readers understand the data-collection and analysis process. However, a model depicting the study design does not serve the same role as a conceptual framework. Researchers need to avoid conflating these constructs by differentiating the researchers’ conceptual framework that guides the study from the research design, when applicable.

Explicitly describing conceptual frameworks is essential in depicting the focus of the study. We have found that being explicit in a conceptual framework means using accepted terminology, referencing prior work, and clearly noting connections between terms. This description can also highlight gaps in the literature or suggest potential contributions to the field of study. A well-elucidated conceptual framework can suggest additional studies that may be warranted. This can also spur other researchers to consider how they would approach the examination of a phenomenon and could result in a revised conceptual framework.

It can be challenging to create conceptual frameworks, but they are important. Below are two resources that could be helpful in constructing and presenting conceptual frameworks in educational research:

• Maxwell, J. A. (2012). Qualitative research design: An interactive approach (3rd ed.). Los Angeles, CA: Sage. Chapter 3 in this book describes how to construct conceptual frameworks.
• Ravitch, S. M., & Riggan, M. (2016). Reason & rigor: How conceptual frameworks guide research . Los Angeles, CA: Sage. This book explains how conceptual frameworks guide the research questions, data collection, data analyses, and interpretation of results.

CONCLUDING THOUGHTS

Literature reviews, theoretical frameworks, and conceptual frameworks are all important in DBER and BER. Robust literature reviews reinforce the importance of a study. Theoretical frameworks connect the study to the base of knowledge in educational theory and specify the researcher’s assumptions. Conceptual frameworks allow researchers to explicitly describe their conceptualization of the relationships among the components of the phenomenon under study. Table 1 provides a general overview of these components in order to assist biology education researchers in thinking about these elements.

It is important to emphasize that these different elements are intertwined. When these elements are aligned and complement one another, the study is coherent, and the study findings contribute to knowledge in the field. When literature reviews, theoretical frameworks, and conceptual frameworks are disconnected from one another, the study suffers. The point of the study is lost, suggested findings are unsupported, or important conclusions are invisible to the researcher. In addition, this misalignment may be costly in terms of time and money.

Conducting a literature review, selecting a theoretical framework, and building a conceptual framework are some of the most difficult elements of a research study. It takes time to understand the relevant research, identify a theoretical framework that provides important insights into the study, and formulate a conceptual framework that organizes the finding. In the research process, there is often a constant back and forth among these elements as the study evolves. With an ongoing refinement of the review of literature, clarification of the theoretical framework, and articulation of a conceptual framework, a sound study can emerge that makes a contribution to the field. This is the goal of BER and education research.

Supplementary Material

• Allee, V. (2000). Knowledge networks and communities of learning . OD Practitioner , 32 ( 4 ), 4–13. [ Google Scholar ]
• Allen, M. (2017). The Sage encyclopedia of communication research methods (Vols. 1–4 ). Los Angeles, CA: Sage. 10.4135/9781483381411 [ CrossRef ] [ Google Scholar ]
• American Association for the Advancement of Science. (2011). Vision and change in undergraduate biology education: A call to action . Washington, DC. [ Google Scholar ]
• Anfara, V. A., Mertz, N. T. (2014). Setting the stage . In Anfara, V. A., Mertz, N. T. (eds.), Theoretical frameworks in qualitative research (pp. 1–22). Sage. [ Google Scholar ]
• Barnes, M. E., Brownell, S. E. (2016). Practices and perspectives of college instructors on addressing religious beliefs when teaching evolution . CBE—Life Sciences Education , 15 ( 2 ), ar18. https://doi.org/10.1187/cbe.15-11-0243 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Boote, D. N., Beile, P. (2005). Scholars before researchers: On the centrality of the dissertation literature review in research preparation . Educational Researcher , 34 ( 6 ), 3–15. 10.3102/0013189x034006003 [ CrossRef ] [ Google Scholar ]
• Booth, A., Sutton, A., Papaioannou, D. (2016a). Systemic approaches to a successful literature review (2nd ed.). Los Angeles, CA: Sage. [ Google Scholar ]
• Booth, W. C., Colomb, G. G., Williams, J. M., Bizup, J., Fitzgerald, W. T. (2016b). The craft of research (4th ed.). Chicago, IL: University of Chicago Press. [ Google Scholar ]
• Brownell, S. E., Kloser, M. J. (2015). Toward a conceptual framework for measuring the effectiveness of course-based undergraduate research experiences in undergraduate biology . Studies in Higher Education , 40 ( 3 ), 525–544. https://doi.org/10.1080/03075079.2015.1004234 [ Google Scholar ]
• Connolly, M. R., Lee, Y. G., Savoy, J. N. (2018). The effects of doctoral teaching development on early-career STEM scholars’ college teaching self-efficacy . CBE—Life Sciences Education , 17 ( 1 ), ar14. https://doi.org/10.1187/cbe.17-02-0039 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Cooper, K. M., Blattman, J. N., Hendrix, T., Brownell, S. E. (2019). The impact of broadly relevant novel discoveries on student project ownership in a traditional lab course turned CURE . CBE—Life Sciences Education , 18 ( 4 ), ar57. https://doi.org/10.1187/cbe.19-06-0113 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Creswell, J. W. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Los Angeles, CA: Sage. [ Google Scholar ]
• DeHaan, R. L. (2011). Education research in the biological sciences: A nine decade review (Paper commissioned by the NAS/NRC Committee on the Status, Contributions, and Future Directions of Discipline Based Education Research) . Washington, DC: National Academies Press. Retrieved May 20, 2022, from www7.nationalacademies.org/bose/DBER_Mee ting2_commissioned_papers_page.html [ Google Scholar ]
• Ding, L. (2019). Theoretical perspectives of quantitative physics education research . Physical Review Physics Education Research , 15 ( 2 ), 020101. [ Google Scholar ]
• Dirks, C. (2011). The current status and future direction of biology education research . Paper presented at: Second Committee Meeting on the Status, Contributions, and Future Directions of Discipline-Based Education Research, 18–19 October (Washington, DC). Retrieved May 20, 2022, from http://sites.nationalacademies.org/DBASSE/BOSE/DBASSE_071087 [ Google Scholar ]
• Duran, R. P., Eisenhart, M. A., Erickson, F. D., Grant, C. A., Green, J. L., Hedges, L. V., Schneider, B. L. (2006). Standards for reporting on empirical social science research in AERA publications: American Educational Research Association . Educational Researcher , 35 ( 6 ), 33–40. [ Google Scholar ]
• Ebert-May, D., Derting, T. L., Henkel, T. P., Middlemis Maher, J., Momsen, J. L., Arnold, B., Passmore, H. A. (2015). Breaking the cycle: Future faculty begin teaching with learner-centered strategies after professional development . CBE—Life Sciences Education , 14 ( 2 ), ar22. https://doi.org/10.1187/cbe.14-12-0222 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Galvan, J. L., Galvan, M. C. (2017). Writing literature reviews: A guide for students of the social and behavioral sciences (7th ed.). New York, NY: Routledge. https://doi.org/10.4324/9781315229386 [ Google Scholar ]
• Gehrke, S., Kezar, A. (2017). The roles of STEM faculty communities of practice in institutional and departmental reform in higher education . American Educational Research Journal , 54 ( 5 ), 803–833. https://doi.org/10.3102/0002831217706736 [ Google Scholar ]
• Ghee, M., Keels, M., Collins, D., Neal-Spence, C., Baker, E. (2016). Fine-tuning summer research programs to promote underrepresented students’ persistence in the STEM pathway . CBE—Life Sciences Education , 15 ( 3 ), ar28. https://doi.org/10.1187/cbe.16-01-0046 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Institute of Education Sciences & National Science Foundation. (2013). Common guidelines for education research and development . Retrieved May 20, 2022, from www.nsf.gov/pubs/2013/nsf13126/nsf13126.pdf
• Jensen, J. L., Lawson, A. (2011). Effects of collaborative group composition and inquiry instruction on reasoning gains and achievement in undergraduate biology . CBE—Life Sciences Education , 10 ( 1 ), 64–73. https://doi.org/10.1187/cbe.19-05-0098 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kolpikova, E. P., Chen, D. C., Doherty, J. H. (2019). Does the format of preclass reading quizzes matter? An evaluation of traditional and gamified, adaptive preclass reading quizzes . CBE—Life Sciences Education , 18 ( 4 ), ar52. https://doi.org/10.1187/cbe.19-05-0098 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Labov, J. B., Reid, A. H., Yamamoto, K. R. (2010). Integrated biology and undergraduate science education: A new biology education for the twenty-first century? CBE—Life Sciences Education , 9 ( 1 ), 10–16. https://doi.org/10.1187/cbe.09-12-0092 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lane, T. B. (2016). Beyond academic and social integration: Understanding the impact of a STEM enrichment program on the retention and degree attainment of underrepresented students . CBE—Life Sciences Education , 15 ( 3 ), ar39. https://doi.org/10.1187/cbe.16-01-0070 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lave, J. (1988). Cognition in practice: Mind, mathematics and culture in everyday life . New York, NY: Cambridge University Press. [ Google Scholar ]
• Lo, S. M., Gardner, G. E., Reid, J., Napoleon-Fanis, V., Carroll, P., Smith, E., Sato, B. K. (2019). Prevailing questions and methodologies in biology education research: A longitudinal analysis of research in CBE — Life Sciences Education and at the Society for the Advancement of Biology Education Research . CBE—Life Sciences Education , 18 ( 1 ), ar9. https://doi.org/10.1187/cbe.18-08-0164 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lysaght, Z. (2011). Epistemological and paradigmatic ecumenism in “Pasteur’s quadrant:” Tales from doctoral research . In Official Conference Proceedings of the Third Asian Conference on Education in Osaka, Japan . Retrieved May 20, 2022, from http://iafor.org/ace2011_offprint/ACE2011_offprint_0254.pdf
• Maxwell, J. A. (2012). Qualitative research design: An interactive approach (3rd ed.). Los Angeles, CA: Sage. [ Google Scholar ]
• Miles, M. B., Huberman, A. M., Saldaña, J. (2014). Qualitative data analysis (3rd ed.). Los Angeles, CA: Sage. [ Google Scholar ]
• Nehm, R. (2019). Biology education research: Building integrative frameworks for teaching and learning about living systems . Disciplinary and Interdisciplinary Science Education Research , 1 , ar15. https://doi.org/10.1186/s43031-019-0017-6 [ Google Scholar ]
• Patton, M. Q. (2015). Qualitative research & evaluation methods: Integrating theory and practice . Los Angeles, CA: Sage. [ Google Scholar ]
• Perry, J., Meir, E., Herron, J. C., Maruca, S., Stal, D. (2008). Evaluating two approaches to helping college students understand evolutionary trees through diagramming tasks . CBE—Life Sciences Education , 7 ( 2 ), 193–201. https://doi.org/10.1187/cbe.07-01-0007 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Posner, G. J., Strike, K. A., Hewson, P. W., Gertzog, W. A. (1982). Accommodation of a scientific conception: Toward a theory of conceptual change . Science Education , 66 ( 2 ), 211–227. [ Google Scholar ]
• Ravitch, S. M., Riggan, M. (2016). Reason & rigor: How conceptual frameworks guide research . Los Angeles, CA: Sage. [ Google Scholar ]
• Reeves, T. D., Marbach-Ad, G., Miller, K. R., Ridgway, J., Gardner, G. E., Schussler, E. E., Wischusen, E. W. (2016). A conceptual framework for graduate teaching assistant professional development evaluation and research . CBE—Life Sciences Education , 15 ( 2 ), es2. https://doi.org/10.1187/cbe.15-10-0225 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Reynolds, J. A., Thaiss, C., Katkin, W., Thompson, R. J. Jr. (2012). Writing-to-learn in undergraduate science education: A community-based, conceptually driven approach . CBE—Life Sciences Education , 11 ( 1 ), 17–25. https://doi.org/10.1187/cbe.11-08-0064 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rocco, T. S., Plakhotnik, M. S. (2009). Literature reviews, conceptual frameworks, and theoretical frameworks: Terms, functions, and distinctions . Human Resource Development Review , 8 ( 1 ), 120–130. https://doi.org/10.1177/1534484309332617 [ Google Scholar ]
• Rodrigo-Peiris, T., Xiang, L., Cassone, V. M. (2018). A low-intensity, hybrid design between a “traditional” and a “course-based” research experience yields positive outcomes for science undergraduate freshmen and shows potential for large-scale application . CBE—Life Sciences Education , 17 ( 4 ), ar53. https://doi.org/10.1187/cbe.17-11-0248 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sabel, J. L., Dauer, J. T., Forbes, C. T. (2017). Introductory biology students’ use of enhanced answer keys and reflection questions to engage in metacognition and enhance understanding . CBE—Life Sciences Education , 16 ( 3 ), ar40. https://doi.org/10.1187/cbe.16-10-0298 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sbeglia, G. C., Goodridge, J. A., Gordon, L. H., Nehm, R. H. (2021). Are faculty changing? How reform frameworks, sampling intensities, and instrument measures impact inferences about student-centered teaching practices . CBE—Life Sciences Education , 20 ( 3 ), ar39. https://doi.org/10.1187/cbe.20-11-0259 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Schwandt, T. A. (2000). Three epistemological stances for qualitative inquiry: Interpretivism, hermeneutics, and social constructionism . In Denzin, N. K., Lincoln, Y. S. (Eds.), Handbook of qualitative research (2nd ed., pp. 189–213). Los Angeles, CA: Sage. [ Google Scholar ]
• Sickel, A. J., Friedrichsen, P. (2013). Examining the evolution education literature with a focus on teachers: Major findings, goals for teacher preparation, and directions for future research . Evolution: Education and Outreach , 6 ( 1 ), 23. https://doi.org/10.1186/1936-6434-6-23 [ Google Scholar ]
• Singer, S. R., Nielsen, N. R., Schweingruber, H. A. (2012). Discipline-based education research: Understanding and improving learning in undergraduate science and engineering . Washington, DC: National Academies Press. [ Google Scholar ]
• Todd, A., Romine, W. L., Correa-Menendez, J. (2019). Modeling the transition from a phenotypic to genotypic conceptualization of genetics in a university-level introductory biology context . Research in Science Education , 49 ( 2 ), 569–589. https://doi.org/10.1007/s11165-017-9626-2 [ Google Scholar ]
• Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes . Cambridge, MA: Harvard University Press. [ Google Scholar ]
• Wenger, E. (1998). Communities of practice: Learning as a social system . Systems Thinker , 9 ( 5 ), 2–3. [ Google Scholar ]
• Ziadie, M. A., Andrews, T. C. (2018). Moving evolution education forward: A systematic analysis of literature to identify gaps in collective knowledge for teaching . CBE—Life Sciences Education , 17 ( 1 ), ar11. https://doi.org/10.1187/cbe.17-08-0190 [ PMC free article ] [ PubMed ] [ Google Scholar ]

Educational resources and simple solutions for your research journey

What is a Conceptual Framework and How to Make It (with Examples)

A strong conceptual framework underpins good research. A conceptual framework in research is used to understand a research problem and guide the development and analysis of the research. It serves as a roadmap to conceptualize and structure the work by providing an outline that connects different ideas, concepts, and theories within the field of study. A conceptual framework pictorially or verbally depicts presumed relationships among the study variables.

The purpose of a conceptual framework is to serve as a scheme for organizing and categorizing knowledge and thereby help researchers in developing theories and hypotheses and conducting empirical studies.

In this post, we explain what is a conceptual framework, and provide expert advice on how to make a conceptual framework, along with conceptual framework examples.

Table of Contents

What is a Conceptual Framework in Research

Definition of a conceptual framework.

A conceptual framework includes key concepts, variables, relationships, and assumptions that guide the academic inquiry. It establishes the theoretical underpinnings and provides a lens through which researchers can analyze and interpret data. A conceptual framework draws upon existing theories, models, or established bodies of knowledge to provide a structure for understanding the research problem. It defines the scope of research, identifying relevant variables, establishing research questions, and guiding the selection of appropriate methodologies and data analysis techniques.

Conceptual frameworks can be written or visual. Other types of conceptual framework representations might be taxonomic (verbal description categorizing phenomena into classes without showing relationships between classes) or mathematical descriptions (expression of phenomena in the form of mathematical equations).

Figure 1: Definition of a conceptual framework explained diagrammatically

Conceptual Framework Origin

The term conceptual framework appears to have originated in philosophy and systems theory, being used for the first time in the 1930s by the philosopher Alfred North Whitehead. He bridged the theological, social, and physical sciences by providing a common conceptual framework. The use of the conceptual framework began early in accountancy and can be traced back to publications by William A. Paton and John B. Canning in the first quarter of the 20 th century. Thus, in the original framework, financial issues were addressed, such as useful features, basic elements, and variables needed to prepare financial statements. Nevertheless, a conceptual framework approach should be considered when starting your research journey in any field, from finance to social sciences to applied sciences.

Purpose and Importance of a Conceptual Framework in Research

The importance of a conceptual framework in research cannot be understated, irrespective of the field of study. It is important for the following reasons:

• It clarifies the context of the study.
• It justifies the study to the reader.
• It helps you check your own understanding of the problem and the need for the study.
• It illustrates the expected relationship between the variables and defines the objectives for the research.
• It helps further refine the study objectives and choose the methods appropriate to meet them.

What to Include in a Conceptual Framework

Essential elements that a conceptual framework should include are as follows:

• Overarching research question(s)
• Study parameters
• Study variables
• Potential relationships between those variables.

The sources for these elements of a conceptual framework are literature, theory, and experience or prior knowledge.

How to Make a Conceptual Framework

Now that you know the essential elements, your next question will be how to make a conceptual framework.

For this, start by identifying the most suitable set of questions that your research aims to answer. Next, categorize the various variables. Finally, perform a rigorous analysis of the collected data and compile the final results to establish connections between the variables.

In short, the steps are as follows:

• Choose appropriate research questions.
• Define the different types of variables involved.
• Determine the cause-and-effect relationships.

Be sure to make use of arrows and lines to depict the presence or absence of correlational linkages among the variables.

Developing a Conceptual Framework

Researchers should be adept at developing a conceptual framework. Here are the steps for developing a conceptual framework:

1. Identify a research question

Your research question guides your entire study, making it imperative to invest time and effort in formulating a question that aligns with your research goals and contributes to the existing body of knowledge. This step involves the following:

• Choose a broad topic of interest
• Conduct background research
• Narrow down the focus
• Define your goals
• Make it specific and answerable
• Consider significance and novelty
• Seek feedback.

 2. Choose independent and dependent variables

The dependent variable is the main outcome you want to measure, explain, or predict in your study. It should be a variable that can be observed, measured, or assessed quantitatively or qualitatively. Independent variables are the factors or variables that may influence, explain, or predict changes in the dependent variable.

Choose independent and dependent variables for your study according to the research objectives, the nature of the phenomenon being studied, and the specific research design. The identification of variables is rooted in existing literature, theories, or your own observations.

3. Consider cause-and-effect relationships

To better understand and communicate the relationships between variables in your study, cause-and-effect relationships need to be visualized. This can be done by using path diagrams, cause-and-effect matrices, time series plots, scatter plots, bar charts, or heatmaps.

4. Identify other influencing variables

Besides the independent and dependent variables, researchers must understand and consider the following types of variables:

• Moderating variable: A variable that influences the strength or direction of the relationship between an independent variable and a dependent variable.
• Mediating variable: A variable that explains the relationship between an independent variable and a dependent variable and clarifies how the independent variable affects the dependent variable.
• Control variable: A variable that is kept constant or controlled to avoid the influence of other factors that may affect the relationship between the independent and dependent variables.
• Confounding variable: A type of unmeasured variable that is related to both the independent and dependent variables.

Example of a Conceptual Framework

Let us examine the following conceptual framework example. Let’s say your research topic is “ The Impact of Social Media Usage on Academic Performance among College Students .” Here, you want to investigate how social media usage affects academic performance in college students. Social media usage (encompassing frequency of social media use, time spent on social media platforms, and types of social media platforms used) is the independent variable, and academic performance (covering grades, exam scores, and class attendance) is the dependent variable.

This conceptual framework example also includes a mediating variable, study habits, which may explain how social media usage affects academic performance. Study habits (time spent studying, study environment, and use of study aids or resources) can act as a mechanism through which social media usage influences academic outcomes. Additionally, a moderating variable, self-discipline (level of self-control and self-regulation, ability to manage distractions, and prioritization skills), is included to examine how individual differences in self-control and discipline may influence the relationship between social media usage and academic performance.

Confounding variables are also identified (socioeconomic status, prior academic achievement), which are potential factors that may influence both social media usage and academic performance. These variables need to be considered and controlled in the study to ensure that any observed effects are specifically attributed to social media usage. A visual representation of this conceptual framework example is seen in Figure 2.

Figure 2: Visual representation of a conceptual framework for the topic “The Impact of Social Media Usage on Academic Performance among College Students”

Key Takeaways

Here is a snapshot of the basics of a conceptual framework in research:

• A conceptual framework is an idea or model representing the subject or phenomena you intend to study.
• It is primarily a researcher’s perception of the research problem. It can be used to develop hypotheses or testable research questions.
• It provides a preliminary understanding of the factors at play, their interrelationships, and the underlying reasons.
• It guides your research by aiding in the formulation of meaningful research questions, selection of appropriate methods, and identification of potential challenges to the validity of your findings.
• It provides a structure for organizing and understanding data.
• It allows you to chalk out the relationships between concepts and variables to understand them.
• Variables besides dependent and independent variables (moderating, mediating, control, and confounding variables) must be considered when developing a conceptual framework.

Frequently Asked Questions

What is the difference between a moderating variable and a mediating variable.

Moderating and mediating variables are easily confused. A moderating variable affects the direction and strength of this relationship, whereas a mediating explains how two variables relate.

What is the difference between independent variables, dependent variables, and confounding variables?

Independent variables are the variables manipulated to affect the outcome of an experiment (e.g., the dose of a fat-loss drug administered to rats). Dependent variables are variables being measured or observed in an experiment (e.g., changes in rat body weight as a result of the drug). A confounding variable distorts or masks the effects of the variables being studied because it is associated both with dependent variable and with the independent variable. For instance, in this example, pre-existing metabolic dysfunction in some rats could interact differently with the drug being studied and also affect rat body weight.

Should I have more than one dependent or independent variable in a study?

The need for more than one dependent or independent variable in a study depends on the research question, study design, and relationships being investigated. Note the following when making this decision for your research:

• If your research question involves exploring the relationships between multiple variables or factors, it may be appropriate to have more than one dependent or independent variable.
• If you have specific hypotheses about the relationships between several variables, it may be necessary to include multiple dependent or independent variables.
• Adequate resources, sample size, and data collection methods should be considered when determining the number of dependent and independent variables to include.

What is a confounding variable?

A confounding variable is not the main focus of the study but can unintentionally influence the relationship between the independent and dependent variables. Confounding variables can introduce bias and give rise to misleading conclusions. These variables must be controlled to ensure that any observed relationship is genuinely due to the independent variable.

What is a control variable?

A control variable is something not of interest to the study’s objectives but is kept constant because it could influence the outcomes. Control variables can help prevent research biases and allow for a more accurate assessment of the relationship between the independent and dependent variables. Examples are (i) testing all participants at the same time (e.g., in the morning) to minimize the potential effects of circadian rhythms, (ii) ensuring that instruments are calibrated consistently before each measurement to minimize the influence of measurement errors, and (iii) randomization of participants across study groups.

R Discovery is a literature search and research reading platform that accelerates your research discovery journey by keeping you updated on the latest, most relevant scholarly content. With 250M+ research articles sourced from trusted aggregators like CrossRef, Unpaywall, PubMed, PubMed Central, Open Alex and top publishing houses like Springer Nature, JAMA, IOP, Taylor & Francis, NEJM, BMJ, Karger, SAGE, Emerald Publishing and more, R Discovery puts a world of research at your fingertips.  

Try R Discovery Prime FREE for 1 week or upgrade at just US$72 a year to access premium features that let you listen to research on the go, read in your language, collaborate with peers, auto sync with reference managers, and much more. Choose a simpler, smarter way to find and read research – Download the app and start your free 7-day trial today !  

Related Posts

Research in Shorts: R Discovery’s New Feature Helps Academics Assess Relevant Papers in 2mins 

How Does R Discovery’s Interplatform Capability Enhance Research Accessibility 

The Ultimate Guide to Qualitative Research - Part 1: The Basics

• Introduction and overview
• What is qualitative research?
• What is qualitative data?
• Examples of qualitative data
• Qualitative vs. quantitative research
• Mixed methods
• Qualitative research preparation
• Theoretical perspective
• Theoretical framework
• Literature reviews
• Research question
• Introduction

Understanding conceptual frameworks

Selecting and developing your framework, variables in a conceptual framework.

• Conceptual vs. theoretical framework
• Data collection
• Qualitative research methods
• Focus groups
• Observational research
• Case studies
• Ethnographical research
• Ethical considerations
• Confidentiality and privacy
• Power dynamics
• Reflexivity

Conceptual framework: Definition and theory

Theoretical and conceptual frameworks ultimately go hand in hand, but while there is significant overlap with theoretical perspectives and theoretical frameworks, understanding the essential differences is important when designing your research project.

Let's explore the idea of a conceptual framework, provide a few common examples, and discuss how to choose a framework for your study. Keep in mind that a conceptual framework will differ from a theoretical framework and that we will explore these differences in the next section.

In this section, we'll delve into the intricacies of conceptual frameworks and their role in qualitative research . They are essentially the scaffolding on which you hang your research questions and analysis . They define the concepts that you'll study and articulate the relationships among them.

Developing conceptual frameworks in research

At the most basic level, a conceptual framework is a visual or written product that explains, either graphically or in narrative form, the main things to be studied, the key factors, variables, or constructs, and any presumed relationships among them. It acts as a road map guiding the course of your research, directing what will be studied, and helping to organize and analyze the data.

The purpose of a conceptual framework

A conceptual framework serves multiple functions in a research project. It helps in clarifying the research problem and purpose, assists in refining the research questions, and guides the data collection and analysis process. It's the tool that ties all aspects of the study together, offering a coherent perspective for the researcher and readers to understand the research more holistically.

Relation between theoretical perspectives and conceptual frameworks

Theoretical perspectives offer overarching philosophies and assumptions that guide the research process, while conceptual frameworks are the specific devices that are derived from these perspectives to operationalize the study. If a theoretical perspective is the broad philosophical underpinning, a conceptual framework is a pragmatic approach that puts that philosophy into practice in the context of the study.

For instance, if you're working from a feminist theoretical perspective, your conceptual framework might involve specific constructs like gender roles, power dynamics , and societal norms, as well as the relationships between these constructs. The conceptual framework would be the lens through which you examine and interpret your data, guided by your theoretical perspective.

Critical theory

Critical theory is a theoretical perspective that seeks to confront social, historical, and ideological forces and structures that produce and constrain social problems. The corresponding conceptual framework might focus on constructs like power relations, historical context, and societal structures. For instance, a study on income inequality might have a conceptual framework involving constructs of socioeconomic status, institutional policies, and the distribution of resources.

Feminist theory

Feminist theory emphasizes the societal roles of gender and power relationships. A conceptual framework derived from this theory might involve constructs like gender roles, power dynamics, and societal norms. In a study about gender representation in media, a feminist conceptual framework could involve constructs such as stereotyping, representation, and societal expectations of gender.

Design your study with ATLAS.ti

Data analysis starts with a solid study design. Make it happen with ATLAS.ti, starting with a free trial.

Choosing and developing your conceptual framework is a pivotal process in your research design. This framework will help guide your study, inform your methodology , and shape your analysis .

Factors to consider when choosing a framework

Your conceptual framework should be derived from and align with your chosen theoretical perspective , but there are other considerations as well. It should resonate with your research question , problem, or purpose and be applicable to the specific context or population you are studying. You should also consider the feasibility of operationalizing the constructs in your framework.

When selecting a conceptual framework, consider the following questions:

1. How does this framework relate to my research topic? 2. Can I use this framework to effectively address my research question(s)? 3. Does this framework resonate with the population and context I'm studying? 4. Can the constructs in this framework be feasibly operationalized in my study?

Steps in developing a conceptual framework

Developing your conceptual framework involves a few key steps:

1. Identify key constructs: Based on your theoretical perspective and research question(s) , what are the main constructs or variables that you need to explore in your study? 2. Clarify relationships among constructs: How do these constructs relate to each other? Are there presumed causal relationships, correlations, or other types of associations? 3. Define each construct: Clearly define what each construct means in the context of your study. This might also involve operationalizing each construct or defining the indicators you will use to measure or identify each construct. 4. Create a visual representation : It is often extremely helpful to create a visual representation of your conceptual framework to illustrate the constructs and their relationships. Map out the relationships among constructs to develop a holistic understanding of what you want to study.

Remember, your conceptual framework is not set in stone. You can start creating your conceptual framework based on your literature review and your own critical reflections. As you proceed with your study, you might need to refine or adapt your conceptual framework based on what you're learning from your data. Developing a robust framework is an iterative process that requires critical thinking, creativity, and flexibility.

A strong conceptual framework includes variables that refer to the constructs or characteristics that are being studied. They are the building blocks of your research study. It might be helpful to think about how the variables in your conceptual framework could be categorized as independent and dependent variables, which respectively influence and are influenced within the research study.

Independent variables and dependent variables

An independent variable is the characteristic or condition that is manipulated or selected by the researcher to determine its effect on the dependent variable. For example, in a study exploring the impact of classroom size on student engagement, classroom size would be the independent variable.

The dependent variable is the main outcome that the researcher is interested in studying or explaining. In the example given above, student engagement would be the dependent variable, as it's the outcome being observed for any changes in response to the independent variable (classroom size). In essence, defining these variables can help you identify the cause-and-effect relationships in your study. While it might be difficult to know beforehand exactly which variables will be important and how they relate to one another, this is a helpful thought exercise to flesh out potential relationships among variables you may want to study.

Relationships among variables

Within a conceptual framework, the dependent and independent variables are listed in addition to their proposed relationships to each other. The ways in which these variables influence one another form the crux of the propositions or assumptions that guide your research.

In a conceptual framework based on the theoretical perspective of constructivism, for instance, the independent variable might be a teaching method (as constructivists would argue that methods of instruction can shape learning), and the dependent variable could be the depth of student understanding. The proposed relationship between these variables might be that student-centered teaching methods lead to a deeper understanding, which would guide the data collection and analysis such that this proposition could be explored.

However, it is important to note that the terminology of independent and dependent variables is more typical of quantitative research , in which independent and dependent variables are operationalized in hypotheses that will be tested based on pre-established theory. In qualitative research , the relationships between variables are more fluid and open-ended because the focus is often more on understanding the phenomenon as a whole and building a contextualized understanding of the research problem. This can involve including new or unexpected variables and interrelationships that emerge during the study, thus extending previous theory or understanding that didn’t initially predict these relationships.

Thus, in your conceptual framework, rather than solely focusing on identifying independent and dependent variables, consider how various factors interact and influence one another within the context of your study. Your conceptual framework should provide a holistic picture of the complexity of the phenomenon you are studying.

Ready to jumpstart your research with ATLAS.ti?

Conceptualize your research project with our intuitive data analysis interface. Download a free trial today.

What is a good example of a conceptual framework?

Last updated

18 April 2023

Reviewed by

Miroslav Damyanov

Short on time? Get an AI generated summary of this article instead

A well-designed study doesn’t just happen. Researchers work hard to ensure the studies they conduct will be scientifically valid and will advance understanding in their field.

Make research less tedious

Dovetail streamlines research to help you uncover and share actionable insights

• The importance of a conceptual framework

The main purpose of a conceptual framework is to improve the quality of a research study. A conceptual framework achieves this by identifying important information about the topic and providing a clear roadmap for researchers to study it.

Through the process of developing this information, researchers will be able to improve the quality of their studies in a few key ways.

Clarify research goals and objectives

A conceptual framework helps researchers create a clear research goal. Research projects often become vague and lose their focus, which makes them less useful. However, a well-designed conceptual framework helps researchers maintain focus. It reinforces the project’s scope, ensuring it stays on track and produces meaningful results.

Provide a theoretical basis for the study

Forming a hypothesis requires knowledge of the key variables and their relationship to each other. Researchers need to identify these variables early on to create a conceptual framework. This ensures researchers have developed a strong understanding of the topic before finalizing the study design. It also helps them select the most appropriate research and analysis methods.

Guide the research design

As they develop their conceptual framework, researchers often uncover information that can help them further refine their work.

Here are some examples:

Confounding variables they hadn’t previously considered

Sources of bias they will have to take into account when designing the project

Whether or not the information they were going to study has already been covered—this allows them to pivot to a more meaningful goal that brings new and relevant information to their field

• Steps to develop a conceptual framework

There are four major steps researchers will follow to develop a conceptual framework. Each step will be described in detail in the sections that follow. You’ll also find examples of how each might be applied in a range of fields.

Step 1: Choose the research question

The first step in creating a conceptual framework is choosing a research question . The goal of this step is to create a question that’s specific and focused.

By developing a clear question, researchers can more easily identify the variables they will need to account for and keep their research focused. Without it, the next steps will be more difficult and less effective.

Here are some examples of good research questions in a few common fields:

Natural sciences: How does exposure to ultraviolet radiation affect the growth rate of a particular type of algae?

Health sciences: What is the effectiveness of cognitive-behavioral therapy for treating depression in adolescents?

Business: What factors contribute to the success of small businesses in a particular industry?

Education: How does implementing technology in the classroom impact student learning outcomes?

Step 2: Select the independent and dependent variables

Once the research question has been chosen, it’s time to identify the dependent and independent variables .

The independent variable is the variable researchers think will affect the dependent variable . Without this information, researchers cannot develop a meaningful hypothesis or design a way to test it.

The dependent and independent variables for our example questions above are:

Natural sciences

Independent variable: exposure to ultraviolet radiation

Dependent variable: the growth rate of a particular type of algae

Health sciences

Independent variable: cognitive-behavioral therapy

Dependent variable: depression in adolescents

Independent variables: factors contributing to the business’s success

Dependent variable: sales, return on investment (ROI), or another concrete metric

Independent variable: implementation of technology in the classroom

Dependent variable: student learning outcomes, such as test scores, GPAs, or exam results

Step 3: Visualize the cause-and-effect relationship

This step is where researchers actually develop their hypothesis. They will predict how the independent variable will impact the dependent variable based on their knowledge of the field and their intuition.

With a hypothesis formed, researchers can more accurately determine what data to collect and how to analyze it. They will then visualize their hypothesis by creating a diagram. This visualization will serve as a framework to help guide their research.

The diagrams for our examples might be used as follows:

Natural sciences : how exposure to radiation affects the biological processes in the algae that contribute to its growth rate

Health sciences : how different aspects of cognitive behavioral therapy can affect how patients experience symptoms of depression

Business : how factors such as market demand, managerial expertise, and financial resources influence a business’s success

Education : how different types of technology interact with different aspects of the learning process and alter student learning outcomes

Step 4: Identify other influencing variables

The independent and dependent variables are only part of the equation. Moderating, mediating, and control variables are also important parts of a well-designed study. These variables can impact the relationship between the two main variables and must be accounted for.

A moderating variable is one that can change how the independent variable affects the dependent variable. A mediating variable explains the relationship between the two. Control variables are kept the same to eliminate their impact on the results. Examples of each are given below:

Moderating variable: water temperature (might impact how algae respond to radiation exposure)

Mediating variable: chlorophyll production (might explain how radiation exposure affects algae growth rate)

Control variable: nutrient levels in the water

Moderating variable: the severity of depression symptoms at baseline might impact how effective the therapy is for different adolescents

Mediating variable: social support might explain how cognitive-behavioral therapy leads to improvements in depression

Control variable: other forms of treatment received before or during the study

Moderating variable: the size of the business (might impact how different factors contribute to market share, sales, ROI, and other key success metrics)

Mediating variable: customer satisfaction (might explain how different factors impact business success)

Control variable: industry competition

Moderating variable: student age (might impact how effective technology is for different students)

Mediating variable: teacher training (might explain how technology leads to improvements in learning outcomes)

Control variable: student learning style

• Conceptual versus theoretical frameworks

Although they sound similar, conceptual and theoretical frameworks have different goals and are used in different contexts. Understanding which to use will help researchers craft better studies.

Conceptual frameworks describe a broad overview of the subject and outline key concepts, variables, and the relationships between them. They provide structure to studies that are more exploratory in nature, where the relationships between the variables are still being established. They are particularly helpful in studies that are complex or interdisciplinary because they help researchers better organize the factors involved in the study.

Theoretical frameworks, on the other hand, are used when the research question is more clearly defined and there’s an existing body of work to draw upon. They define the relationships between the variables and help researchers predict outcomes. They are particularly helpful when researchers want to refine the existing body of knowledge rather than establish it.

Should you be using a customer insights hub?

Do you want to discover previous research faster?

Do you share your research findings with others?

Do you analyze research data?

Start for free today, add your research, and get to key insights faster

Editor’s picks

Last updated: 18 April 2023

Last updated: 27 February 2023

Last updated: 22 August 2024

Last updated: 5 February 2023

Last updated: 16 August 2024

Last updated: 9 March 2023

Last updated: 30 April 2024

Last updated: 12 December 2023

Last updated: 11 March 2024

Last updated: 4 July 2024

Last updated: 6 March 2024

Last updated: 5 March 2024

Last updated: 13 May 2024

Latest articles

Related topics, .css-je19u9{-webkit-align-items:flex-end;-webkit-box-align:flex-end;-ms-flex-align:flex-end;align-items:flex-end;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-flex-direction:row;-ms-flex-direction:row;flex-direction:row;-webkit-box-flex-wrap:wrap;-webkit-flex-wrap:wrap;-ms-flex-wrap:wrap;flex-wrap:wrap;-webkit-box-pack:center;-ms-flex-pack:center;-webkit-justify-content:center;justify-content:center;row-gap:0;text-align:center;max-width:671px;}@media (max-width: 1079px){.css-je19u9{max-width:400px;}.css-je19u9>span{white-space:pre;}}@media (max-width: 799px){.css-je19u9{max-width:400px;}.css-je19u9>span{white-space:pre;}} decide what to .css-1kiodld{max-height:56px;display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}@media (max-width: 1079px){.css-1kiodld{display:none;}} build next, decide what to build next, log in or sign up.

Get started for free

Conceptual review papers: revisiting existing research to develop and refine theory

• Theory/Conceptual
• Published: 29 April 2020
• Volume 10 , pages 27–35, ( 2020 )

Cite this article

• John Hulland 1  

6359 Accesses

66 Citations

8 Altmetric

Explore all metrics

Conceptual review papers can theoretically enrich the field of marketing by reviewing extant knowledge, noting tensions and inconsistencies, identifying important gaps as well as key insights, and proposing agendas for future research. The result of this process is a theoretical contribution that refines, reconceptualizes, or even replaces existing ways of viewing a phenomenon. This paper spells out the primary aims of conceptual reviews and clarifies how they differ from other theory development efforts. It also describes elements essential to a strong conceptual review paper and offers a specific set of best practices that can be used to distinguish a strong conceptual review from a weak one.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Meta-analysis: integrating accumulated knowledge.

Designing conceptual articles: four approaches

Contours of the marketing literature: text, context, point-of-view, research horizons, interpretation, and influence in marketing, explore related subjects.

• Artificial Intelligence

Palmatier et al. ( 2018 ) reference a study of the frequency with which review papers were published in top marketing journals during the 2012–2016 period. Focusing on the top six journals included in the Financial Times (( FT-50 ) journal list, the study found that “ JAMS has become the most common outlet … publishing 31% of all review papers that appeared in the top six marketing journals.”

The bifurcation here between theory development “from scratch” versus through conceptual review is potentially somewhat misleading, since the latter can also result in novel theoretical insights. Furthermore, many conceptual papers make significant theoretical contributions by building on existing theory without themselves being review papers. Nonetheless, conceptual reviews necessarily involve working with extant, published work.

This focus is quite distinct from the approach proposed by Zeithaml et al. ( 2020 ). Their emphasis is on “an approach that is ideally suited to the development of theories in marketing: the ‘theories-in-use’ (TIU) approach” (p. 32). They propose it as an alternative inductive methodology (vs. case studies and ethnographies) to developing grounded theory.

These elements are drawn from Hulland & Houston ( 2020 ), MacInnis ( 2011 ), Palmatier et al. ( 2018 ), and Yadav ( 2010 ). Houston ( 2020 ), MacInnis ( 2011 ), Palmatier, Houston & Hulland et al. ( 2018 ), and Yadav ( 2010 ).

These underlying assumptions are a crucial component in developing strong arguments for theory development (Toulmin 1958 ).

MacInnis ( 2011 ) describes eight critical skills for conceptual thinking that are arrayed across four dimensions: envisioning (identifying vs. revising), explicating (delineating vs. summarizing), relating (differentiating vs. integrating, and debating (advocating vs. refuting). For conceptual review papers, summarizing and revising represent critical skills that need to be harnessed by the author (whereas identifying and delineating are skills more critical to uncovering new ideas). For the other two dimensions (relating and debating), a more balanced use of the associated skills is needed (i.e., both differentiating and integrating are important, and both advocating and refuting are important).

In her paper, Jaakkola ( 2020 ) describes four different types of research designs for conceptual reviews: (1) theory synthesis, (2) theory adaptation, (3) typology, and (4) model. In the current paper, elements from all four of these types are discussed.

In doing so, Khamitov et al. discover seven overarching insights that reveal gaps in the interfaces between the three streams. This highlighting of gaps represents stage four in the theory refinement process.

Not all of the gaps in a specific domain are necessarily valuable, however. Just because no one has studied a phenomenon in a particular industry or region, or with a particular method does not mean that a filling of that gap is required (or even valued).

Antonakis, J., Bartardox, N., Liu, Y., & Schriesheim, C. A. (2014). What makes articles highly cited? The Leadership Quarterly, 25 (1), 152–179.

Article   Google Scholar  

Barczak, G. (2017). From the editor: Writing a review article. Journal of Product Innovation Management, 34 (2), 120–121.

Bem, D.J. (1995). Writing a review article for Psychological Bulletin . Psychological Bulletin , 118(2), 172–177.

Bettencourt, L. A., & Houston, M. B. (2001). Assessing the impact of article method type and subject area on citation frequency and reference diversity. Marketing Letters, 12 (4), 327–340.

Dekimpe, M. G., & Deleersnyder, B. (2018). Business cycle research in marketing: A review and research agenda. Journal of the Academy of Marketing Science, 46 (1), 31–58.

Dowling, K., Guhl, D., Klapper, D., Spann, M., Stich, L., & Yegoryan, N. (2020). Behavioral biases in marketing. Journal of the Academy of Marketing Science in press , 48 (3), 449–477.

Gilson, L. L., & Goldberg, C. B. (2015). Editors’ comment: So, what is a conceptual paper? Group & Organization Management, 40 (2), 127–130.

Grewal, D., Puccinelli, N. M., & Monroe, K. B. (2018). Meta-analysis: Integrating accumulated knowledge. Journal of the Academy of Marketing Science, 46 (1), 9–30.

Houston, M. B. (2019). Four facets of rigor. Journal of the Academy of Marketing Science, 47 (4), 570–573.

Hulland, J., & Houston, M. B. (2020). Why systematic review papers and meta-analyses matter: An introduction to the special issue on generalizations in marketing. Journal of the Academy of Marketing Science, 48 (3) in press, 351–359.

Hulland, J., Baumgartner, H., & Smith, K. M. (2018). Marketing survey research best practices: Evidence and recommendations from a review of JAMS articles. Journal of the Academy of Marketing Science, 46 (1), 92–108.

Jaakkola, E. (2020). Designing conceptual articles: Four approaches. Journal of the Academy of Marketing Science in press.

Khamitov, M., Gregoire, Y., & Suri, A. (2020). A systematic review of brand transgression, service failure recovery and product-harm crisis: Integration and guiding insights. Journal of the Academy of Marketing Science in press , 48 (3), 519–542.

Kozlenkova, I. V., Samaha, S. A., & Palmatier, R. W. (2014). Resource-based theory in marketing. Journal of the Academy of Marketing Science, 42 (1), 1–21.

Lamberton, C., & Stephen, A. T. (2016). A thematic exploration of digital, social media, and mobile marketing: Research evolution from 2000 to 2015 and an agenda for future inquiry. Journal of Marketing, 80 (November), 146–172.

Littell, J. H., Corcoran, J., & Pillai, V. (2008). Systematic reviews and meta-analysis . New York: Oxford University Press.

MacInnis, D. J. (2011). A framework for conceptual contributions in marketing. Journal of Marketing, 75 (July), 136–154.

Palmatier, R. W. (2016). Improving publishing success at JAMS : Contribution and positioning. Journal of the Academy of Marketing Science, 44 (6), 655–659.

Palmatier, R. W., Houston, M. B., & Hulland, J. (2018). Review articles: Purpose, process, and structure. Journal of the Academy of Marketing Science, 46 (1), 1–5.

Rindfleisch, A., & Heide, J. B. (1997). Transaction cost analysis: Past, present, and future applications. Journal of Marketing, 61 (4), 30–54.

Rosario, A. B., de Valck, K., & Sotgiu, F. (2020). Conceptualizing the electronic word-of-mouth process: What we know and need to know about eWOM creation, exposure, and evaluation. Journal of the Academy of Marketing Science in press.

Samiee, S. (1994). Customer evaluation of products in a global market. Journal of International Business Studies, 25 (3), 579–604.

Sample, K. L., Hagtvedt, H., & Brasel, S. A. (2020). Components of visual perception in marketing contexts: A conceptual framework and review. Journal of the Academy of Marketing Science in press , 48 (3), 405–421.

Short, J. (2009). The art of writing a review article. Journal of Management, 35 (6), 1312–1317.

Sorescu, A., Warren, N. L., & Ertekin, L. (2017). Event study methodology in the marketing literature: An overview. Journal of the Academy of Marketing Science, 45 (2), 186–207.

Steinhoff, L., Arli, D., Weaven, S., & Kozlenkova, I. V. (2019). Online relationship marketing. Journal of the Academy of Marketing Science, 47 (3), 369–393.

Stewart, D. W., & Zinkhan, G. M. (2006). Enhancing marketing theory in academic research. Journal of the Academy of Marketing Science, 34 (Fall), 477–480.

Sutton, R. I., & Staw, B. M. (1995). What theory is not . Administrative Science Quarterly, 40 (3), 371–384.

Toulmin, S. (1958). The uses of argument . Cambridge, U.K.: Cambridge University Press.

Google Scholar  

Wade, M., & Hulland, J. (2004). The resource-based view and information systems research: Review, extension, and suggestions for future research. MIS Quarterly, 28 (1), 107–142.

Yadav, M. S. (2010). The decline of conceptual articles and implications for knowledge development. Journal of Marketing, 74 (January), 1–19.

Zeithaml, V. A., Jaworski, B. J., Kohli, A. K., Tuli, K. R., Ulaga, W., & Zaltman, G. (2020). A theories-in-use approach to building marketing theory. Journal of Marketing, 84 (1), 32–51.

Download references

Author information

Authors and affiliations.

Terry College of Business, University of Georgia, Athens, GA, 30602, USA

John Hulland

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to John Hulland .

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Hulland, J. Conceptual review papers: revisiting existing research to develop and refine theory. AMS Rev 10 , 27–35 (2020). https://doi.org/10.1007/s13162-020-00168-7

Download citation

Received : 11 March 2020

Accepted : 01 April 2020

Published : 29 April 2020

Issue Date : June 2020

DOI : https://doi.org/10.1007/s13162-020-00168-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Conceptual review papers
• Marketing theory
• Find a journal
• Publish with us
• Track your research

• Link to facebook
• Link to linkedin
• Link to twitter
• Link to youtube
• Writing Tips

How to Make a Conceptual Framework

6-minute read

• 2nd January 2022

What is a conceptual framework? And why is it important?

A conceptual framework illustrates the relationship between the variables of a research question. It’s an outline of what you’d expect to find in a research project.

Conceptual frameworks should be constructed before data collection and are vital because they map out the actions needed in the study. This should be the first step of an undergraduate or graduate research project.

What Is In a Conceptual Framework?

In a conceptual framework, you’ll find a visual representation of the key concepts and relationships that are central to a research study or project . This can be in form of a diagram, flow chart, or any other visual representation. Overall, a conceptual framework serves as a guide for understanding the problem being studied and the methods being used to investigate it.

Steps to Developing the Perfect Conceptual Framework

• Pick a question
• Conduct a literature review
• Identify your variables
• Create your conceptual framework

1. Pick a Question

You should already have some idea of the broad area of your research project. Try to narrow down your research field to a manageable topic in terms of time and resources. From there, you need to formulate your research question. A research question answers the researcher’s query: “What do I want to know about my topic?” Research questions should be focused, concise, arguable and, ideally, should address a topic of importance within your field of research.

An example of a simple research question is: “What is the relationship between sunny days and ice cream sales?”

2. Conduct a Literature Review

A literature review is an analysis of the scholarly publications on a chosen topic. To undertake a literature review, search for articles with the same theme as your research question. Choose updated and relevant articles to analyze and use peer-reviewed and well-respected journals whenever possible.

For the above example, the literature review would investigate publications that discuss how ice cream sales are affected by the weather. The literature review should reveal the variables involved and any current hypotheses about this relationship.

3. Identify Your Variables

There are two key variables in every experiment: independent and dependent variables.

Independent Variables

The independent variable (otherwise known as the predictor or explanatory variable) is the expected cause of the experiment: what the scientist changes or changes on its own. In our example, the independent variable would be “the number of sunny days.”

Dependent Variables

The dependent variable (otherwise known as the response or outcome variable) is the expected effect of the experiment: what is being studied or measured. In our example, the dependent variable would be “the quantity of ice cream sold.”

Next, there are control variables.

Control Variables

A control variable is a variable that may impact the dependent variable but whose effects are not going to be measured in the research project. In our example, a control variable could be “the socioeconomic status of participants.” Control variables should be kept constant to isolate the effects of the other variables in the experiment.

Finally, there are intervening and extraneous variables.

Find this useful?

Subscribe to our newsletter and get writing tips from our editors straight to your inbox.

Intervening Variables

Intervening variables link the independent and dependent variables and clarify their connection. In our example, an intervening variable could be “temperature.”

Extraneous Variables

Extraneous variables are any variables that are not being investigated but could impact the outcomes of the study. Some instances of extraneous variables for our example would be “the average price of ice cream” or “the number of varieties of ice cream available.” If you control an extraneous variable, it becomes a control variable.

4. Create Your Conceptual Framework

Having picked your research question, undertaken a literature review, and identified the relevant variables, it’s now time to construct your conceptual framework. Conceptual frameworks are clear and often visual representations of the relationships between variables.

We’ll start with the basics: the independent and dependent variables.

Our hypothesis is that the quantity of ice cream sold directly depends on the number of sunny days; hence, there is a cause-and-effect relationship between the independent variable (the number of sunny days) and the dependent and independent variable (the quantity of ice cream sold).

Next, introduce a control variable. Remember, this is anything that might directly affect the dependent variable but is not being measured in the experiment:

Finally, introduce the intervening and extraneous variables. 

The intervening variable (temperature) clarifies the relationship between the independent variable (the number of sunny days) and the dependent variable (the quantity of ice cream sold). Extraneous variables, such as the average price of ice cream, are variables that are not controlled and can potentially impact the dependent variable.

Are Conceptual Frameworks and Research Paradigms the Same?

In simple terms, the research paradigm is what informs your conceptual framework. In defining our research paradigm we ask the big questions—Is there an objective truth and how can we understand it? If we decide the answer is yes, we may be working with a positivist research paradigm and will choose to build a conceptual framework that displays the relationship between fixed variables. If not, we may be working with a constructivist research paradigm, and thus our conceptual framework will be more of a loose amalgamation of ideas, theories, and themes (a qualitative study). If this is confusing–don’t worry! We have an excellent blog post explaining research paradigms in more detail.

Where is the Conceptual Framework Located in a Thesis?

This will depend on your discipline, research type, and school’s guidelines, but most papers will include a section presenting the conceptual framework in the introduction, literature review, or opening chapter. It’s best to present your conceptual framework after presenting your research question, but before outlining your methodology.

Can a Conceptual Framework be Used in a Qualitative Study?

Yes. Despite being less clear-cut than a quantitative study, all studies should present some form of a conceptual framework. Let’s say you were doing a study on care home practices and happiness, and you came across a “happiness model” constructed by a relevant theorist in your literature review. Your conceptual framework could be an outline or a visual depiction of how you will use this model to collect and interpret qualitative data for your own study (such as interview responses). Check out this useful resource showing other examples of conceptual frameworks for qualitative studies .

Expert Proofreading for Researchers

Whether you’re a seasoned academic or not, you will want your research paper to be error-free and fluently written. That’s where proofreading comes in. Our editors are on hand 24 hours a day to ensure your writing is concise, clear, and precise. Submit a free sample of your writing today to try our services.

Share this article:

Post A New Comment

Got content that needs a quick turnaround? Let us polish your work. Explore our editorial business services.

5-minute read

Free Email Newsletter Template (2024)

Promoting a brand means sharing valuable insights to connect more deeply with your audience, and...

How to Write a Nonprofit Grant Proposal

If you’re seeking funding to support your charitable endeavors as a nonprofit organization, you’ll need...

9-minute read

How to Use Infographics to Boost Your Presentation

Is your content getting noticed? Capturing and maintaining an audience’s attention is a challenge when...

8-minute read

Why Interactive PDFs Are Better for Engagement

Are you looking to enhance engagement and captivate your audience through your professional documents? Interactive...

7-minute read

Seven Key Strategies for Voice Search Optimization

Voice search optimization is rapidly shaping the digital landscape, requiring content professionals to adapt their...

4-minute read

Five Creative Ways to Showcase Your Digital Portfolio

Are you a creative freelancer looking to make a lasting impression on potential clients or...

Make sure your writing is the best it can be with our expert English proofreading and editing.

Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, automatically generate references for free.

• Knowledge Base
• Methodology
• What Is a Conceptual Framework? | Tips & Examples

What Is a Conceptual Framework? | Tips & Examples

Published on 4 May 2022 by Bas Swaen and Tegan George. Revised on 18 March 2024.

A conceptual framework illustrates the expected relationship between your variables. It defines the relevant objectives for your research process and maps out how they come together to draw coherent conclusions.

Keep reading for a step-by-step guide to help you construct your own conceptual framework.

Table of contents

Developing a conceptual framework in research, step 1: choose your research question, step 2: select your independent and dependent variables, step 3: visualise your cause-and-effect relationship, step 4: identify other influencing variables, frequently asked questions about conceptual models.

A conceptual framework is a representation of the relationship you expect to see between your variables, or the characteristics or properties that you want to study.

Conceptual frameworks can be written or visual and are generally developed based on a literature review of existing studies about your topic.

Your research question guides your work by determining exactly what you want to find out, giving your research process a clear focus.

However, before you start collecting your data, consider constructing a conceptual framework. This will help you map out which variables you will measure and how you expect them to relate to one another.

In order to move forward with your research question and test a cause-and-effect relationship, you must first identify at least two key variables: your independent and dependent variables .

• The expected cause, ‘hours of study’, is the independent variable (the predictor, or explanatory variable)
• The expected effect, ‘exam score’, is the dependent variable (the response, or outcome variable).

Note that causal relationships often involve several independent variables that affect the dependent variable. For the purpose of this example, we’ll work with just one independent variable (‘hours of study’).

Now that you’ve figured out your research question and variables, the first step in designing your conceptual framework is visualising your expected cause-and-effect relationship.

It’s crucial to identify other variables that can influence the relationship between your independent and dependent variables early in your research process.

Some common variables to include are moderating, mediating, and control variables.

Moderating variables

Moderating variable (or moderators) alter the effect that an independent variable has on a dependent variable. In other words, moderators change the ‘effect’ component of the cause-and-effect relationship.

Let’s add the moderator ‘IQ’. Here, a student’s IQ level can change the effect that the variable ‘hours of study’ has on the exam score. The higher the IQ, the fewer hours of study are needed to do well on the exam.

Let’s take a look at how this might work. The graph below shows how the number of hours spent studying affects exam score. As expected, the more hours you study, the better your results. Here, a student who studies for 20 hours will get a perfect score.

But the graph looks different when we add our ‘IQ’ moderator of 120. A student with this IQ will achieve a perfect score after just 15 hours of study.

Below, the value of the ‘IQ’ moderator has been increased to 150. A student with this IQ will only need to invest five hours of study in order to get a perfect score.

Here, we see that a moderating variable does indeed change the cause-and-effect relationship between two variables.

Mediating variables

Now we’ll expand the framework by adding a mediating variable . Mediating variables link the independent and dependent variables, allowing the relationship between them to be better explained.

Here’s how the conceptual framework might look if a mediator variable were involved:

In this case, the mediator helps explain why studying more hours leads to a higher exam score. The more hours a student studies, the more practice problems they will complete; the more practice problems completed, the higher the student’s exam score will be.

Moderator vs mediator

It’s important not to confuse moderating and mediating variables. To remember the difference, you can think of them in relation to the independent variable:

• A moderating variable is not affected by the independent variable, even though it affects the dependent variable. For example, no matter how many hours you study (the independent variable), your IQ will not get higher.
• A mediating variable is affected by the independent variable. In turn, it also affects the dependent variable. Therefore, it links the two variables and helps explain the relationship between them.

Control variables

Lastly,  control variables must also be taken into account. These are variables that are held constant so that they don’t interfere with the results. Even though you aren’t interested in measuring them for your study, it’s crucial to be aware of as many of them as you can be.

A mediator variable explains the process through which two variables are related, while a moderator variable affects the strength and direction of that relationship.

No. The value of a dependent variable depends on an independent variable, so a variable cannot be both independent and dependent at the same time. It must be either the cause or the effect, not both.

Yes, but including more than one of either type requires multiple research questions .

For example, if you are interested in the effect of a diet on health, you can use multiple measures of health: blood sugar, blood pressure, weight, pulse, and many more. Each of these is its own dependent variable with its own research question.

You could also choose to look at the effect of exercise levels as well as diet, or even the additional effect of the two combined. Each of these is a separate independent variable .

To ensure the internal validity of an experiment , you should only change one independent variable at a time.

A control variable is any variable that’s held constant in a research study. It’s not a variable of interest in the study, but it’s controlled because it could influence the outcomes.

A confounding variable , also called a confounder or confounding factor, is a third variable in a study examining a potential cause-and-effect relationship.

A confounding variable is related to both the supposed cause and the supposed effect of the study. It can be difficult to separate the true effect of the independent variable from the effect of the confounding variable.

In your research design , it’s important to identify potential confounding variables and plan how you will reduce their impact.

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the ‘Cite this Scribbr article’ button to automatically add the citation to our free Reference Generator.

Swaen, B. & George, T. (2024, March 18). What Is a Conceptual Framework? | Tips & Examples. Scribbr. Retrieved 3 September 2024, from https://www.scribbr.co.uk/research-methods/conceptual-frameworks/

Is this article helpful?

Other students also liked

Mediator vs moderator variables | differences & examples, independent vs dependent variables | definition & examples, what are control variables | definition & examples.

Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, generate accurate citations for free.

• Knowledge Base

Methodology

• How to Write a Literature Review | Guide, Examples, & Templates

How to Write a Literature Review | Guide, Examples, & Templates

Published on January 2, 2023 by Shona McCombes . Revised on September 11, 2023.

What is a literature review? A literature review is a survey of scholarly sources on a specific topic. It provides an overview of current knowledge, allowing you to identify relevant theories, methods, and gaps in the existing research that you can later apply to your paper, thesis, or dissertation topic .

There are five key steps to writing a literature review:

• Search for relevant literature
• Evaluate sources
• Identify themes, debates, and gaps
• Outline the structure
• Write your literature review

A good literature review doesn’t just summarize sources—it analyzes, synthesizes , and critically evaluates to give a clear picture of the state of knowledge on the subject.

Instantly correct all language mistakes in your text

Upload your document to correct all your mistakes in minutes

Table of contents

What is the purpose of a literature review, examples of literature reviews, step 1 – search for relevant literature, step 2 – evaluate and select sources, step 3 – identify themes, debates, and gaps, step 4 – outline your literature review’s structure, step 5 – write your literature review, free lecture slides, other interesting articles, frequently asked questions, introduction.

• Quick Run-through
• Step 1 & 2

When you write a thesis , dissertation , or research paper , you will likely have to conduct a literature review to situate your research within existing knowledge. The literature review gives you a chance to:

• Demonstrate your familiarity with the topic and its scholarly context
• Develop a theoretical framework and methodology for your research
• Position your work in relation to other researchers and theorists
• Show how your research addresses a gap or contributes to a debate
• Evaluate the current state of research and demonstrate your knowledge of the scholarly debates around your topic.

Writing literature reviews is a particularly important skill if you want to apply for graduate school or pursue a career in research. We’ve written a step-by-step guide that you can follow below.

Receive feedback on language, structure, and formatting

Professional editors proofread and edit your paper by focusing on:

• Academic style
• Vague sentences
• Style consistency

See an example

Writing literature reviews can be quite challenging! A good starting point could be to look at some examples, depending on what kind of literature review you’d like to write.

• Example literature review #1: “Why Do People Migrate? A Review of the Theoretical Literature” ( Theoretical literature review about the development of economic migration theory from the 1950s to today.)
• Example literature review #2: “Literature review as a research methodology: An overview and guidelines” ( Methodological literature review about interdisciplinary knowledge acquisition and production.)
• Example literature review #3: “The Use of Technology in English Language Learning: A Literature Review” ( Thematic literature review about the effects of technology on language acquisition.)
• Example literature review #4: “Learners’ Listening Comprehension Difficulties in English Language Learning: A Literature Review” ( Chronological literature review about how the concept of listening skills has changed over time.)

You can also check out our templates with literature review examples and sample outlines at the links below.

Download Word doc Download Google doc

Before you begin searching for literature, you need a clearly defined topic .

If you are writing the literature review section of a dissertation or research paper, you will search for literature related to your research problem and questions .

Make a list of keywords

Start by creating a list of keywords related to your research question. Include each of the key concepts or variables you’re interested in, and list any synonyms and related terms. You can add to this list as you discover new keywords in the process of your literature search.

• Social media, Facebook, Instagram, Twitter, Snapchat, TikTok
• Body image, self-perception, self-esteem, mental health
• Generation Z, teenagers, adolescents, youth

Search for relevant sources

Use your keywords to begin searching for sources. Some useful databases to search for journals and articles include:

• Your university’s library catalogue
• Google Scholar
• Project Muse (humanities and social sciences)
• Medline (life sciences and biomedicine)
• EconLit (economics)
• Inspec (physics, engineering and computer science)

You can also use boolean operators to help narrow down your search.

Make sure to read the abstract to find out whether an article is relevant to your question. When you find a useful book or article, you can check the bibliography to find other relevant sources.

You likely won’t be able to read absolutely everything that has been written on your topic, so it will be necessary to evaluate which sources are most relevant to your research question.

For each publication, ask yourself:

• What question or problem is the author addressing?
• What are the key concepts and how are they defined?
• What are the key theories, models, and methods?
• Does the research use established frameworks or take an innovative approach?
• What are the results and conclusions of the study?
• How does the publication relate to other literature in the field? Does it confirm, add to, or challenge established knowledge?
• What are the strengths and weaknesses of the research?

Make sure the sources you use are credible , and make sure you read any landmark studies and major theories in your field of research.

You can use our template to summarize and evaluate sources you’re thinking about using. Click on either button below to download.

Take notes and cite your sources

As you read, you should also begin the writing process. Take notes that you can later incorporate into the text of your literature review.

It is important to keep track of your sources with citations to avoid plagiarism . It can be helpful to make an annotated bibliography , where you compile full citation information and write a paragraph of summary and analysis for each source. This helps you remember what you read and saves time later in the process.

Don't submit your assignments before you do this

The academic proofreading tool has been trained on 1000s of academic texts. Making it the most accurate and reliable proofreading tool for students. Free citation check included.

Try for free

To begin organizing your literature review’s argument and structure, be sure you understand the connections and relationships between the sources you’ve read. Based on your reading and notes, you can look for:

• Trends and patterns (in theory, method or results): do certain approaches become more or less popular over time?
• Themes: what questions or concepts recur across the literature?
• Debates, conflicts and contradictions: where do sources disagree?
• Pivotal publications: are there any influential theories or studies that changed the direction of the field?
• Gaps: what is missing from the literature? Are there weaknesses that need to be addressed?

This step will help you work out the structure of your literature review and (if applicable) show how your own research will contribute to existing knowledge.

• Most research has focused on young women.
• There is an increasing interest in the visual aspects of social media.
• But there is still a lack of robust research on highly visual platforms like Instagram and Snapchat—this is a gap that you could address in your own research.

There are various approaches to organizing the body of a literature review. Depending on the length of your literature review, you can combine several of these strategies (for example, your overall structure might be thematic, but each theme is discussed chronologically).

Chronological

The simplest approach is to trace the development of the topic over time. However, if you choose this strategy, be careful to avoid simply listing and summarizing sources in order.

Try to analyze patterns, turning points and key debates that have shaped the direction of the field. Give your interpretation of how and why certain developments occurred.

If you have found some recurring central themes, you can organize your literature review into subsections that address different aspects of the topic.

For example, if you are reviewing literature about inequalities in migrant health outcomes, key themes might include healthcare policy, language barriers, cultural attitudes, legal status, and economic access.

Methodological

If you draw your sources from different disciplines or fields that use a variety of research methods , you might want to compare the results and conclusions that emerge from different approaches. For example:

• Look at what results have emerged in qualitative versus quantitative research
• Discuss how the topic has been approached by empirical versus theoretical scholarship
• Divide the literature into sociological, historical, and cultural sources

Theoretical

A literature review is often the foundation for a theoretical framework . You can use it to discuss various theories, models, and definitions of key concepts.

You might argue for the relevance of a specific theoretical approach, or combine various theoretical concepts to create a framework for your research.

Like any other academic text , your literature review should have an introduction , a main body, and a conclusion . What you include in each depends on the objective of your literature review.

The introduction should clearly establish the focus and purpose of the literature review.

Depending on the length of your literature review, you might want to divide the body into subsections. You can use a subheading for each theme, time period, or methodological approach.

As you write, you can follow these tips:

• Summarize and synthesize: give an overview of the main points of each source and combine them into a coherent whole
• Analyze and interpret: don’t just paraphrase other researchers — add your own interpretations where possible, discussing the significance of findings in relation to the literature as a whole
• Critically evaluate: mention the strengths and weaknesses of your sources
• Write in well-structured paragraphs: use transition words and topic sentences to draw connections, comparisons and contrasts

In the conclusion, you should summarize the key findings you have taken from the literature and emphasize their significance.

When you’ve finished writing and revising your literature review, don’t forget to proofread thoroughly before submitting. Not a language expert? Check out Scribbr’s professional proofreading services !

This article has been adapted into lecture slides that you can use to teach your students about writing a literature review.

Scribbr slides are free to use, customize, and distribute for educational purposes.

Open Google Slides Download PowerPoint

If you want to know more about the research process , methodology , research bias , or statistics , make sure to check out some of our other articles with explanations and examples.

• Sampling methods
• Simple random sampling
• Stratified sampling
• Cluster sampling
• Likert scales
• Reproducibility

 Statistics

• Null hypothesis
• Statistical power
• Probability distribution
• Effect size
• Poisson distribution

Research bias

• Optimism bias
• Cognitive bias
• Implicit bias
• Hawthorne effect
• Anchoring bias
• Explicit bias

A literature review is a survey of scholarly sources (such as books, journal articles, and theses) related to a specific topic or research question .

It is often written as part of a thesis, dissertation , or research paper , in order to situate your work in relation to existing knowledge.

There are several reasons to conduct a literature review at the beginning of a research project:

• To familiarize yourself with the current state of knowledge on your topic
• To ensure that you’re not just repeating what others have already done
• To identify gaps in knowledge and unresolved problems that your research can address
• To develop your theoretical framework and methodology
• To provide an overview of the key findings and debates on the topic

Writing the literature review shows your reader how your work relates to existing research and what new insights it will contribute.

The literature review usually comes near the beginning of your thesis or dissertation . After the introduction , it grounds your research in a scholarly field and leads directly to your theoretical framework or methodology .

A literature review is a survey of credible sources on a topic, often used in dissertations , theses, and research papers . Literature reviews give an overview of knowledge on a subject, helping you identify relevant theories and methods, as well as gaps in existing research. Literature reviews are set up similarly to other  academic texts , with an introduction , a main body, and a conclusion .

An  annotated bibliography is a list of  source references that has a short description (called an annotation ) for each of the sources. It is often assigned as part of the research process for a  paper .  

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.

McCombes, S. (2023, September 11). How to Write a Literature Review | Guide, Examples, & Templates. Scribbr. Retrieved September 3, 2024, from https://www.scribbr.com/dissertation/literature-review/

Is this article helpful?

Shona McCombes

Other students also liked, what is a theoretical framework | guide to organizing, what is a research methodology | steps & tips, how to write a research proposal | examples & templates, "i thought ai proofreading was useless but..".

I've been using Scribbr for years now and I know it's a service that won't disappoint. It does a good job spotting mistakes”

How To Structure Your Literature Review

3 options to help structure your chapter.

By: Amy Rommelspacher (PhD) | Reviewer: Dr Eunice Rautenbach | November 2020 (Updated May 2023)

Writing the literature review chapter can seem pretty daunting when you’re piecing together your dissertation or thesis. As  we’ve discussed before , a good literature review needs to achieve a few very important objectives – it should:

• Demonstrate your knowledge of the research topic
• Identify the gaps in the literature and show how your research links to these
• Provide the foundation for your conceptual framework (if you have one)
• Inform your own  methodology and research design

To achieve this, your literature review needs a well-thought-out structure . Get the structure of your literature review chapter wrong and you’ll struggle to achieve these objectives. Don’t worry though – in this post, we’ll look at how to structure your literature review for maximum impact (and marks!).

But wait – is this the right time?

Deciding on the structure of your literature review should come towards the end of the literature review process – after you have collected and digested the literature, but before you start writing the chapter. 

In other words, you need to first develop a rich understanding of the literature before you even attempt to map out a structure. There’s no use trying to develop a structure before you’ve fully wrapped your head around the existing research.

Equally importantly, you need to have a structure in place before you start writing , or your literature review will most likely end up a rambling, disjointed mess. 

Importantly, don’t feel that once you’ve defined a structure you can’t iterate on it. It’s perfectly natural to adjust as you engage in the writing process. As we’ve discussed before , writing is a way of developing your thinking, so it’s quite common for your thinking to change – and therefore, for your chapter structure to change – as you write. 

Need a helping hand?

Like any other chapter in your thesis or dissertation, your literature review needs to have a clear, logical structure. At a minimum, it should have three essential components – an  introduction , a  body   and a  conclusion . 

Let’s take a closer look at each of these.

1: The Introduction Section

Just like any good introduction, the introduction section of your literature review should introduce the purpose and layout (organisation) of the chapter. In other words, your introduction needs to give the reader a taste of what’s to come, and how you’re going to lay that out. Essentially, you should provide the reader with a high-level roadmap of your chapter to give them a taste of the journey that lies ahead.

Here’s an example of the layout visualised in a literature review introduction:

Your introduction should also outline your topic (including any tricky terminology or jargon) and provide an explanation of the scope of your literature review – in other words, what you  will   and  won’t   be covering (the delimitations ). This helps ringfence your review and achieve a clear focus . The clearer and narrower your focus, the deeper you can dive into the topic (which is typically where the magic lies). 

Depending on the nature of your project, you could also present your stance or point of view at this stage. In other words, after grappling with the literature you’ll have an opinion about what the trends and concerns are in the field as well as what’s lacking. The introduction section can then present these ideas so that it is clear to examiners that you’re aware of how your research connects with existing knowledge .

2: The Body Section

The body of your literature review is the centre of your work. This is where you’ll present, analyse, evaluate and synthesise the existing research. In other words, this is where you’re going to earn (or lose) the most marks. Therefore, it’s important to carefully think about how you will organise your discussion to present it in a clear way. 

The body of your literature review should do just as the description of this chapter suggests. It should “review” the literature – in other words, identify, analyse, and synthesise it. So, when thinking about structuring your literature review, you need to think about which structural approach will provide the best “review” for your specific type of research and objectives (we’ll get to this shortly).

There are (broadly speaking)  three options  for organising your literature review.

Option 1: Chronological (according to date)

Organising the literature chronologically is one of the simplest ways to structure your literature review. You start with what was published first and work your way through the literature until you reach the work published most recently. Pretty straightforward.

The benefit of this option is that it makes it easy to discuss the developments and debates in the field as they emerged over time. Organising your literature chronologically also allows you to highlight how specific articles or pieces of work might have changed the course of the field – in other words, which research has had the most impact . Therefore, this approach is very useful when your research is aimed at understanding how the topic has unfolded over time and is often used by scholars in the field of history. That said, this approach can be utilised by anyone that wants to explore change over time .

For example , if a student of politics is investigating how the understanding of democracy has evolved over time, they could use the chronological approach to provide a narrative that demonstrates how this understanding has changed through the ages.

Here are some questions you can ask yourself to help you structure your literature review chronologically.

• What is the earliest literature published relating to this topic?
• How has the field changed over time? Why?
• What are the most recent discoveries/theories?

In some ways, chronology plays a part whichever way you decide to structure your literature review, because you will always, to a certain extent, be analysing how the literature has developed. However, with the chronological approach, the emphasis is very firmly on how the discussion has evolved over time , as opposed to how all the literature links together (which we’ll discuss next ).

Option 2: Thematic (grouped by theme)

The thematic approach to structuring a literature review means organising your literature by theme or category – for example, by independent variables (i.e. factors that have an impact on a specific outcome).

As you’ve been collecting and synthesising literature , you’ll likely have started seeing some themes or patterns emerging. You can then use these themes or patterns as a structure for your body discussion. The thematic approach is the most common approach and is useful for structuring literature reviews in most fields.

For example, if you were researching which factors contributed towards people trusting an organisation, you might find themes such as consumers’ perceptions of an organisation’s competence, benevolence and integrity. Structuring your literature review thematically would mean structuring your literature review’s body section to discuss each of these themes, one section at a time.

Here are some questions to ask yourself when structuring your literature review by themes:

• Are there any patterns that have come to light in the literature?
• What are the central themes and categories used by the researchers?
• Do I have enough evidence of these themes?

PS – you can see an example of a thematically structured literature review in our literature review sample walkthrough video here.

Option 3: Methodological

The methodological option is a way of structuring your literature review by the research methodologies used . In other words, organising your discussion based on the angle from which each piece of research was approached – for example, qualitative , quantitative or mixed  methodologies.

Structuring your literature review by methodology can be useful if you are drawing research from a variety of disciplines and are critiquing different methodologies. The point of this approach is to question  how  existing research has been conducted, as opposed to  what  the conclusions and/or findings the research were.

For example, a sociologist might centre their research around critiquing specific fieldwork practices. Their literature review will then be a summary of the fieldwork methodologies used by different studies.

Here are some questions you can ask yourself when structuring your literature review according to methodology:

• Which methodologies have been utilised in this field?
• Which methodology is the most popular (and why)?
• What are the strengths and weaknesses of the various methodologies?
• How can the existing methodologies inform my own methodology?

3: The Conclusion Section

Once you’ve completed the body section of your literature review using one of the structural approaches we discussed above, you’ll need to “wrap up” your literature review and pull all the pieces together to set the direction for the rest of your dissertation or thesis.

The conclusion is where you’ll present the key findings of your literature review. In this section, you should emphasise the research that is especially important to your research questions and highlight the gaps that exist in the literature. Based on this, you need to make it clear what you will add to the literature – in other words, justify your own research by showing how it will help fill one or more of the gaps you just identified.

Last but not least, if it’s your intention to develop a conceptual framework for your dissertation or thesis, the conclusion section is a good place to present this.

Example: Thematically Structured Review

In the video below, we unpack a literature review chapter so that you can see an example of a thematically structure review in practice.

Let’s Recap

In this article, we’ve  discussed how to structure your literature review for maximum impact. Here’s a quick recap of what  you need to keep in mind when deciding on your literature review structure:

• Just like other chapters, your literature review needs a clear introduction , body and conclusion .
• The introduction section should provide an overview of what you will discuss in your literature review.
• The body section of your literature review can be organised by chronology , theme or methodology . The right structural approach depends on what you’re trying to achieve with your research.
• The conclusion section should draw together the key findings of your literature review and link them to your research questions.

If you’re ready to get started, be sure to download our free literature review template to fast-track your chapter outline.

Psst… there’s more!

This post is an extract from our bestselling short course, Literature Review Bootcamp . If you want to work smart, you don't want to miss this .

28 Comments

Great work. This is exactly what I was looking for and helps a lot together with your previous post on literature review. One last thing is missing: a link to a great literature chapter of an journal article (maybe with comments of the different sections in this review chapter). Do you know any great literature review chapters?

I agree with you Marin… A great piece

I agree with Marin. This would be quite helpful if you annotate a nicely structured literature from previously published research articles.

Awesome article for my research.

I thank you immensely for this wonderful guide

It is indeed thought and supportive work for the futurist researcher and students

Very educative and good time to get guide. Thank you

Great work, very insightful. Thank you.

Thanks for this wonderful presentation. My question is that do I put all the variables into a single conceptual framework or each hypothesis will have it own conceptual framework?

Thank you very much, very helpful

This is very educative and precise . Thank you very much for dropping this kind of write up .

Pheeww, so damn helpful, thank you for this informative piece.

I’m doing a research project topic ; stool analysis for parasitic worm (enteric) worm, how do I structure it, thanks.

comprehensive explanation. Help us by pasting the URL of some good “literature review” for better understanding.

great piece. thanks for the awesome explanation. it is really worth sharing. I have a little question, if anyone can help me out, which of the options in the body of literature can be best fit if you are writing an architectural thesis that deals with design?

I am doing a research on nanofluids how can l structure it?

Beautifully clear.nThank you!

Lucid! Thankyou!

Brilliant work, well understood, many thanks

I like how this was so clear with simple language 😊😊 thank you so much 😊 for these information 😊

Insightful. I was struggling to come up with a sensible literature review but this has been really helpful. Thank you!

You have given thought-provoking information about the review of the literature.

Thank you. It has made my own research better and to impart your work to students I teach

I learnt a lot from this teaching. It’s a great piece.

I am doing research on EFL teacher motivation for his/her job. How Can I structure it? Is there any detailed template, additional to this?

You are so cool! I do not think I’ve read through something like this before. So nice to find somebody with some genuine thoughts on this issue. Seriously.. thank you for starting this up. This site is one thing that is required on the internet, someone with a little originality!

I’m asked to do conceptual, theoretical and empirical literature, and i just don’t know how to structure it

Asking questions are actually fastidious thing if you are not understanding anything fully, but this article presents good understanding yet.

Submit a Comment Cancel reply

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

• Print Friendly

How to Build a Conceptual Framework From a Systematic Literature Review Study: Insights From the Technology Foresight Domain

• November 2022
• Publisher: SAGE Research Methods Cases
• ISBN: 9781529626537

• University of Leeds

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations
• Milan Marinković

• Samuel Ankrah

• INT J MANAG REV

• TECHNOL FORECAST SOC

• Creativ Innovat Manag

• Pam Pringle
• Gwen Speranzini

• J. E. Mcgrath
• Matthew B. Miles
• A. Michael Huberman
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

• Open access
• Published: 02 September 2024

Clinician perspectives and recommendations regarding design of clinical prediction models for deteriorating patients in acute care

• Robin Blythe   ORCID: orcid.org/0000-0002-3643-4332 1 ,
• Sundresan Naicker   ORCID: orcid.org/0000-0002-2392-4981 1 ,
• Nicole White   ORCID: orcid.org/0000-0002-9292-0773 1 ,
• Raelene Donovan   ORCID: orcid.org/0000-0003-0737-7719 2 ,
• Ian A. Scott   ORCID: orcid.org/0000-0002-7596-0837 3 , 4 ,
• Andrew McKelliget 2 &
• Steven M McPhail   ORCID: orcid.org/0000-0002-1463-662X 1 , 4  

BMC Medical Informatics and Decision Making volume  24 , Article number:  241 ( 2024 ) Cite this article

Metrics details

Successful deployment of clinical prediction models for clinical deterioration relates not only to predictive performance but to integration into the decision making process. Models may demonstrate good discrimination and calibration, but fail to match the needs of practising acute care clinicians who receive, interpret, and act upon model outputs or alerts. We sought to understand how prediction models for clinical deterioration, also known as early warning scores (EWS), influence the decision-making of clinicians who regularly use them and elicit their perspectives on model design to guide future deterioration model development and implementation.

Nurses and doctors who regularly receive or respond to EWS alerts in two digital metropolitan hospitals were interviewed for up to one hour between February 2022 and March 2023 using semi-structured formats. We grouped interview data into sub-themes and then into general themes using reflexive thematic analysis. Themes were then mapped to a model of clinical decision making using deductive framework mapping to develop a set of practical recommendations for future deterioration model development and deployment.

Fifteen nurses ( n  = 8) and doctors ( n  = 7) were interviewed for a mean duration of 42 min. Participants emphasised the importance of using predictive tools for supporting rather than supplanting critical thinking, avoiding over-protocolising care, incorporating important contextual information and focusing on how clinicians generate, test, and select diagnostic hypotheses when managing deteriorating patients. These themes were incorporated into a conceptual model which informed recommendations that clinical deterioration prediction models demonstrate transparency and interactivity, generate outputs tailored to the tasks and responsibilities of end-users, avoid priming clinicians with potential diagnoses before patients were physically assessed, and support the process of deciding upon subsequent management.

Conclusions

Prediction models for deteriorating inpatients may be more impactful if they are designed in accordance with the decision-making processes of acute care clinicians. Models should produce actionable outputs that assist with, rather than supplant, critical thinking.

• This article explored decision-making processes of clinicians using a clinical prediction model for deteriorating patients, also known as an early warning score.

• Our study identified that the clinical utility of deterioration models may lie in their assistance in generating, evaluating, and selecting diagnostic hypotheses, an important part of clinical decision making that is underrepresented in the prediction modelling literature.

• Nurses in particular stressed the need for models that encourage critical thinking and further investigation rather than prescribe strict care protocols.

Peer Review reports

The number of ‘clinical prediction model’ articles published on PubMed has grown rapidly over the past two decades, from 1,918 articles identified with these search terms published in 2002 to 26,326 published in 2022. A clinical prediction model is defined as any multivariable model that provides patient-level estimates of the probability or risk of a disease, condition or future event [ 1 , 2 , 3 ].

Recent systematic and scoping reviews report a lack of evidence that clinical decision support systems based on prediction models are associated with improved patient outcomes once implemented in acute care [ 4 , 5 , 6 , 7 ]. One potential reason may be that some models are not superior to clinical judgment in reducing missed diagnoses or correctly classifying non-diseased patients [ 8 ]. While improving predictive accuracy is important, this appears insufficient for improving patient outcomes, suggesting that more attention should be paid to the process and justification of how prediction models are designed and deployed [ 9 , 10 ].

If model predictions are to influence clinical decision-making, they must not only demonstrate acceptable accuracy, but also be implemented and adopted at scale in clinical settings. This requires consideration of how they are integrated into clinical workflows, how they generate value for users, and how clinicians perceive and respond to their outputs of predicted risks [ 11 , 12 ]. These concepts are tenets of user-centred design, which focuses on building systems based on the needs and responsibilities of those who will use them. User-centred decision support tools can be designed in a variety of ways, but may benefit from understanding the characteristics of the users and the local environment in which tools are implemented, [ 13 ] the nature of the tasks end-users are expected to perform, [ 14 ] and the interface between the user and the tools [ 15 ].

Prediction models for clinical deterioration

A common task for prediction models integrated into clinical decision support systems is in predicting or recognising clinical deterioration, also known as early warning scores. Clinical deterioration is defined as the transition of a patient from their current health state to a worse one that puts them at greater risk of adverse events and death [ 16 ]. Early warning scores were initially designed to get the attention of skilled clinicians when patients began to deteriorate, but have since morphed into complex multivariable prediction models [ 17 ]. As with many other clinical prediction models, early warning scores often fail to demonstrate better patient outcomes once deployed [ 4 , 18 ]. The clinical utility of early warning scores likely rests on two key contextual elements: the presence of uncertainty, both in terms of diagnosis and prognosis, and the potential for undesirable patient outcomes if an appropriate care pathway is delayed or an inappropriate one is chosen [ 19 ].

The overarching goal of this qualitative study was to determine how prediction models for clinical deterioration, or early warning scores, could be better tailored to the needs of end-users to improve inpatient care. This study had three aims. First, to understand the experiences and perspectives of nurses and doctors who use early warning scores. Second, to identify the tasks these clinicians performed when managing deteriorating patients, the decision-making processes that guided these tasks, and how these could be conceptualised schematically. Finally, to address these tasks and needs with actionable, practical recommendations for enhancing future deterioration prediction model development and deployment.

To achieve our study aims, we conducted semi-structured interviews of nurses and doctors at two large, digitally mature hospitals. We first asked clinicians to describe their backgrounds, perspectives, and experience with early warning scores to give context to our analysis. We then examined the tasks and responsibilities of participants and the decision-making processes that guided these tasks using reflexive thematic analysis, an inductive method that facilitated the identification of general themes. We then identified a conceptual decision-making framework from the literature to which we mapped these themes to understand how they may lead to better decision support tools. Finally, we used this framework to formulate recommendations for deterioration prediction model design and deployment. These steps are presented graphically in a flow diagram (Fig.  1 ).

Schema of study goal, aims and methods

The study was conducted at one large tertiary and one medium-sized metropolitan hospital in Brisbane, Australia. The large hospital contained over 1,000 beds, handling over 116,000 admissions and approximately 150,000 deterioration alerts per year in 2019. Over the same period, the medium hospital contained 175 beds, handling over 31,000 admissions and approximately 42,000 deterioration alerts per year. These facilities had a high level of digital maturity, including fully integrated electronic medical records.

Clinical prediction model for deteriorating patients

The deterioration monitoring system used at both hospitals was the Queensland Adult Deterioration Detection System (Q-ADDS) [ 20 , 21 ]. Q-ADDS uses an underlying prediction model to convert patient-level vital signs from a single time of observation into an ordinal risk score describing an adult patient’s risk of acute deterioration. Vital signs collected are respiratory rate (breaths/minute), oxygen flow rate (L/minute), arterial oxygen saturation (percent), blood pressure (mmHg), heart rate (beats/minute), temperature (degrees Celsius), level of consciousness (Alert-Voice-Pain-Unresponsive) and increased or new onset agitation. Increased pain and urine output are collected but not used for score calculation [ 21 ]. The Q-ADDS tool is included in the supplementary material.

Vital signs are entered into the patient’s electronic medical record, either imported from the vital signs monitoring device at the patient’s bedside or from manual entry by nurses. Calculations are made automatically within Q-ADDS to generate an ordinal risk score per patient observation. Scores can be elevated to levels requiring a tiered escalation response if a single vital sign is greatly deranged, or if several observations are deranged by varying degrees. Scores range from 0 to 8+, with automated alerts and escalation protocols ranging from more frequent observations for lower scores to immediate activation of the medical emergency team (MET) at higher scores.

The escalation process for Q-ADDS is highly structured, mandated and well documented [ 21 ]. Briefly, when a patient’s vital signs meet a required alert threshold, the patient’s nurse is required to physically assess the patient and, depending on the level of severity predicted by Q-ADDS, notify the patient’s doctor (escalation). The doctor is then required to be notified of the patient’s Q-ADDS score, potentially review the patient, and discuss any potential changes to care with the nurse. Both nurses and doctors can escalate straight to MET calls or an emergency ‘code blue’ call (requiring cardiopulmonary resuscitation or assisted ventilation) at any time if necessary.

Participant recruitment

Participant recruitment began in February 2022 and concluded in March 2023, disrupted by the COVID-19 pandemic. Eligibility criteria were nurses or doctors at each hospital with direct patient contact who either receive or respond, respectively, to Q-ADDS alerts. An anticipated target sample size of 15 participants was established prior to recruitment, based on expected constraints in recruitment due to clinician workloads and the expected length of interviews relative to their scope, as guided by prior research [ 22 ]. As the analysis plan involved coding interviews iteratively as they were conducted, the main justification for ceasing recruitment was when no new themes relating to the study objectives were generated during successive interviews as the target sample size was approached [ 23 ].

Study information was broadly distributed via email to nurses and doctors in patient-facing roles across hospitals. Nurse unit managers were followed up during regular nursing committee meetings to participate or assist with recruitment within their assigned wards. Doctors were followed up by face-to-face rounding. Snowball sampling, in which participants were encouraged to refer their colleagues for study participation, was employed whenever possible. In all cases, study authors explained study goals and distributed participant consent forms prior to interview scheduling with the explicit proviso that participation was completely voluntary and anonymous to all but two study authors (RB and SN).

Interview process

We used a reflexive framework method to develop an open-ended interview template [ 24 ] that aligned with our study aims. Interview questions were informed by the non-adoption, abandonment, scale-up, spread and sustainability (NASSS) framework [ 25 ]. The NASSS framework relates the end-user perceptions of the technology being evaluated to its value proposition for the clinical situation to which it is being applied. We selected a reflexive method based on the NASSS for our study as we wanted to allow end-users to speak freely about the barriers they faced when using prediction models for clinical deterioration, but did not limit participants to discussing only topics that could fit within the NASSS framework.

Participants were first asked about their background and clinical expertise. They were then invited to share their experiences and perspectives with using early warning scores to manage deteriorating patients. This was used as a segue for participants to describe the primary tasks required of them when evaluating and treating a deteriorating patient. Participants were encouraged to talk through their decision-making process when fulfilling these tasks, and to identify any barriers or obstacles to achieving those tasks that were related to prediction models for deteriorating patients. Participants were specifically encouraged to identify any sources of information that were useful for managing deteriorating patients, including prediction models for other, related disease groups like sepsis, and to think of any barriers or facilitators for making that information more accessible. Finally, participants were invited to suggest ways to improve early warning scores, and how those changes may lead to benefits for patients and clinicians.

As we employed a reflexive methodology to allow clinicians to speak freely about their perspectives and opinions, answers to interview questions were optional and open-ended, allowing participants to discuss relevant tangents. Separate interview guides were developed for nurses and doctors as the responsibilities and information needs of these two disciplines in managing deteriorating patients often differ. Nurses are generally charged with receiving and passing on deterioration alerts, while doctors are generally charged with responding to alerts and making any required changes to patient care plans [ 4 ]. Interview guides are contained in the supplement.

Due to clinician workloads, member checking, a form of post-interview validation in which participants retrospectively confirm their interview answers, was not used. To ensure participants perceived the interviewers as being impartial, two study authors not employed by the hospital network and not involved in direct patient care (RB and SN) were solely responsible for conducting interviews and interrogating interview transcripts. Interviews were recorded and transcribed verbatim, then re-checked for accuracy.

Inductive thematic analysis

Transcripts were analysed using a reflexive thematic methodology informed by Braun and Clarke [ 26 ]. This method was selected because it facilitated exploring the research objectives rather than being restricted to the domains of a specific technology adoption framework, which may limit generalisability [ 27 ]. Interviews were analysed over five steps to identify emergent themes.

Each interview was broken down into segments by RB and SN, where segments corresponded to a distinct opinion.

Whenever appropriate, representative quotes for each distinct concept were extracted.

Segments were grouped into sub-themes.

Sub-themes were grouped into higher-order themes, or general concepts.

Steps 1 through 4 were iteratively repeated by RB and supervised by SN.

As reflexive methods incorporate the experiences and expertise of the analysts, our goal was to extract any sub-themes relevant to the study aims and able to be analysed in the context of early warning scores, prediction models, or decision support tools for clinical deterioration. The concepts explored during this process were not exhaustive, but repeated analysis and re-analysis of participant transcripts helped to ensure all themes could be interpreted in the context of our three study aims: background and perspectives, tasks and decision-making, and recommendations for future practice.

Deductive mapping to a clinical decision-making framework

Once the emergent themes from the inductive analysis were defined, we conducted a brief scan of PubMed for English-language studies that investigated how the design of clinical decision support systems relate to clinical decision-making frameworks. The purpose of this exercise was to identify a framework against which we could map the previously elicited contexts, tasks, and decision-making of end-users in developing a decision-making model that could then be used to support the third aim of formulating recommendations to enhance prediction model development and deployment.

RB and SN then mapped higher-order themes from the inductive analysis to the decision-making model based on whether there was a clear relationship between each theme and a node in the model (see Results).

Recommendations for improving prediction model design were derived by reformatting the inductive themes based on the stated preferences of the participants. These recommendations were then assessed by the remaining authors and the process repeated iteratively until authors were confident that all recommendations were concordant with the decision-making model.

Participant characteristics

Our sample included 8 nurses and 7 doctors of varying levels of expertise and clinical specialties; further information is contained in the supplement. Compared to doctors, nurse participants were generally more experienced, often participating in training or mentoring less experienced staff. Clinical specialities of nurses were diverse, including orthopaedics, cancer services, medical assessment and planning unit, general medicine, and pain management services. Doctor participants ranged from interns with less than a year of clinical experience up to consultant level, including three doctors doing training rotations and two surgical registrars. Clinical specialties of doctors included geriatric medicine, colorectal surgery, and medical education.

Interviews and thematic analysis

Eleven interviews were conducted jointly by RB and SN, one conducted by RB, and three by SN. Interviews were scheduled for up to one hour, with a mean duration of 42 min. Six higher-order themes were identified. These were: added value of more information; communication of model outputs; validation of clinical intuition; capability for objective measurement; over-protocolisation of care; and model transparency and interactivity (Table  1 ). Some aspects of care, including the need for critical thinking and the informational value of discerning trends in patient observations, were discussed in several contexts, making them relevant to more than one higher-order theme.

Added value of other information

Clinicians identified that additional data or variables important for decision making were often omitted from the Q-ADDS digital interface. Such variables included current medical conditions, prescribed medications and prior observations, which were important for interpreting current patient data in the context of their baseline observations under normal circumstances (e.g., habitually low arterial oxygen saturation due to chronic obstructive pulmonary disease) or in response to an acute stimulus (e.g., expected hypotension for next 4 to 8 h while treatment for septic shock is underway).

“The trend is the biggest thing [when] looking at the data , because sometimes people’s observations are deranged forever and it’s not abnormal for them to be tachycardic , whereas for someone else , if it’s new and acute , then that’s a worry.” – Registrar.

Participants frequently emphasised the critical importance of looking at patients holistically, or that patients were more than the sum of the variables used to predict risk. Senior nurses stressed that prediction models were only one part of patient evaluation, and clinicians should be encouraged to incorporate both model outputs and their own knowledge and experiences in decision making rather than trust models implicitly. Doctors also emphasised this holistic approach, adding that they placed more importance on hearing a nurse was concerned for the patient than seeing the model output. Critical thinking about future management was frequently raised in this context, with both nurses and doctors insisting that model predictions and the information required for contextualising risk scores should be communicated together when escalating the patient’s care to more senior clinicians.

Model outputs

Model outputs were discussed in two contexts. First, doctors perceived that ordinal risk scores generated by Q-ADDS felt arbitrary compared to receiving probabilities of a future event, for example cardiorespiratory decompensation, that required a response such as resuscitation or high-level treatment. However, nurses did not wholly embrace probabilities as outputs, instead suggesting that recommendations for how they should respond to different Q-ADDS scores were more important. This difference may reflect the different roles of alert receivers (nurses) and alert responders (doctors).

“[It’s helpful] if you use probabilities… If your patient has a sedation score of 2 and a respiratory rate of 10 , [giving them] a probability of respiratory depression would be helpful. However , I don’t find many clinicians , and certainly beginning practitioners , think in terms of probabilities.” – Clinical nurse consultant.

Second, there was frequent mention of alert fatigue in the context of model outputs. One doctor and two nurses felt there was insufficient leeway for nurses to exercise discretion in responding to risk scores, leading to many unnecessary alert-initiated actions. More nuance in the way Q-ADDS outputs were delivered to clinicians with different roles was deemed important to avoid model alerts being perceived as repetitive and unwarranted. However, three other doctors warned against altering MET call criteria in response to repetitive and seemingly unchanging risk scores and that at-risk patients should, as a standard of care, remain under frequent observation. Frustrations centred more often around rigidly tying repetitive Q-ADDS outputs to certain mandated actions, leading to multiple clinical reviews in a row for a patient whose trajectory was predictable, for example a patient with stable heart failure having a constantly low blood pressure. This led to duplication of nursing effort (e.g., repeatedly checking the blood pressure) and the perception that prediction models were overly sensitive.

“It takes away a lot of nurses’ critical judgement. If someone’s baseline systolic [blood pressure] is 95 [mmHg] , they’re asymptomatic and I would never hear about it previously. We’re all aware that this is where they sit and that’s fine. Now they are required to notify me in the middle of the night , “Just so you know , they’ve dropped to 89 [below an alert threshold of 90mmHg].“” – Junior doctor.

Validation of clinical intuition

Clinicians identified the ability of prediction models to validate their clinical intuition as both a benefit and a hindrance, depending on how outputs were interpreted and acted upon. Junior clinicians appreciated early warning scores giving them more support to escalate care to senior clinicians, as a conversation starter or framing a request for discussion. Clinicians described how assessing the patient holistically first, then obtaining model outputs to add context and validate their diagnostic hypotheses, was very useful in deciding what care should be initiated and when.

“You kind of rule [hypotheses] out… you go to the worst extreme: is it something you need to really be concerned about , especially if their [score] is quite high? You’re thinking of common complications like blood clots , so that presents as tachycardic… I’m thinking of a PE [pulmonary embolism] , then you do the nursing interventions.” – Clinical nurse manager.

While deterioration alerts were often seen as triggers to think about potential causes for deterioration, participants noted that decision making could be compromised if clinicians were primed by model outputs to think of different diagnoses before they had fully assessed the patient at the bedside. Clinicians described the dangers of tunnel vision or, before considering all available clinical information, investigating favoured diagnoses to the exclusion of more likely causes.

“[Diagnosis-specific warnings are] great , [but] that’s one of those things that can lead to a bit of confirmation bias… It’s a good trigger to articulate , “I need to look for sources of infection when I go to escalate"… but then , people can get a little bit sidetracked with that and ignore something more blatant in front of them. I’ve seen people go down this rabbit warren of being obsessed with the “fact” that it was sepsis , but it was something very , very unrelated.” – Nurse educator.

Objective measurement

Clinicians perceived that prediction models were useful as more objective measures of patients’ clinical status that could ameliorate clinical uncertainty or mitigate cognitive biases. In contrast to the risk of confirmation bias arising from front-loading model outputs suggesting specific diagnoses, prediction models could offer a second opinion that could help clinicians recognise opposing signals in noisy data that, in particular, assisted in considering serious diagnoses that shouldn’t be missed (e.g., sepsis), or more frequent and easily treated diagnoses (e.g., dehydration). Prediction models were also useful when they disclosed several small, early changes in patient status that provided an opportunity for early intervention.

“Maybe [the patient has] a low grade fever , they’re a bit tachycardic. Maybe [sepsis] isn’t completely out of the blue for this person. If there was some sort of tool , that said there’s a reasonable chance that they could have sepsis here , I would use that to justify the option of going for blood cultures and maybe a full septic screen. If [I’m indecisive] , that sort of information could certainly push me in that direction.” – Junior doctor.

Clinicians frequently mentioned that prediction models would have been more useful when first starting clinical practice, but become less useful with experience. However, clinicians noted that at any experience level, risk scoring was considered most useful as a triage/prioritisation tool, helping decide which patients to see first, or which clinical concerns to address first.

“[Doctors] can easily triage a patient who’s scoring 4 to 5 versus 1 to 3. If they’re swamped , they can change the escalation process , or triage appropriately with better communication.” – Clinical nurse manager.

Clinicians also stressed that predictions were not necessarily accurate because measurement error or random variation, especially one-off outlier values for certain variables, was a significant contributor to false alerts and inappropriate responses. For example, a single unusually high respiratory rate generated an unusually high risk score, prompting an unnecessary alert.

Over-protocolisation of care

The sentiment most commonly expressed by all experienced nursing participants and some doctors was that nurses were increasingly being trained to solely react to model outputs with fixed response protocols, rather than think critically about what is happening to patients and why. It was perceived that prediction models may actually reduce the capacity for clinicians to process and internalise important information. For example, several nurses observed their staff failing to act on their own clinical suspicions that patients were deteriorating because the risk score had not exceeded a response threshold.

“We’ve had patients on the ward that have had quite a high tachycardia , but it’s not triggering because it’s below the threshold to trigger… [I often need to make my staff] make the clinical decision that they can call the MET anyway , because they have clinical concern with the patient.” – Clinical nurse consultant.

A source of great frustration for many nurses was the lack of critical thinking by their colleagues of possible causes when assessing deteriorating patients. They wanted their staff to investigate whether early warning score outputs or other changes in patient status were caused by simple, easily fixable issues such as fitting the oxygen mask properly and helping the patient sit up to breathe more easily, or whether they indicated more serious underlying pathophysiology. Nurses repeatedly referenced the need for clinicians to always be asking why something was happening, not simply reacting to what was happening.

“[Models should also be] trying to get back to critical thinking. What I’m seeing doesn’t add up with the monitor , so I should investigate further than just simply calling the code.” – Clinical nurse educator.

Model transparency and interactivity

Clinicians frequently requested more transparent and interactive prediction models. These included a desire to receive more training in how prediction models worked and how risk estimates were generated mathematically, and being able to visualise important predictors of deterioration and the absolute magnitude of their effects (effect sizes) in intuitive ways. For example, despite receiving training in Q-ADDS, nurses expressed frustrations that nobody at the hospital seemed to understand how it worked in generating risk scores. Doctors were interested in being able to visualise the relative size and direction of effect of different model variables, potentially using colour-coding, combined with other contextual patient data like current vital sign trends and medications, and presented on one single screen.

The ability to modify threshold values for model variables and see how this impacted risk scores, and what this may then mean for altering MET calling criteria, was also discussed. For example, in an older patient with an acute ischaemic stroke, a persistently high, asymptomatic blood pressure value is an expected bodily response to this acute insult over the first 24–48 h. In the absence of any change to alert criteria, recurrent alerts would be triggered which may encourage overtreatment and precipitous lowering of the blood pressure with potential to cause harm. Altering the criteria to an acceptable or “normal” value for this clinical scenario (i.e. a higher than normal blood pressure) may generate a lower, more patient-centred risk estimate and less propensity to overtreat. This ability to tinker with the model may also enhance understanding of how it works.

“I wish I could alter criteria and see what the score is after that , with another set of observations. A lot of the time… I wonder what they’re sitting at , now that I’ve [altered] the bit that I’m not concerned about… It would be quite helpful to refresh it and have their score refreshed as the new score.” – Junior doctor.

Derivation of the decision-making model

Guided by the responses of our participants regarding their decision-making processes, our literature search identified a narrative review by Banning (2008) that reported previous work by O’Neill et al. (2005) [ 28 , 29 ]. While these studies referred to models of nurse decision-making, we selected a model (Fig.  2 ) that also appropriately described the responses of doctors in our participant group and matched the context of using clinical decision support systems to support clinical judgement. As an example, when clinicians referenced needing to look for certain data points to give context to a patient assessment, this was mapped to nodes relating to “Current patient data,” “Changes to patient status/data,” and “Hypothesis-driven assessment.”

Decision-making model(Adapted from Neill’s clinical decision making framework [2005] and modified by Banning [2006]) with sequential decision nodes

Mapping of themes to decision-making model

The themes from Table  1 were mapped to the nodes in the decision-making model based on close alignment with participant responses (see Fig.  3 ). This mapping is further explained below, where the nodes in the model are described in parentheses.

Value of additional information for decision-making : participants stressed the importance of understanding not only the data going into the prediction model, but also how that data changed over time as trends, and the data that were not included in the model. (Current patient data, changes to patient status/data)

Format, frequency, and relevance of outputs : participants suggested a change in patient data should not always lead to an alert. Doctors, but not necessarily nurses, proposed outputs displayed as probabilities rather than scores, tying model predictions to potential diagnoses or prognoses. (Changes to patient status/data, hypothesis generation)

Using models to validate but not supersede clinical intuition : Depending on the exact timing of model outputs within the pathway of patient assessment, participants found predictions could either augment or hinder the hypothesis generation process. (Hypothesis generation)

Measuring risks objectively : Risk scores can assist with triaging or prioritising patients by urgency or prognostic risk, thereby potentially leading to early intervention to identify and/or prevent adverse events. (Clinician concerns, hypothesis generation)

Supporting critical thinking and reducing over-protocolised care : by acting as triggers for further assessment, participants suggested prediction models can support or discount diagnostic hypotheses, lead to root-cause identification, and facilitate interim cares, for example by ensuring good fit of nasal prongs. (Provision of interim care, hypothesis generation, hypothesis-driven assessment)

Model transparency and interactivity : understanding how prediction models worked, being able to modify or add necessary context to model predictions, and understanding the relative contribution of different predictors could better assist the generation and selection of different hypotheses that may explain a given risk score. (Hypothesis generation, recognition of clinical pattern and hypothesis selection)

Mapping of the perceived relationships between higher-order themes and nodes in the decision-making model shown in Fig.  2

Recommendations for improving the design of prediction models

Based on the mapping of themes to the decision-making model, we formulated four recommendations for enhancing the development and deployment of prediction models for clinical deterioration.

Improve accessibility and transparency of data included in the model. Provide an interface that allows end-users to see what predictor variables are included in the model, their relative contributions to model outputs, and facilitate easy access to data not included in the model but still relevant for model-informed decisions, e.g., trends of predictor variables over time.

Present model outputs that are relevant to the end-user receiving those outputs, their responsibilities, and the tasks they may be obliged to perform, while preserving the ability of clinicians to apply their own discretionary judgement.

In situations associated with diagnostic uncertainty, avoid tunnel vision from priming clinicians with possible diagnostic explanations based on model outputs, prior to more detailed clinical assessment of the patient.

Support critical thinking whereby clinicians can apply a more holistic view of the patient’s condition, take all relevant contextual factors into account, and be more thoughtful in generating and selecting causal hypotheses.

This qualitative study involving front-line acute care clinicians who respond to early warning score alerts has generated several insights into how clinicians perceive the use of prediction models for clinical deterioration. Clinicians preferred models that facilitated critical thinking, allowed an understanding of the impact of variables included and excluded from the model, provided model outputs specific to the tasks and responsibilities of different disciplines of clinicians, and supported decision-making processes in terms of hypotheses and choice of management, rather than simply responding to alerts in a pre-specified, mandated manner. In particular, preventing prediction models from supplanting critical thinking was repeatedly emphasised.

Reduced staffing ratios, less time spent with patients, greater reliance on more junior workforce, and increasing dependence on automated activation of protocolised management are all pressures that could lead to a decline in clinical reasoning skills. This problem could be exacerbated by adding yet more predictive algorithms and accompanying protocols for other clinical scenarios, which may intensify alert fatigue and disrupt essential clinical care. However, extrapolating our results to areas other than clinical deterioration should be done with caution. An opposing view may be that using prediction models to reduce the burden of routine surveillance may allow redirection of critical thinking skills towards more useful tasks, a question that has not been explored in depth in the clinical informatics literature.

Clinicians expressed interest in models capable of providing causal insights into clinical deterioration. This is neither a function nor capability of most risk prediction models, requiring different assumptions and theoretical frameworks [ 30 ]. Despite this limitation, risk nomograms, visualisations of changes in risk with changes in predictor variables, and other interactive tools for estimating risk may be useful adjuncts for clinical decision-making due to the ease with which input values can be manipulated.

Contributions to the literature

Our research supports and extends the literature on the acceptability of risk prediction models within clinical decision support systems. Common themes in the literature supporting good practices in clinical informatics and which are also reflected in our study include: alert fatigue; the delivery of more relevant contextual information; [ 31 ] the value of patient histories; [ 32 , 33 ] ranking relevant information by clinical importance, including colour-coding; [ 34 , 35 ] not using computerised tools to replace clinical judgement; [ 32 , 36 , 37 ] and understanding the analytic methods underpinning the tool [ 38 ]. One other study has investigated the perspectives of clinicians of relatively simple, rules-based prediction models similar to Q-ADDS. Kappen et al [ 12 ] conducted an impact study of a prediction model for postoperative nausea and vomiting and also found that clinicians frequently made decisions in an intuitive manner that incorporated information both included and absent from prediction models. However, the authors recommended a more directive than assistive approach to model-based recommendations, possibly due to a greater focus on timely prescribing of effective prophylaxis or treatment.

The unique contribution of our study is a better understanding of how clinicians may use prediction models to generate and validate diagnostic hypotheses. The central role of critical thinking and back-and-forth interactions between clinician and model in our results provide a basis for future research using more direct investigative approaches like cognitive task analysis [ 39 ]. Our study has yielded a set of cognitive insights into decision making that can be applied in tandem with statistical best practice in designing, validating and implementing prediction models. [ 19 , 40 , 41 ].

Relevance to machine learning and artificial intelligence prediction models for deterioration

Our results may generalise to prediction models based on machine learning (ML) and artificial intelligence (AI), according to results of several recent studies. Tonekaboni et al [ 42 ] investigated clinician preferences for ML models in the intensive care unit and emergency department using hypothetical scenarios. Several themes appear both in our results and theirs: a need to understand the impact of both included and excluded predictors on model performance; the role of uncertain or noisy data in prediction accuracy; and the influence of trends or patient trajectories in decision making. Their recommendations for more transparent models and the delivery of model outputs designed for the task at hand align closely with ours. The authors’ focus on clinicians’ trust in the model was not echoed by our participants.

Eini-Porat et al [ 43 ] conducted a comprehensive case study of ML models in both adult and paediatric critical care. Their results present several findings supported by our participants despite differences in clinical environments: the value of trends and smaller changes in several vital signs that could cumulatively signal future deterioration; the utility of triage and prioritisation in time-poor settings; and the use of models as triggers for investigating the cause of deterioration.

As ML/AI models proliferate in the clinical deterioration prediction space, [ 44 ] it is important to deeply understand the factors that may influence clinician acceptance of more complex approaches. As a general principle, these methods often strive to input as many variables or transformations of those variables as possible into the model development process to improve predictive accuracy, incorporating dynamic updating to refine model performance. While this functionality may be powerful, highly complex models are not easily explainable, require careful consideration of generalisability, and can prevent clinicians from knowing when a model is producing inaccurate predictions, with potential for patient harm when critical healthcare decisions are being made [ 45 , 46 , 47 ]. Given that our clinicians emphasised the need to understand the model, know which variables are included and excluded, and correctly interpret the format of the output, ML/AI models in the future will need to be transparent in their development and their outputs easily interpretable.

Limitations

The primary limitations of our study were that our sample was drawn from two hospitals with high levels of digital maturity in a metropolitan region of a developed country, with a context specific to clinical deterioration. Our sample of 15 participants may be considered small but is similar to that of other studies with a narrow focus on clinical perspectives [ 42 , 43 ]. All these factors can limit generalisability to other settings or to other prediction models. As described in the methods, we used open-ended interview templates and generated our inductive themes reflexively, which is vulnerable to different types of biases compared to more structured preference elicitation methods with rigidly defined analysis plans. Member checking may have mitigated this bias, but was not possible due to the time required from busy clinical staff.

Our study does not directly deal with methodological issues in prediction model development, [ 41 , 48 ] nor does it provide explicit guidance on how model predictions should be used in clinical practice. Our findings should also not be considered an exhaustive list of concerns clinicians have with prediction models for clinical deterioration, nor may they necessarily apply to highly specialised clinical areas, such as critical care. Our choice of decision making framework was selected because it demonstrated a clear, intuitive causal pathway for model developers to support the clinical decision-making process. However, other, equally valid frameworks may have led to different conclusions, and we encourage more research in this area.

This study elicited clinician perspectives of models designed to predict and manage impending clinical deterioration. Applying these perspectives to a decision-making model, we formulated four recommendations for the design of future prediction models for deteriorating patients: improved transparency and interactivity, tailoring models to the tasks and responsibilities of different end-users, avoiding priming clinicians with diagnostic predictions prior to in-depth clinical review, and finally, facilitating the diagnostic hypothesis generation and assessment process.

Availability of data and materials

Due to privacy concerns and the potential identifiability of participants, interview transcripts are not available. However, interview guides are available in the supplement.

Jenkins DA, Martin GP, Sperrin M, Riley RD, Debray TPA, Collins GS, Peek N. Continual updating and monitoring of clinical prediction models: time for dynamic prediction systems? Diagn Prognostic Res. 2021;5(1):1.

Article   Google Scholar  

Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015;350:g7594.

Article   PubMed   Google Scholar  

Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, et al. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162(1):W1–73.

Blythe R, Parsons R, White NM, Cook D, McPhail SM. A scoping review of real-time automated clinical deterioration alerts and evidence of impacts on hospitalised patient outcomes. BMJ Qual Saf. 2022;31(10):725–34.

Fahey M, Crayton E, Wolfe C, Douiri A. Clinical prediction models for mortality and functional outcome following ischemic stroke: a systematic review and meta-analysis. PLoS ONE. 2018;13(1):e0185402.

Article   PubMed   PubMed Central   Google Scholar  

Fleuren LM, Klausch TLT, Zwager CL, Schoonmade LJ, Guo T, Roggeveen LF, et al. Machine learning for the prediction of sepsis: a systematic review and meta-analysis of diagnostic test accuracy. Intensive Care Med. 2020;46(3):383–400.

White NM, Carter HE, Kularatna S, Borg DN, Brain DC, Tariq A, et al. Evaluating the costs and consequences of computerized clinical decision support systems in hospitals: a scoping review and recommendations for future practice. J Am Med Inform Assoc. 2023;30(6):1205–18.

Sanders S, Doust J, Glasziou P. A systematic review of studies comparing diagnostic clinical prediction rules with clinical judgment. PLoS ONE. 2015;10(6):e0128233.

Abell B, Naicker S, Rodwell D, Donovan T, Tariq A, Baysari M, et al. Identifying barriers and facilitators to successful implementation of computerized clinical decision support systems in hospitals: a NASSS framework-informed scoping review. Implement Sci. 2023;18(1):32.

van der Vegt AH, Campbell V, Mitchell I, Malycha J, Simpson J, Flenady T, et al. Systematic review and longitudinal analysis of implementing Artificial Intelligence to predict clinical deterioration in adult hospitals: what is known and what remains uncertain. J Am Med Inf Assoc. 2024;31(2):509–24.

Davenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future Healthc J. 2019;6(2):94–8.

Kappen TH, van Loon K, Kappen MA, van Wolfswinkel L, Vergouwe Y, van Klei WA, et al. Barriers and facilitators perceived by physicians when using prediction models in practice. J Clin Epidemiol. 2016;70:136–45.

Witteman HO, Dansokho SC, Colquhoun H, Coulter A, Dugas M, Fagerlin A, Giguere AM, Glouberman S, Haslett L, Hoffman A, Ivers N. User-centered design and the development of patient decision aids: protocol for a systematic review. Systematic reviews. 2015;4:1−8.

Zhang J, Norman DA. Representations in distributed cognitive tasks. Cogn Sci. 1994;18(1):87–122.

Johnson CM, Johnson TR, Zhang J. A user-centered framework for redesigning health care interfaces. J Biomed Inf. 2005;38(1):75–87.

Jones D, Mitchell I, Hillman K, Story D. Defining clinical deterioration. Resuscitation. 2013;84(8):1029–34.

Morgan RJ, Wright MM. In defence of early warning scores. Br J Anaesth. 2007;99(5):747–8.

Article   CAS   PubMed   Google Scholar  

Smith ME, Chiovaro JC, O’Neil M, Kansagara D, Quinones AR, Freeman M, et al. Early warning system scores for clinical deterioration in hospitalized patients: a systematic review. Annals Am Thorac Soc. 2014;11(9):1454–65.

Baker T, Gerdin M. The clinical usefulness of prognostic prediction models in critical illness. Eur J Intern Med. 2017;45:37–40.

Campbell V, Conway R, Carey K, Tran K, Visser A, Gifford S, et al. Predicting clinical deterioration with Q-ADDS compared to NEWS, between the flags, and eCART track and trigger tools. Resuscitation. 2020;153:28–34.

The Australian Commission on Safety and Quality in Health is the publisher, and the publisher location is Sydney, Australia. https://www.safetyandquality.gov.au/sites/default/files/migrated/35981-ChartDevelopment.pdf .

Vasileiou K, Barnett J, Thorpe S, Young T. Characterising and justifying sample size sufficiency in interview-based studies: systematic analysis of qualitative health research over a 15-year period. BMC Med Res Methodol. 2018;18(1):148.

Hennink MM, Kaiser BN, Marconi VC. Code saturation versus meaning saturation: how many interviews are Enough? Qual Health Res. 2017;27(4):591–608.

Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013;13(1):1–8.

Greenhalgh T, Wherton J, Papoutsi C, Lynch J, Hughes G, A’Court C, et al. Beyond adoption: a New Framework for Theorizing and evaluating nonadoption, abandonment, and challenges to the Scale-Up, Spread, and sustainability of Health and Care technologies. J Med Internet Res. 2017;19(11):e367.

Braun V, Clarke V. One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Res Psychol. 2021;18(3):328–52.

Campbell KA, Orr E, Durepos P, Nguyen L, Li L, Whitmore C, et al. Reflexive thematic analysis for applied qualitative health research. Qualitative Rep. 2021;26(6):2011–28.

Google Scholar  

Banning M. A review of clinical decision making: models and current research. J Clin Nurs. 2008;17(2):187–95.

O’Neill ES, Dluhy NM, Chin E. Modelling novice clinical reasoning for a computerized decision support system. J Adv Nurs. 2005;49(1):68–77.

Arnold KF, Davies V, de Kamps M, Tennant PWG, Mbotwa J, Gilthorpe MS. Reflection on modern methods: generalized linear models for prognosis and intervention—theory, practice and implications for machine learning. Int J Epidemiol. 2020;49(6):2074–82.

Article   PubMed Central   Google Scholar  

Westerbeek L, Ploegmakers KJ, de Bruijn GJ, Linn AJ, van Weert JCM, Daams JG, et al. Barriers and facilitators influencing medication-related CDSS acceptance according to clinicians: a systematic review. Int J Med Informatics. 2021;152:104506.

Henshall C, Marzano L, Smith K, Attenburrow MJ, Puntis S, Zlodre J, et al. A web-based clinical decision tool to support treatment decision-making in psychiatry: a pilot focus group study with clinicians, patients and carers. BMC Psychiatry. 2017;17(1):265.

Weingart SN, Simchowitz B, Shiman L, Brouillard D, Cyrulik A, Davis RB, et al. Clinicians’ assessments of electronic medication safety alerts in ambulatory care. Arch Intern Med. 2009;169(17):1627–32.

Baysari MT, Zheng WY, Van Dort B, Reid-Anderson H, Gronski M, Kenny E. A late attempt to involve end users in the design of medication-related alerts: Survey Study. J Med Internet Res. 2020;22(3):e14855.

Trafton J, Martins S, Michel M, Lewis E, Wang D, Combs A, et al. Evaluation of the acceptability and usability of a decision support system to encourage safe and effective use of opioid therapy for chronic, noncancer pain by primary care providers. Pain Med. 2010;11(4):575–85.

Wipfli R, Betrancourt M, Guardia A, Lovis C. A qualitative analysis of prescription activity and alert usage in a computerized physician order entry system. Stud Health Technol Inform. 2011;169:940–4.

PubMed   Google Scholar  

Cornu P, Steurbaut S, De Beukeleer M, Putman K, van de Velde R, Dupont AG. Physician’s expectations regarding prescribing clinical decision support systems in a Belgian hospital. Acta Clin Belg. 2014;69(3):157–64.

Ahearn MD, Kerr SJ. General practitioners’ perceptions of the pharmaceutical decision-support tools in their prescribing software. Med J Australia. 2003;179(1):34–7.

Swaby L, Shu P, Hind D, Sutherland K. The use of cognitive task analysis in clinical and health services research - a systematic review. Pilot Feasibility Stud. 2022;8(1):57.

Steyerberg EW. Applications of prediction models. In: Steyerberg EW, editor. Clinical prediction models. New York: Springer-; 2009. pp. 11–31.

Chapter   Google Scholar  

Steyerberg EW, Vergouwe Y. Towards better clinical prediction models: seven steps for development and an ABCD for validation. Eur Heart J. 2014;35(29):1925–31.

Tonekaboni S, Joshi S, McCradden MD, Goldenberg A. What Clinicians Want: Contextualizing Explainable Machine Learning for Clinical End Use. In: Doshi-Velez F, Fackler J, Jung K, Kale D, Ranganath R, Wallace B, Wiens J, editors. Proceedings of the 4th Machine Learning for Healthcare Conference; Proceedings of Machine Learning Research: PMLR; 2019;106:359–80.

Eini-Porat B, Amir O, Eytan D, Shalit U. Tell me something interesting: clinical utility of machine learning prediction models in the ICU. J Biomed Inform. 2022;132:104107.

Muralitharan S, Nelson W, Di S, McGillion M, Devereaux PJ, Barr NG, Petch J. Machine learning-based early warning systems for clinical deterioration: systematic scoping review. J Med Internet Res. 2021;23(2):e25187.

Rudin C. Stop Explaining Black Box Machine Learning Models for high stakes decisions and use interpretable models instead. Nat Mach Intell. 2019;1(5):206–15.

Blythe R, Parsons R, Barnett AG, McPhail SM, White NM. Vital signs-based deterioration prediction model assumptions can lead to losses in prediction performance. J Clin Epidemiol. 2023;159:106–15.

Futoma J, Simons M, Panch T, Doshi-Velez F, Celi LA. The myth of generalisability in clinical research and machine learning in health care. Lancet Digit Health. 2020;2(9):e489–92.

Steyerberg EW, Uno H, Ioannidis JPA, van Calster B, Collaborators. Poor performance of clinical prediction models: the harm of commonly applied methods. J Clin Epidemiol. 2018;98:133–43.

Download references

Acknowledgements

We would like to thank the participants who made time in their busy clinical schedules to speak to us and offer their support in recruitment.

This work was supported by the Digital Health Cooperative Research Centre (“DHCRC”). DHCRC is funded under the Commonwealth’s Cooperative Research Centres (CRC) Program. SMM was supported by an NHMRC-administered fellowships (#1181138).

Author information

Authors and affiliations.

Australian Centre for Health Services Innovation and Centre for Healthcare Transformation, School of Public Health and Social Work, Faculty of Health, Queensland University of Technology, 60 Musk Ave, Kelvin Grove, Brisbane, QLD, 4059, Australia

Robin Blythe, Sundresan Naicker, Nicole White & Steven M McPhail

Princess Alexandra Hospital, Metro South Health, Woolloongabba, QLD, Australia

Raelene Donovan & Andrew McKelliget

Queensland Digital Health Centre, Faculty of Medicine, University of Queensland, Brisbane, QLD, Australia

Ian A. Scott

Digital Health and Informatics Directorate, Metro South Health, Woolloongabba, QLD, Australia

Ian A. Scott & Steven M McPhail

You can also search for this author in PubMed   Google Scholar

Contributions

RB: conceptualisation, data acquisition, analysis, interpretation, writing. SN: data acquisition, analysis, interpretation, writing. NW: interpretation, writing. RD: data acquisition, interpretation, writing. IS: data acquisition, analysis, interpretation, writing. AM: data acquisition, interpretation, writing. SM: conceptualisation, data acquisition, analysis, interpretation, writing. All authors have approved the submitted version and agree to be accountable for the integrity and accuracy of the work.

Corresponding author

Correspondence to Robin Blythe .

Ethics declarations

Ethics approval and consent to participate.

This study was approved by the Metro South Human Research Ethics Committee (HREC/2022/QMS/84205). Informed consent was obtained prior to interview scheduling, with all participants filling out a participant information and consent form. Consent forms were approved by the ethics committee. Participation was entirely voluntary, and could be withdrawn at any time. All responses were explicitly deemed confidential, with only the first two study authors and the participant privy to the research data. Interviews were then conducted in accordance with Metro South Health and Queensland University of Technology qualitative research regulations. For further information, please contact the corresponding author.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Blythe, R., Naicker, S., White, N. et al. Clinician perspectives and recommendations regarding design of clinical prediction models for deteriorating patients in acute care. BMC Med Inform Decis Mak 24 , 241 (2024). https://doi.org/10.1186/s12911-024-02647-4

Download citation

Received : 06 September 2023

Accepted : 23 August 2024

Published : 02 September 2024

DOI : https://doi.org/10.1186/s12911-024-02647-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Clinical prediction models
• Clinical decision support systems
• Early warning score
• Clinical deterioration
• Clinical decision-making

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

• General enquiries: [email protected]