The Why and How of the Integrative Review

Abstract

An effective integrative review can provide important insight into the current state of research on a topic and can recommend future research directions. This article discusses different types of reviews and outlines an approach to writing an integrative review. It includes guidance regarding challenges encountered when composing integrative reviews, such as fair representation of different perspectives and synthesizing that knowledge to yield new insights. An integrative review is of unique value among other types of knowledge-synthesis vehicles, such as narrative or systematic reviews and meta-analyses. Because each has distinctive but important approaches to synthesizing empirical knowledge, our protocol for writing integrative reviews is designed to complement these other knowledge-synthesis vehicles to best advance organization science.

Keywords

philosophy of science meta-analysis quantitative research research design

An integrative review synthesizes and evaluates current knowledge of a topic to provide new insights therein (Torraco, 2005). Integrative reviews are central to Huff’s (2008) “sensemaking/sensegiving circle of scholarship” (p. 4). A thorough and well-written integrative review synthesizes the current state of knowledge of a topic and brings together different “conversations” (to use Huff’s term) thereof that might be rooted in different paradigms (Burrell & Morgan, 1979). An example of this is the integration of information from those studying “identity” from an interpretivist-sociological perspective (Stryker, 1980) and those approaching it from a functionalist-psychological perspective (Baumeister & Tice, 1986). Such reviews offer opportunities for scholars interested in the topic to develop new research programs that might not have emerged from within a single community of practice, with the new research reifying current research trends (Haslam et al., 2017). An integrative review enables the synthesis of knowledge from across research approaches in a fragmented field. Addressing that fragmentation is an overarching goal therein, as evidenced by the high impact factors of journals like the International Journal of Management Reviews or the Academy of Management Annals and the review issues published by other prestigious journals, such as Journal of Management. Given the range of paradigms and traditions in management science, integration of this knowledge and the different ways in which it is created is of significant importance (Cronin & Bendersky, 2012; Gatrell & Breslin, 2017).

Integrative reviews are a unique vehicle for synthesizing existing knowledge. Whereas other vehicles, such as theory development or meta-analyses, and other review types, such as narrative or systematic, share protocols and a general purpose, the unique value of any one of these approaches in the field of organizational science remains unclear. This is in part because even if a particular vehicle is well described, for instance as meta-analyses are, its unique value is rarely compared with other knowledge-synthesis vehicles. If anything, articles on one type of vehicle tend toward critique of other approaches (e.g., for meta-analyses in particular, see Borenstein et al., 2009; Eden, 2002; Williams, 1998). This article seeks to clarify the distinct function integrative reviews serve as a part of the broader panoply of knowledge-synthesis approaches. This clarity should inform how such reviews should be constructed for maximum impact and how they support other types of knowledge integration.

Fundamentally, good integrative reviews provide insight on a topic by synthesizing knowledge across communities of practice (Brown & Duguid, 1991), as do the different conversations in Huff’s (2008) framework. We use the word insight deliberately; the review does not merely point researchers toward new landscapes but allows them to see existing ones from a different perspective. We also use the term communities of practice deliberately to acknowledge discrete groups of researchers who study similar topics using different paradigms, conceptual language, and research traditions, even in their metatheoretical analyses (Burrell & Morgan, 1979; Hogg et al., 1990; Laudan, 1978; Wenger, 1999). Integrating communities of practice counters the tendency for fragmentation among organization scientists (Pfeffer, 1993; Whitley, 1984), where scholars encompass a range of disciplinary backgrounds (Glick et al., 2007; Tranfield et al., 2003; Tranfield & Starkey, 1998) and may self-sort into silos accordingly. The integrative review presents both a unique opportunity and a unique challenge in the integration of knowledge from different communities of practice. Assimilation of different perspectives can spur innovation within a community (Cronin & Loewenstein, 2018), but synthesizing these perspectives can be problematic (Bechky, 2003; Carlile, 2004; Cronin & Weingart, 2007).

This article describes the process through which critical decision points guide the writing of such a review, a schematic for which is presented in Figure 1. First, we discuss whether the integrative review is the most appropriate vehicle for an author to adjudicate between different findings or redirect the current stream of work. We then detail how to compile and characterize existing knowledge from across communities of practice. Finally, we explore how authors might abstract themes from within each community and integrate them into insights that can influence the trajectory of future research. The exploration of a topic via multiple communities’ perspectives provides sensemaking for the authors, and the integration of the different perspectives allows authors to engage in sensegiving therein (Huff, 2008). We thus believe that integrative reviews can propel the field of organizational research forward toward becoming a more integrated science.

Figure 1.

Flow chart for synthesizing research.

Choosing a Knowledge-Synthesis Vehicle

Science accumulates knowledge in part by organizing and reconciling findings from different studies on a given topic. The knowledge-synthesis vehicle is a key component thereof that draws conclusions by assembling specific findings from across existing studies (Greenhalgh, 1997) and evaluating how they fit together. We use topic to refer to an area of study (Cooper, 1998); these can be broad, narrow, general, or contextualized and may be phenomenon-driven or abstract. Topics have associated domain beliefs that are justified by empirical findings within a given community. When domain knowledge is well organized, one can even make sense of ad hoc topics, for instance, “trust among top management teams attempting to structure mergers and acquisitions,” by interpreting the assembly of relevant studies on the components of this topic.

Knowledge-synthesis vehicles are published works that help to organize domain knowledge into coherent and consistent structures. To map the process to Huff’s (2008) cycle of scholarship (Figure 2), authors review academic publications as sensegiving about a topic, but authors must use their own sensemaking processes to arrive at a position on what the collective body of work demonstrates. This is the bottom link in Huff’s cycle, and although it is a single arrow, it does represent an iterative process between the authors and the published work. When written, the authors’ vehicle will be a means of sensegiving back to the field, thus completing Huff’s cycle when the work is presented or published. This upper link is largely outside of the authors’ control but is most effective when the right knowledge-synthesis vehicle is chosen.

Figure 2.

Huff’s (2008) cycle of sensemaking/sensegiving.

Knowledge-synthesis vehicles include theories, meta-analyses, and different types of reviews, all of which possess certain commonalities. Each includes a literature review and is designed to draw broad conclusions, and each advances research on the topic by suggesting new methods and areas of research (Cooper, 1998; Denyer & Tranfield, 2009; Paré et al., 2015; Whetten, 1989). What distinguishes these vehicles from one another is their typical sensegiving function, which is also supported by the synthesis process used to draw conclusions about the topic of review. Each vehicle can perform a variety of sensegiving functions, but the different synthesis processes used by each vehicle make them more or less suited for giving depending on the authors’ purpose and the evidence (i.e., published studies) available. The first question is whether the authors are interested in adjudication or redirection.

Adjudication organizes domain knowledge by eliminating error and producing “settled science” with respect to a topic (Davis, 2015). Redirection organizes domain knowledge by structuring it in such a way that insights that promote new kinds of research emerge. The purposes of adjudication and redirection in knowledge synthesis closely parallel Turner et al.’s (2017) functions of convergent and holistic triangulation in theory development. Convergent triangulation aims to demonstrate more valid conclusions by finding support for a theory across a variety of research strategies, whereas holistic triangulation aims to produces fuller understanding of a phenomenon through the unique insights/perspectives gained by using different research strategies. The spirit of adjudication and redirection are also to be found in Edmondson and McManus’s (2007) discussion of when quantitative versus qualitative methods should be used. When looking to test claims (i.e., settle them), one uses quantitative methods. When looking to explore new claims, one uses qualitative. We draw these parallels to demonstrate how the choice of adjudication versus redirection parallels choices made in designing theoretical and empirical work. Yet because the intended output is potentially a review article, meta-analysis, or theory, the nature of the choice has important distinctions.

Is Adjudication Desirable and Possible?

Authors seeking to adjudicate on a topic will take a verifiable position about what is true with respect to a topic given the existing evidence. For example, authors seeking to assess whether task conflict is or is not helpful for team performance (De Dreu & Weingart, 2003) or those seeking to narrow down the drivers, facilitators, and hindrances of business model innovation (Foss & Saebi, 2017) would be seeking to adjudicate. Adjudication is best suited to meta-analysis or systematic reviews because these approaches seek definitive answers about potential causal relationships based on objective and systematic standards (Paré et al., 2015) and thus provide high confidence in what the preponderance of the evidence says with respect to a given relationship.

Adjudication provides sensegiving to the field by clarifying and strengthening the foundation of what is known. This can scaffold further research. For example, Joshi and Roh’s (2009) meta-analysis of the role of context in the relationship between work team diversity and firm performance has been cited more than 450 times and has triggered a wealth of studies examining potential moderators, such as educational diversity in top management teams in technology firms (del Carmen Triana et al., 2019) or perceptions of diversity management practices in firms across multiple sectors in the United Kingdom (Otaye-Ebede, 2019).

Adjudication helps produce reliable knowledge that can be used by both researchers and practitioners. Mathieu et al. (2019) reviewed evidence across a range of studies on team structural and compositional features that most reliably affected performance. Such adjudication was useful for those interested in both studying team design as well as those seeking to use evidence-based principles to design teams. Sometime adjudication can be constructed specifically for practitioners. The systematic review process by Briner and Rousseau (2011, Table 3, p. 17) on whether flexible working conditions affect employee health and well-being would be directly relevant for practicing human resource professionals. No matter the audience for whom sensegiving is intended, adjudication requires enough comparable studies to determine what findings are valid and reliable. For this reason, the effectiveness of adjudication is very much dependent on the evidence available, and this also affects the choice of adjudication vehicle.

Choosing the adjudication vehicle

By its nature, adjudication addresses a specific question in light of several studies to arrive at a definitive conclusion. Meta-analyses are a natural fit to this purpose, but they tend to have a high evidentiary demand. They are most effective when specific studies measure and analyze similar constructs in controlled settings with consistency and clarity. Such conditions enable quantification of effect sizes, variable interdependence, and homogeneity as well as potential artifacts and moderators (Hunter & Schmidt, 2004). However, such quantitative studies are not always possible for certain issues and concerns, particularly outside purely scientific fields (Geyskens et al., 2009). In addition, one may wish to adjudicate a question with respect to a particular context, intervention, mechanism, or outcome where there are simply not many studies. When study constructs are not easily quantified or when adjudication applies to a narrower set of circumstances, one may not have the quantitative evidence required to run a meta-analysis.

When there is insufficient quantitative evidence, a systematic review is still a useful adjudicative vehicle. This is a judgment call, but Greenalgh (1997) discussed criteria that can inform the choice. Should one decide on a systematic review, evidence comparison remains algorithmic and mostly objective, even among studies with different quantitative methods (Denyer et al., 2008; Denyer & Tranfield, 2009). There are, for instance, hierarchies of evidence used by those who advocate systematic reviews to evaluate the credibility of primary source studies (see Denyer & Tranfield, 2009, Table 39.1). There are also guides for literature inclusion and bounding the reviewed phenomenon (Briner & Rousseau, 2011).

Other synthesis vehicles are less well adapted for adjudication. By definition, a theory presents relationships that are untested and thus cannot adjudicate. Narrative reviews, which correspond to the typical literature review, may only seek to organize findings rather than settle what is and is not true (Cropanzano, 2009). Furthermore, often the studies included in a narrative review are at the discretion of the authors rather than some algorithm (as they are with systematic reviews). This idiosyncratic study-selection process is why some have criticized the accuracy of the position taken by narrative reviews (Greenhalgh, 1997; Williams, 1998). As we argue shortly, theory and narrative reviews are better suited for redirection.

Integrative reviews are typically used to bridge different communities of practice. Thus, integrative reviews could, conceivably, use an algorithmic approach to collect studies from across communities of practice. The problem emerges when communities of practice have varying perspectives that shape each community’s assumptions about the nature of reality and about which methods should be applied to draw inferences (Burrell & Morgan, 1979). When studies are based in different metatheoretical or paradigmatic differences (Hogg et al., 1990; Miller, 2007; Wagner & Berger, 1985), then without a single set of standards by which to judge the validity of studies’ findings, adjudication may not be feasible. It means that what makes an integrative review useful, the integration across communities of practice, diminishes its capacity to adjudicate.

Limits to adjudication

Although some communities of practice use different language to describe the same phenomenon, they typically share a common paradigm. For example, the communities that study status versus respect are examining the same phenomenon (see Blader & Yu, 2017), and even though their research traditions are different, both share a common experimental social-psychological approach. This may not always be true. Consider research on employees’ choices to help each other. Those who study this as a mixed-motive situation will have a coherent metatheory that guides both the construction of studies and the interpretation of the results. Those who study the choice to help as a question of identity will have a different yet still coherent metatheory. The conclusions from each community cannot be easily reconciled given the fundamental differences between the research approach of each community. The very term identity might not be a legitimate construct to a game theorist, whereas using mathematical evidence to substantiate outcomes might not be legitimate to an experimental psychologist. Each community will explain a deviation from predicted outcomes in such a way as to not contradict their underlying ontological or epistemological assumptions. In such cases, where there is no shared paradigm or theoretical framework with which to resolve a dispute, adjudication is inappropriate.

Choosing Redirection

To redirect is to propose an alteration to the field’s perspective on a topic. Thus, authors seeking to redirect must determine the potentially fruitful avenues of exploration with respect to a topic given the existing research practices. For example, Jang et al. (2018) argued that negotiation research needed to break out of its experimental monoculture, and E. George and colleagues (2006) argued that institutional persistence versus change ought to be studied from the perspectives of individual decision makers. Redirection is best suited to theory, narrative reviews, and integrative reviews because the act of enriching or critiquing a particular area of knowledge is more of a judgmental task, as opposed to an intellective one where correctness is demonstrable (Laughlin, 1980). The authors must make a compelling case for their interpretation of the evidence they review and for what this implies for how a topic is understood.

Redirection changes the thinking about a topic and thus initially provides sensegiving primarily to those already researching that topic. Although redirection of researchers may eventually affect what people believe to be true or how practices are enacted, redirection will at first be more about asking different questions than finding answers. Importantly, redirection need not only guide research on the substance of a topic (e.g., highlighting the role of social class in decision-making in organizations; Coté, 2011), it can also shift the way in which a topic is researched. For example, Cronin and Bezrukova (2019) suggested the study of conflict using system-dynamics approaches. Although this did not change how conflict had been understood, it did expand the ways in which the dynamics of conflict could be empirically tested.

Choosing the redirection vehicle

By its nature, redirection requires finding new insight about a topic through the juxtaposition of several studies. Redirection, therefore, involves disciplined imagination to develop new kinds of ideas that are necessarily speculative out of current domain knowledge, and it foregrounds aspects of the domain in need of more frontline empirical¹ work. Theory is thus well suited to redirection because its function is to illuminate new questions about a topic that could optimally push research therein into unexplored arenas (Bacharach, 1989; Ferris et al., 2012; Whetten, 1989). Narrative and integrative reviews can do this as well because they require authors’ disciplined imagination to glean insights from surveying a topic. To choose whether a theory or a review is most appropriate requires that the authors consider the way that published evidence will be used to support their position.

Theories can be speculative about what existing studies imply, whereas reviews focus on the implications already present in the literature. This means that the more an author’s position requires inferences based on what is not explicit in existing research, the more theoretical it will be. We characterize this decision as the determination whether the studies available provide direct or circumstantial evidence for the author’s position. Direct evidence requires little additional inference to support it because what has been reported thus far is acceptably unambiguous with regard to the claim in question. With circumstantial evidence, additional assumptions must be made in support of a claim. For example, Martin and Coté (2019) developed a theory about the experience of individuals who transition between social classes using a mix of findings across domains ranging from cross-cultural psychology (Hong et al., 2000) to expatriate adjustment (Zhu et al., 2016). Although these articles made no claims about social class, Martin and Coté used these findings to reify the cultural challenges that social-class transitioners face. Hence, the Hong et al. (2000) and Zhu et al. (2016) citations represent circumstantial evidence for Martin and Coté’s claim.

Unfortunately, in practice, the distinction between direct and circumstantial evidence is less obvious. Consider the claim that most conflict research is not dynamic because conflict researchers are not trained to think in terms of dynamic causality. Even in the presence of empirical data showing the majority of conflict studies do not test dynamic causality, the explanation for why remains circumstantial, requiring inferences about why the state of conflict research is as it is. Supposing the authors found an article that validates their claim about how conflict researchers are trained, such as Waller et al. (2016); what was strictly circumstantial evidence now contains a component of direct evidence. As more studies are brought in to validate aspects of the claim—that conflict research design favors single-period studies (Okhuysen & Richardson, 2007) or that most conflict studies, even longitudinal ones, use an experimental paradigm where causality is unidirectional (Jang et al., 2018)—the more the authors’ conclusion is based on direct rather than circumstantial evidence. This example also illustrates how the evidence base as a whole determines whether a claim has direct evidence to support it.

For redirection, it is therefore important to choose between building a theory and composing a review based on the evidence available rather than the nature of redirection. Both can change some aspect of the approach to research on a topic under study. But whereas the former discusses what those in a community of practice think might be true, the latter focuses on what has emerged as being true when experts step back and look at the collected data. Metaphorically, theory creates blueprints for houses that could be built, whereas reviews report on houses that already exist.²

Choosing an integrative review

To continue with the house metaphor, the choice between narrative and integrative review thus comes down to whether one will glean insights by reflecting on the houses in one’s own community of practice, or neighborhood, or by comparing across different communities of practice, such as neighborhoods across different regions. Communities of practice are not only defined by disciplines but also by people who study such topics in distinctly different ways. The microfoundations movement in strategy, for instance, is predominantly concerned with human information processing, much like microlevel researchers who study human cognition. Yet what is distinctive about the microfoundations perspective is that it is foregrounding social structure and placing less weight on human agency than those who study human cognition (Felin et al., 2015).

Communities of practice may also cluster under a topic, sharing metatheory and paradigm, but still operate in discrete, if adjacent, spaces. Posen et al. (2018) reported that research on how firms alter their behavior in response to performance shortfalls overlooked how organizations identify the cause of the shortfall and conflated discrete elements of the search, including process, outcomes, and start/stop rules. An extreme example of interwoven but distinct communities of practice are the status and respect communities; both are concerned with group member esteem, but each evolved independently, and they remain separate (Blader & Yu, 2017). Communities of practice can be less about disciplines and more about the silos that emerge when different researchers share assumptions about how to characterize and study a topic.

Narrative reviews serve an important function in taking stock of the state of research within a particular community of practice and then suggesting ways in which that community can improve. The integrative review sets itself apart by gathering what is known on a topic from a variety of different communities of practice and juxtaposing these perspectives to spawn new areas of research (Torraco, 2005). It is the integration across communities that is an integrative review’s primary purpose; when narrative reviews integrate, they do so from a singular perspective. Systematic reviews can also integrate multiple perspectives but tend to do so in light of the particular cause-effect being adjudicated.

A schematic of the decision-making process is represented as a flowchart in Figure 3. Although knowledge-synthesis vehicles can both adjudicate and redirect, one or the other is usually selected as the primary goal. For example, the position that status and respect are identical (Blader & Yu, 2017) involves adjudication but is intended to redirect research on esteem in teams. Similarly, the position that despite theory to the contrary, task conflict has a negative effect on team performance (De Dreu & Weingart, 2003) is first an adjudication that may secondarily redirect research on conflict.

Figure 3.

Choice of synthesis vehicle.

An integrative review is primarily a tool for redirection through the synthesis of knowledge from different communities of practice. The integrative review thus performs a critical function in the field of management by building bridges across communities of practice in the field and uncovering connections to other related disciplines. As such, it can provide a breadth of view beyond that of other synthesis vehicles. Next, we provide an approach for writing an integrative review to capitalize on these attributes.

Writing the Integrative Review

An integrative review expands and diversifies knowledge on a topic by integrating knowledge from across the communities of practice that study it (Brown & Duguid, 1991; Carlile, 2002; Wenger, 1999). It examines what, how, and why the topic has been studied in each community. The integrative review borrows techniques from other knowledge-synthesis vehicles: It gathers and evaluates studies (systematic review), describes the landscape of research on a topic (narrative review), evaluates study conclusions regarding specific constructs (meta-analysis), and determines implications for how and why a topic should be studied (theory) moving forward. The representation and knowledge synthesis across different communities of practice, however, change the manner in which these common processes are executed.

Consider the process of collecting and reviewing literature (Phase 2, Figure 2). In some ways, this process is typical—one looks for studies on the topic of review and then distills what these studies show into sets of findings. Yet for an integrative review, an author must explicitly look to find everything that has been published on a topic (i.e., a complete review), which means going into unfamiliar research communities and reconciling unfamiliar concepts and methodologies. In addition, the author must strive to give each community of practice a voice (i.e., balance) in terms of presenting the important findings. With a complete and balanced literature review as the foundation, authors are better equipped to determine what the collection of findings implies for the study of their topic going forward and to thus develop an insightful position on the reviewed work. That position emerges and is refined during thematic synthesis (Phase 3, Figure 2). Because the authors must work with direct evidence, the insights will emerge from examining patterns at higher levels of abstraction. Because the review must be integrative, abstraction must be carried out with coherence and consistency of thematic fit in mind. Although we discuss these steps and phases as a sequential process, we do note that they sometimes require multiple cycles.

Literature Review Process

Study selection determines the raw data that will form the basis of the review. Much like with meta-analyses and other systematic reviews, including a complete and balanced representation of findings relevant to a topic is essential to guard against bias arising from gaps in the included data. A singular preference for publications in high-status journals (Kepes et al., 2012) or for one’s own community of practice can undermine this process and interfere with completeness and balance. Completeness can be challenging if different schools of thought use different jargon for the same construct. Balance is also challenging in that it requires judgments about the quality of findings across studies from different traditions. Authors must reconcile ontological and paradigmatic differences by acknowledging and transcending them. Completeness and balance might seem to be competing goals because as completeness increases, so does the number of perspectives that need to be integrated. When focused properly, however, they are complementary given that balanced representation can distill a complete body of studies into a manageable set of findings.

Completeness

Knowledge-synthesis vehicles all strive for completeness—to have necessary and sufficient relevant studies that bear on the authors’ position. Yet the operationalization of completeness is different depending on the knowledge-synthesis vehicle. In adjudicative vehicles, evidence is gathered on an effect to compare when the effect holds and when it does not and to identify parameters that alter the strength of the effect (Egger et al., 1997). In the simplest case, adjudication asks whether and how X → Y. In this case, completeness means that the full range of data on the X→Y relationship, even those from unpublished studies, must be considered. For the integrative review, an X → Y relationship is assessed as to the different ways X and Y have been conceptualized and juxtaposed in addition to the strength and moderation of the arrow. Thus, completeness means finding the range of communities of practice studying a topic, which in turn implies deliberately considering which other communities might be studying the same topic, albeit using different nomenclature.

To start the sensemaking process in Huff’s (2008) cycle, an efficient search process would be for authors to go first to familiar literatures in their domains of expertise. Much like with systematic reviews, capturing everything that has been written on the topic helps to minimize bias in the results (Briner & Rousseau, 2011; Tranfield et al., 2003). Yet for integrative reviews, the bias is more likely to be toward views consistent with the authors’ community of practice as opposed to significant findings. Authors should aim to uncover all that has been found on the topic, including what is not yet well established or is not part of the dominant or mainstream view. These findings should also be reducible to concepts, mechanisms, and methodologies used to support them.

The next step is to find the other communities that have researched the same topic. Sometimes collecting articles on the topic itself can lead one to find different communities, including some from different disciplines, researching that topic. The topic of compliance in the context of cybersecurity has been studied from a range of perspectives, from computer science (Wash et al., 2014) and cognition (Benenson et al., 2012) to economics (Beautement et al., 2009) and law (Good et al., 2006). These communities draw from different ontologies of concepts based on different metatheories to describe the same phenomenon. In cybersecurity, those leveraging cognitive science characterize password-compliance failure in terms of memory-capacity challenges, whereas those leveraging the economic perspective characterize it in terms of incentives. The intent in this step would be to capture and organize what all these different communities have discovered about the common topic. Research on status and respect, both concerned with esteem for group members, developed and operated separately from one another (Blader & Yu, 2017) despite their paradigmatic proximity. Similar examples of literatures on a common topic developing in parallel are included in reviews by Akinci and Sadler-Smith (2012) and Carpini et al. (2017). The point is that one must deliberately look for different communities; otherwise, their relevance to one another may be missed.

There are multiple ways of identifying communities of practice. One is by examining citation patterns in a manner analogous to factor analysis, wherein relational structures among studies effectively identify distinct communities. Certain software packages, such as CitNetExplorer, can help authors achieve completeness by examining citation patterns to identify core works in parallel fields. These core works often spawn streams of research that have matured into communities of practice.

Other ways to distinguish perspectives include using time period (Akinci & Sadler-Smith, 2012; Carpini et al., 2017) or publication type, as in books, articles in the lay press, and peer-reviewed research. Akinici and Sadler-Smith (2012) defined the different perspectives on intuition in the behavioral, biological, and brain sciences using decades. Although the chronological periods were not deliberate communities per se, each decade nonetheless represented distinct perspectives as the work of behavioral decision theorists (Kahneman et al., 1982) influenced the somatic marker hypotheses of neurologists (Bechara et al., 1997) and then views on cognition in organizations adopted by management scholars (Hodgkinson & Healey, 2008).

Different communities of practice are also identifiable by the conceptual language they use (Carlile, 2004; Cronin & Weingart, 2007; Dougherty, 1992). This presents a practical challenge in how to find what different communities of practice have to say about a topic. But by initiating literature searches using familiar terms in unfamiliar databases, the searches can uncover others’ concepts that are synonymous to the one from one’s own community. This can expand one’s reach both within (cf. Anteby et al., 2016) and beyond one’s disciplinary boundaries (cf. Butler et al., 2016). Most databases contain a thesaurus of index terms that define the preferred names for concepts and provide relationships to similar concepts. This is especially helpful across disciplines, where such definitions can highlight instances where different terms are used to indicate the same concept (e.g., that “problem construction” in macro research is functionally equivalent to “problem finding” in micro research). It can also help uncover instances where different disciplines use the same word to apply nonequivalent concepts (e.g., in management, dynamic refers to change over time; in the system dynamics community, dynamic refers to change over time that stems from an endogenous feedback process).

Balance

Once completeness has been established, the resulting list usually contains more studies than one could review in detail. Thus, the author needs to decide what findings need to be reported from across all those surveyed. This can be guided by a determination of a finding’s credibility. This process would be analogous to those used in systematic reviews to determine the level of faith one should have in a finding based on the method used to demonstrate it (cf. Denyer & Tranfield, 2009, Figure 39.1, p. 675). Yet because studies come from different communities of practice, it necessitates a qualitative approach because different communities of practice use different paradigms to produce different types of evidence, and they also use different conceptual frameworks to communicate findings. As Burrell and Morgan (1979) stated, “Each set of research assumptions identifies a quite separate social-scientific reality. To be located in a particular paradigm is to view the world in a particular way” (p. 24).

A challenge in determining importance outside of one’s home community is that it is tempting to evaluate findings in ways that conform to one’s own norms and traditions (Bartko, 1982; Campanario, 1998; Marsh et al., 2008). This can be an automatic response based in researchers’ functional training (Cronin & Weingart, 2007). Those trained on bench research will automatically focus on the importance of randomization, positive and negative controls, and blinding; those trained in qualitative methods may not have generalizability in focus (Stake, 2005); and still others assume most complex phenomena are not understandable without computational simulation (Sterman, 2000). But because one is reviewing what another community found, it is critical for authors to deliberately set their own norms and traditions aside and use what the published research from each community dictates as important.

Within communities, it is easy to mistake quantity of citations for importance or impact. Yet if impact is measured by the number of studies showing an effect, authors risk sidelining minority viewpoints and overweighing some types of studies. Balancing the findings of studies from multiple methodological traditions or paradigms is necessary to ensure findings of speedier types of studies do not overwhelm the findings of those that are more involved, even within a community of practice. In addition, if the method used seems systematically related to the effects found (e.g., 10 Amazon Turk studies show X→ Y, and two field studies show no effect), reconciliation of the discrepant finding should consider how methodological effects might explain the divergence in results. This is not to say an uncommon view should be given the same weight as a common one but rather, that one must be careful to avoid false equivalency. This is especially important because innovation in a field will almost always start out as anomalous findings (Cronin & Loewenstein, 2018; Foss & Saebi, 2017).

Ultimately, a balanced perspective represents knowledge from across many communities of practice. Within a community, one will apply community-held standards to evaluate evidence collected. Across communities, however, one needs to avoid the temptation to evaluate the findings of one community against the findings of another. In the end, balance is about presenting different perspectives’ findings and limitations as objectively as possible. Table 1 outlines the steps to study selection that we have described to achieve completeness and balance.

Table 1.

Literature Review Process.

Step	Purpose	Suggested Process	Result
Articulate the topic locally. Articulate the concepts and boundary conditions that will guide the literature search using literature within your community of practice/areas of expertise.	To capture the full range of concepts and relationships as well as the nuance among them using a familiar metatheory/paradigm	1. Conduct a standard keyword search using a range of databases. 2. Capture all of what has been published on this topic. 3. Search across all journal levels.	1. A complete list of studies on the topic from within your community 2. A list of concepts, mechanisms (i.e., mediators, moderators, antecedents), and methodologies used to characterize the topic being reviewed
Find other kinds of researchers studying the topic. Use the studies and concepts from Step 1 to identify other communities of practice studying the same topic.	To capture the full range of ways in which the topic is studied NOTE: This step is about locating the studies, so it is not yet about balance.	1. Use software (CiteNetExplorer) to explore citation maps and develop a snowball sampling approach. 2. Plug familiar concepts (identified in Step 1) into the database thesaurus of unfamiliar databases to suggest additional search terms. This may require some interpolation depending on the distance between communities of practice.	1. A complete list of studies on the topic linked to the communities from which they came 2. A list of concepts, mechanisms, and methodologies that can be mapped onto/structured alongside those found in Step 1
Bound identifiable communities of practice around a topic. Use the studies identified in Steps 1 and 2 to determine the communities of practice that are studying the topic.	To capture the perspectives that need to be given voice for a balanced view about what is known on a topic	1. Attempt to draw boundaries around communities of practice. This might include examining citation patterns, seminal works cited, or temporal periods. 2. Red flags include too few methodologies, author networks, metatheories, and paradigms; these suggest that the author has been too narrow in the inclusion criteria with respect to the topic.	1. A way to name and differentiate the communities of practice 2. Identification of the metatheoretical and paradigmatic differences among communities
Distill primary findings from each community of practice. Use the studies within each community’s literature set to determine what that community has demonstrated about the topic.	To distill what is common and unique among the findings from different communities of practice	1. Identify the most current, significant, and demonstrable conclusions that emerge from within each community. 2. Be certain that each community’s standards are used to evaluate the importance of findings.	1. Identification of where findings from across communities converge, diverge, and are disconnected 2. Identification of possible paradigmatic or metatheoretical reasons for divergence

Thematic-Synthesis Process

Having determined what the different communities of practice demonstrate about a topic, the next task is to take a position on what, collectively, this literature demonstrates. This will be the authors’ position and should weave the knowledge reviewed together to present a new framework for research. This framework can bridge communities of practice and lead to new insights and research directions. The cross-pollination of knowledge can also lead to innovations in thought, substance, and method that advance how the topic is studied and in management science more broadly.

A complete and balanced set of studies illuminates what each of the communities says about a topic. But to make sense of the set, authors must determine what these findings mean as an integrated collection. Thus, study synthesis involves integrating the insights gained across communities of practice. Given the number, scope, and breadth of what is being reviewed, the processes of abstraction and integration mirror what happens when composing a dissertation (C. M. Roberts, 2010). That is, rather than delving into each finding, the integrative review seeks to synthesize each community’s current state of knowledge surrounding the topic, seeking out holistic patterns among elements distilled from different communities of practice. Discovering such patterns requires an abstraction of themes from findings among communities, which must then be integrated into structured conceptual frameworks. Table 2 outlines the steps to study synthesis, which we elaborate further in the following.

Table 2.

Thematic Synthesis Process.

Step	Purpose	Process	Result
Abstract the findings of each community into common themes. Themes synthesize common findings across communities.	To create the foundational building blocks for a framework to structure the manuscript and integrate the insights of the different communities.	1. Look for commonalities among community findings to develop themes. 2. Consider the different ways themes might relate to each other.	1. A list of potential thematic categorizations of the findings, possibly at multiple levels 2. A set of ways that the relationships among themes could be described
Explore how the themes relate to each other. Themes should fit together into a coherent whole.	To articulate the potential structures that can organize the themes into a meaningful framework	1. If possible, develop higher order themes out of the relationships among lower order themes. 2. Use higher order themes to examine how multiple themes can be structured together.	A preliminary set of structures that show how the themes and their interrelationships may coalesce into a broader frameworks of meaning
Integrate themes into a coherent whole. Distill the integration of themes to an essential framework that supports the redirection proposed.	To develop a framework that highlights relationships among themes more than within them	1. Use tables or figures to draw attention to the relationships among elements/cells. 2. Explicitly draw attention to the cross-theme relationships in the text.	A “big picture” structure that allows readers to zoom in and out on aspects of the topic
Refine and refocus. Iterate the processes until a maximally elegant framework emerges.	To reduce the emerging framework to its essentials	1. Seek the minimum number of categories and relationships. 2. Consider which higher order categories make lower order relationships salient and use those for efficiency (e.g., ordinal categories are better than nominal ones).	A framework for understanding that does not overwhelm the reader

Abstract and juxtapose themes emerging from communities’ findings

A critical component of the integrative review, abstraction, draws out higher order themes from lower order elements. In a narrative review, such lower order elements might be results from individual studies that can be merged into demonstrated findings, such as concepts or causal linkages, that represent the knowledge base in a community of practice (Paré et al., 2015). In an integrative review, one treats these findings from different communities of practice as lower order elements from which broader themes can be abstracted. Abstraction is thus about generalizing meaning (sensemaking) from the studies reviewed (sensegiving) at a higher order than in a traditional review. Because most of the important findings are going to be localized to their communities of practice, this means the author often will need to construct these higher order themes.

The most common way thematic abstraction is done is bottom-up. Lower order themes are a function of the findings of individual studies, and higher order structures link and organize the lower order themes. Hennessey and Amabile’s (2010) review of the literature on creativity illustrated this mostly bottom-up process. The authors chose to organize the research by level of analysis as the highest order theme, namely, neurological, mental, individual, group, social, cultural, and systems. But within each level, lower order themes were identified using the substance of the investigation of the constituent studies; for instance, within the mental level, studies were themed into those that examined cognition, affect, or training. To be more explicit, J. M. George and Zhou (2007) reported that their study was about affect, and so this motivated Hennesy and Amabile to use affect as a theme within the mental level of abstraction. This kind of bottom-up classification helps identify and organize local communities of practice, but it makes integration difficult. It is why in Hennessey and Amabile’s review, each level of analysis contained its own set of lower order themes that did not naturally integrate; that is, the topics studied in group-level creativity research had no direct connection to those studied by researchers interested in the cultural aspects of creativity.

To be fully integrative, it is not merely the themes but the relationships among them that must be brought into focus. Taking a more top-down approach helps in this regard. To continue with the Hennessey and Amabile (2010) example, their review would be integrative to the degree that the themes at the mental level were juxtaposed with themes in the individual and group levels. Thus, one must take a more top-down approach when abstracting higher order themes. Such higher order themes function to illuminate the ways that lower order themes can coalesce into broader frameworks of meaning, which will ultimately link communities together more explicitly.

Abstracting higher order themes requires consideration of the meaning of the relational structures among lower order themes. Higher order themes must still follow directly from how the reviewed studies are described as direct evidence. Typically, however, because these themes are broad abstractions, they may not be the direct focus of any individual study and instead are only implicit in the research. Zhao et al. (2019) examined how leaders affect conflict in groups in an attempt to integrate two separate communities, leadership research and conflict research. Zhao et al. abstracted two higher order themes to organize the lower order ones used in the review. The first higher order theme was how leaders were involved with conflict, which organized the lower order themes of instigating it, engaging in it, or managing it. Each lower order theme then organized the findings within that category, such as whether leaders who instigate conflict do so through either intentional or unintentional means. These themes were not explicit in the studies but were categories abstracted by Zhao et al. More critically, such a classification structure was designed to integrate findings from across the leadership and conflict research communities, and as such, the focus was on how the categories fit together into a larger sensemaking structure, such as at what point leaders might enter a group’s conflict.

Abstraction and juxtaposition can also be applied to methodological themes. Both Zhao et al. (2019) and Jang et al. (2018) reviewed conflict research; the former abstracted themes related to concepts and cause-effect findings, whereas the latter abstracted themes with respect to method and translation of research findings into applications. Sometimes a review will do both. Vaara et al. (2016) examined the use of narratives in organizations and then categorized narratives as a methodological tool. Narratives as causes could affect stability and alter organizational processes. In addition, narratives as data could be conceptualized as “truth” to be analyzed within a positivist framework or as hegemonic presentations to be deconstructed with a postmodernist approach. In all cases, the abstraction of latent themes is key to identifying useful patterns that may have escaped notice because of the typical within-community focus. Abstraction is how patterns of connection emerge when one looks across multiple communities.

Zooming out is the most effective way to juxtapose the abstractions from direct evidence. For example, some communities discuss “thought worlds” (Dougherty, 1992), others “mental models” (Salas & Cannon-Bowers, 2001), and still others “frames” (Hogg et al., 1990); each of these describes how people make assumptions that guide information processing and reasoned action. Yet finding commonalities should still seek to preserve nuances and critical distinctions within each community. For instance, mental models are typically about specific problems, schemas are more generalized, and frames are imposed on situations. Preserving such differences can lead to interesting new kinds of questions upon integration, such as tracing how mental models become schemas or the drawbacks of imposing an ill-fitting frame. Such questions suggest links among various communities of practice. The different methods of inquiry encountered can broaden the panoply of methodological tools available to the different communities. For instance, the distributed dynamic decision-making task (Serfaty et al., 1998) has been used for decades on mental models—maybe it could also be used to understand the imposition of frames.

In summary, abstraction and juxtaposition are about using the findings from across communities of practice to develop broader themes among findings that can also illuminate the relationships among these themes. First-order themes may be found directly within the questions asked, constructs defined, and methodologies employed in a research community. But first-order themes should be abstracted so that they can be juxtaposed across communities, allowing higher order themes to emerge. Although higher order themes are often implicit, support with direct evidence remains critical. Because the higher order themes will emerge from the authors’ interpretations of themes and their juxtaposition and because there may be multiple ways to juxtapose the abstracted themes, authors should expect to have to develop their integrative framework through multiple iterations.

Integrate and iterate

An integrative review structures themes into a sensemaking framework that maintains completeness and balance while promoting redirection for the study of that topic. An explicit integration structure is critical given that the more communities and studies are reviewed, the harder it is to communicate how they fit together (Newell & Simon, 1972). This is a particular challenge for the integrative review because it tends to cover wide swaths of literature and the abstracted categories and themes may not easily align when bridging communities of practice (Glick et al., 2007). Effective integration strives for elegance and parsimony, and authors should decide how many categories and relationships are necessary to frame the topic and their position therein. It should be, to quote Einstein, “As simple as possible but no simpler.”

Tables and especially figures are important integration tools because they force authors to think about the relationships among the areas of study and not just what has been found within the areas themselves. As an example, Clough and colleagues (2019) abstracted four different forms of capital necessary for successful entrepreneurial ventures: human, social, financial, and other. The insightful contribution to the research on capital procurement came less from this abstraction and more from how they integrated the research on their interrelationships. In their figure (see Figure 4), the weight of the arrows represents how much research has been done in which one form of capital helps procure another (or more of the same, represented by a looping arrow). The figure aptly illustrates which aspects of capital procurement are over- or understudied. It also shows that financial capital is the most common dependent variable and that recursive relationships lack a substantive body of research outside the realm of social capital.

Figure 4.

Sample figure for thematic synthesis. Reproduced from Clough et al. (2019) Figure 1, p. 243. Note: Thickness of arrows is proportional to the volume and strength of the evidence regarding the association of one form of capital (independent variable) with another (dependent variable). Feedback loops represent literature where the same form of capital is both an independent and dependent variable.

When the number and variety of themes that need to be integrated increases, one should employ multiple levels of abstraction to synthesize knowledge into a coherent framework. In Carpini et al.’s (2017) review of 9,299 studies on work performance conducted over the past 40 years, the researchers identified communities of practice using a bibliometric analysis technique (Rip & Courtial, 1984; van Eck & Waltman, 2009), first identifying 97 clusters of research categories that were further abstracted into successively broader themes through the use and cocitation of terms used across studies. Community boundaries were identified algorithmically, but the authors used their own judgment to abstract themes with an eye toward the integration framework that would ultimately be used to present their findings.

The Carpini et al. (2017) article demonstrated how abstraction and integration can coevolve. Emergent themes from the bibliometric analysis were revealed by placing findings in a temporal framework to illuminate how thought had changed across decades. A different theme of study emerged in each decade, and Carpini et al. integrated these themes using a theoretical model of performance (Griffin et al., 2007) that represented proficiency, adaptivity, and proactivity across levels of analysis. The Griffin et al. (2007) framework was itself a tool for integration that could be imposed on themes found in the literature. Note that because the Griffin et al. framework was well established, applying it to the themes identified by the bibliometric analysis was less of speculation and more of pointing out the fit between the two. We raise this point because integration is another place where it is easy to speculate and venture into theory, especially when linkages are tentative. But when a conceptual framework is established, it is often both useful and appropriate to use that framework to make sense of the emerging abstractions.

A challenge integrating themes and categories from across communities of practice is that they do not always align. Some of those who study mindfulness suggest it is possible to be both more focused on and more open to new information (Weick & Putnam, 2006; Weick & Sutcliffe, 2006); however, those who study information processing contend that limited processing capacity means one must sacrifice one for the other (Newell & Simon, 1972; Simon, 1957). Fitting these together necessitates an overarching framework that may not yet exist. Integrating abstractions from different communities means finding a way to reconcile them or pointing to other research that might (Kudesia, 2019).

The authors’ position must be based not simply on collecting findings from different communities but by demonstrating how they all fit together. Insights that redirect future research emerge from connections among communities even when these communities seem to be working at cross-purposes (e.g., Costa et al., 2018; Dose, 1997) or on different parts of a larger sequence (e.g., Posen et al., 2018).

The integrative framework that emerges from the authors’ review resembles a theoretical model; however, an important difference is that the linkages must be justified with empirical evidence. Abstracting to the full picture can be challenging, and where possible, it is recommended that authors use established models to delineate relationships among higher and lower order categories. This will also help bridge and unify domain knowledge rather than creating yet another silo. Ultimately, the integration of themes must clarify how research should proceed, including recommendations for new questions to address and ways efforts could be redirected.

Integrative Reviews in Concert With Other Knowledge-Synthesis Vehicles

Integrative reviews stake a position by assimilating perspectives from across multiple communities of practice. Whether these communities are in different disciplines or parallel silos within a field of study, the goal is to provide guidance for those interested in conducting future research on the topic. Furthermore, the different knowledge-synthesis vehicles should be understood as complementary rather than competing (cf. Aguinis et al., 2011; Borenstein et al., 2009; Shapiro, 1994). Going back to the materials-houses-neighborhoods analogy, the integrative review lets builders (i.e., researchers) think about what might be useful, usable, and desirable to construct by comparing the broad patterns of practice across neighborhoods. Meta-analyses help determine the robustness of material structures in houses and neighborhoods. Theories help to imagine new kinds of houses that might use building materials in new ways and/or improve a neighborhood. Together, these vehicles help advance what is known in the field of management and organizational science. In this last section, we discuss how integrative reviews can be used with other vehicles.

Theories and integrative reviews both seek to explain patterns of influence among constructs. Speculations about how or why particular causal chains work are best served by theory development, whereas comparing and contrasting diverse explanations for a given phenomenon is best approached with an integrative review. Just as a new theory can spur new empirical work, the integrative review can spawn new theories by presenting existing theories in new contexts. Importantly, the integrative review can also act as a governor on the proliferation of untested theories in organization science (Edwards, 2010), validating some and illuminating flaws in others. The integrative review is uniquely equipped to point out redundancies in existing theories that may be functionally equivalent but are articulated quite differently among different communities of practice. Blader and Yu (2017) used an integrative review to demonstrate that status and respect researchers are studying the same topic with different terminology.

Integrative reviews can also help characterize which independent theories are or are not compatible, thereby circumventing the need for more complex theories or moderators. Dose’s (1997) review of work values uncovered a conflict as to whether values are merely attitudinal preferences or something else and showed how each assumption led to a series of other inconsistencies, such as that attitudinal preferences have objects but values do not or that values have standards but attitudes do not. Dose resolved these issues by developing an overarching integrative framework with the morality perspective of Kluckholn (1951).

There remains debate over the differences between an integrative review and a narrative review, but approaching it as a difference in scope can be a useful guideline. Narrative reviews organize and distill what is known within a community of practice and can be used as scaffolding for an integrative review that will compare and contrast patterns and practices among multiple communities. Referring to narrative reviews within an integrative review to help articulate a given community’s abstractions allows the author to focus on synthesizing the different categories and themes and can help keep an integrative review from becoming unwieldy. The integrative review can be an important tool for building bridges across different communities and breaking down silos within the big tent that is organization science (Glick et al., 2007). However, integration is accommodation, not assimilation. All perspectives must be mindfully integrated (Argyris, 1996) to support a future direction without minimizing differences or pitting one against another.

In some cases, existing methods for knowledge creation might be limited. Jang et al. (2018) reported that the majority of what is known about negotiation is based on experiments focused on the bargaining phase thereof, making a case for expanding the methods of investigation. Such reviews explore a process (e.g., emergence, see Waller et al., 2016) or method (narrative analysis, see Vaara et al., 2016), the upshot of which is relevant to multiple communities of practice. To expand methods and paradigms, an integrative review is best because it compels authors to consider the advantages of different approaches that are typically isolated.

There is also the potential for complementarity between integrative reviews and adjudication vehicles (meta-analyses and traditional systematic reviews). Similar to integrative reviews, meta-analyses and systematic reviews synthesize all findings surrounding a conceptual relationship, albeit with a much narrower focus. At times, however, the two devices have been pitted against one another (Aguinis et al., 2011; Borenstein et al., 2009; Shapiro, 1994) in a false dichotomy. Rather, the two have different purposes (adjudication vs. redirection). In addition, we have found that the methods formalized for meta-analyses (Hunter & Schmidt, 2004) and systematic reviews (Denyer et al., 2008; Denyer & Tranfield, 2009; Field & Gillett, 2010; Geyskens et al., 2009) to ensure completeness and balance could be adopted for integrative reviews. Because completeness and balance are conceptually different when looking across communities of practice, some methods, such as those devoted to determining the robustness of a finding, would need to be modified. But others, such as finding file-drawer studies, could be imported directly.

Ideally, integrative reviews can extend practical findings beyond specific meta-analyses. Meta-analyses tend to focus on a single finding or relationship and to reduce contexts to moderators. However, emergent phenomena are not reducible to lower level elements (Morgeson & Hofmann, 1999), so when meta-analyses aggregate across contexts, even when moderators are employed, the important emergent phenomena in particular contexts such as hospitals (Ramanujam & Rousseau, 2006) or high-reliability organizations (K. H. Roberts, 1990) can be lost. In addition, for organization science to provide practical knowledge to managers, results that are fit to their specific contexts are more useful and usable than “universal” associations. The systematic review can elude this obstacle by using a context-intervention-mechanism-outcome framework (Denyer et al., 2008) as a first step toward creating a contextualized meta-analysis wherein only studies on a particular phenomenological setting are included. Integrative reviews can then build on these meta-analyses and systematic reviews by addressing the validity and utility of concepts defined therein by looking across contexts. Systematic reviews can focus researchers on a problem, and integrative reviews can provide evidence that researchers are focused in the right place.

In cases where conclusive agreement on the conceptualization and study of a topic has been established, a meta-analysis can refine this existing knowledge. But if there is an issue with the approach to the research method in some community of practice, then meta-analysis has the potential to deepen the competency trap. Consider Van Knippenberg and Sitkin’s (2013) review of transactional leadership. As they noted, a great deal of research and meta-analyses (Derue et al., 2011; Judge & Piccolo, 2004) had affirmed the impact of the construct on a variety of outcomes. Yet their integrative review found that many of the studies that undergirded these meta-analyses suffered from conceptual and methodological weaknesses that led to a “garbage in–garbage out” problem and recommended abandoning the concept altogether and instead mining existing research for salvageable components, such as examining the utility of a leader’s capacity to communicate a vision, a part of the broader transformational leadership construct.

Borenstein and colleagues (2009) defended meta-analyses against the garbage in–garbage out criticism by suggesting that researchers “anticipate that effects may vary from one kind to the other. It is a further strength of meta-analysis that these differences, if identified, can be investigated formally” (p. 380). This focus on methodological fixes to inclusion criteria, however, did not account for problems that arose due to deeper conceptual concerns. This example illustrates how different knowledge-synthesis vehicles each serve important but discrete functions in the accumulation of knowledge. Rather than debate whether one tool can serve all functions, it is more productive to recognize that even when reviewing literature, one needs the right tool for the right job. Our work not only delineates what knowledge synthesis vehicles are suited for, but forms the third leg of a stool whose other legs are Edmondson and McManus (2017) and Turner et al. (2017); each asks authors to consider whether they are trying to settle a question or ask a new type of question (i.e., adjudication or redirection). Edmondson and McManus (2007) made this point with empirical work, Turner et al. (2017) made the point for theoretical work, and we hope we have done so for reviews.

Conclusion

The integrative review is best used when different communities of practice seem to be working in parallel and where research therein could be improved if their findings were synthesized. This goal can be achieved by seeking out and assessing all existing research on a topic from across communities of practice and providing a balanced representation of each community’s knowledge. This provides a firm foundation from which to abstract themes from the various communities and integrate them into broader patterns of knowledge that can provide new insight into how research on the topic might be improved.

Because integrative reviews use similar processes to other knowledge-synthesis vehicles, we focused on what is distinctive about each of these vehicles to underscore that the purpose of the integrative review is to bridge different communities of practice studying the same topic. In describing the integrative-review writing process and outlining the types of situations that call for an integrative review, we hope to improve the efficacy of producing integrative reviews. We also hope to clarify selection of the knowledge-synthesis vehicle and how these complement each other. Ultimately, we hope this will improve the entire field of organizational research.

The integrative review is a powerful tool for studying organizational phenomena. Writing an integrative review starts with thinking about the intended impact of the authors’ position, and its effectiveness is maximized when it is used in concert with other knowledge-synthesis vehicles, such as theories and meta-analyses. We hope that this article helps organization science researchers and writers in using this tool to advance the field.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Matthew A. Cronin

Notes

References

Aguinis

Pierce

C. A.

Bosco

F. A.

Dalton

D. R.

Dalton

C. M.

(2011). Debunking myths and urban legends about meta-analysis. Organizational Research Methods, 14(2), 306–331.

Akinci

Sadler-Smith

. (2012). Intuition in management research: A historical review. International Journal of Management Reviews, 14(1), 104–122.

Anteby

Chan

C. K.

DiBenigno

(2016). Three lenses on occupations and professions in organizations: Becoming, doing, and relating. The Academy of Management Annals, 10(1), 183–244.

Argyris

(1996). Crossroads—Unrecognized defenses of scholars: Impact on theory and research. Organization Science, 7(1), 79–87.

Bacharach

S. B.

(1989). Organizational theories: Some criteria for evaluation. Academy of Management Review, 14(4), 496–515.

Bartko

J. J.

(1982). The fate of published articles, submitted again. Behavioral and Brain Sciences, 5(2), 199.

Baumeister

R. E.

Tice

D. M.

(1986). Four selves, two motives, and a substitute process self-regulation model. In Baumeister

R. F.

(Ed.), Public self and private self (pp. 63–74). Springer.

Beautement

Sasse

M. A.

Wonham

(2009). The compliance budget: Managing security behaviour in organisations. Proceedings of the 2008 New Security Paradigms Workshop . https://www.nspw.org/papers/2008/nspw2008-beautement.pdf

Bechara

Damasio

Tranel

Damasio

A. R.

(1997). Deciding advantageously before knowing the advantageous strategy. Science, 275(5304), 1293–1295.

10.

Bechky

B. A.

(2003). Sharing meaning across occupational communities: The transformation of understanding on a production floor. Organization Science, 14(3), 312–330.

11.

Benenson

Kroll-Peters

Krupp

. (2012). Attitudes to IT security when using a smartphone. In Federated Conference on Computer Science and Information Systems (FedCSIS) (pp. 1179–1183). Wroclaw.

12.

Blader

S. L.

(2017). Are status and respect different or two sides of the same coin? Academy of Management Annals, 11(2), 800–824.

13.

Borenstein

Hedges

L. V.

Higgins

J. P.

Rothstein

H. R

. (2009). Introduction to meta-analysis. Wiley.

14.

Briner

R. B.

Rousseau

D. M.

(2011). Evidence-based I–O psychology: Not there yet. Industrial and Organizational Psychology, 4(1), 3–22.

15.

Brown

J. S.

Duguid

(1991). Organizational learning and communities-of-practice: Toward a unified view of working, learning, and innovation. Organization Science, 2(1), 40–57.

16.

Burrell

W. G.

Morgan

. (1979). Sociological paradigms and organizational analysis. Heinemann.

17.

Butler

M. J.

O’Broin

H. L.

Lee

Senior

(2016). How organizational cognitive neuroscience can deepen understanding of managerial decision-making: A review of the recent literature and future directions. International Journal of Management Reviews, 18(4), 542–559.

18.

Campanario

J. M.

(1998). Peer review for journals as it stands today—Part 2. Science Communication, 19(4), 277–306.

19.

Carlile

P. R.

(2002). A pragmatic view of knowledge and boundaries: Boundary objects in new product development. Organization Science, 13(4), 442–455.

20.

Carlile

P. R.

(2004). Transferring, translating, and transforming: An integrative framework for managing knowledge across boundaries. Organization Science, 15(5), 555–568.

21.

Carpini

J. A.

Parker

S. K.

Griffin

M. A.

(2017). A look back and a leap forward: A review and synthesis of the individual work performance literature. Academy of Management Annals, 11(2), 825–885.

22.

Clough

D. R.

Fang

T. P.

Vissa

(2019). Turning lead into gold: How do entrepreneurs mobilize resources to exploit opportunities? Academy of Management Annals, 13(1), 240–271.

23.

Cooper

H. M.

(1998). Synthesizing research: A guide for literature reviews (Vol. 2). Sage.

24.

Côté

(2011). How social class shapes thoughts and actions in organizations. Research in Organizational Behavior, 31, 43–71.

25.

Costa

A. C.

Fulmer

C. A.

Anderson

N. R.

(2018). Trust in work teams: An integrative review, multilevel model, and future directions. Journal of Organizational Behavior, 39(2), 169–184.

26.

Cronin

M. A.

Bendersky

(2012). The supply chain for producing quality organizational knowledge. Organizational Psychology Review, 2(1), 54–70.

27.

Cronin

M. A.

Bezrukova

(2019). Conflict management through the lens of system dynamics. Academy of Management Annals, 13(2), 770–806.

28.

Cronin

M. A.

Loewenstein

. (2018). The craft of creativity. Stanford University Press.

29.

Cronin

M. A.

Weingart

L. R.

(2007). Representational gaps, information processing and conflict in functionally diverse teams. Academy of Management Review, 32, 761–773.

30.

Cropanzano

(2009). Writing nonempirical articles for Journal of Management: General thoughts and suggestions. Journal of Management, 35(6), 1304–1311.

31.

Davis

G. F.

(2015). Editorial essay: What is organizational research for? Administrative Science Quarterly, 60(2), 179–188.

32.

De Dreu

C. K. W.

Weingart

L. R.

(2003). Task versus relationship conflict, team performance, and team member satisfaction: A meta-analysis. Journal of Applied Psychology, 88(4), 741–749.

33.

del Carmen Triana

Richard

O. C.

(2019). Gender diversity in senior management, strategic change, and firm performance: Examining the mediating nature of strategic change in high tech firms. Research Policy, 48(7), 1681–1693.

34.

Denyer

Tranfield

(2009). Producing a systematic review. In Buchanan

D. A.

Bryman

(Eds.), The Sage handbook of organizational research methods (pp. 671–689). Sage.

35.

Denyer

Tranfield

Van Aken

J. E.

(2008). Developing design propositions through research synthesis. Organization Studies, 29(3), 393–413.

36.

Derue

D. S.

Nahrgang

J. D.

Wellman

N. E. D.

Humphrey

S. E.

(2011). Trait and behavioral theories of leadership: An integration and meta-analytic test of their relative validity. Personnel Psychology, 64(1), 7–52.

37.

Dose

J. J.

(1997). Work values: An integrative framework and illustrative application to organizational socialization. Journal of Occupational and Organizational Psychology, 70(3), 219–240.

38.

Dougherty

D. J.

(1992). Interpretive barriers to successful product innovation in large firms. Organization Science, 3(2), 179–202.

39.

Eden

(2002). From the editors: Replication, meta-analysis, scientific progress, and AMJ’s publication policy. Academy of Management Journal, 45(5), 841–846.

40.

Edmondson

A. C.

McManus

S. E.

(2007). Methodological fit in management field research. Academy of Management Review, 32(4), 1246–1264.

41.

Edwards

J. R.

(2010). Reconsidering theoretical progress in organizational and management research. Organizational Research Methods, 13(4), 615–619.

42.

Egger

Smith

G. D.

Schneider

Minder

(1997). Bias in meta-analysis detected by a simple, graphical test. The BMJ, 315(7109), 629–634.

43.

Felin

Foss

N. J.

Ployhart

R. E.

(2015). The microfoundations movement in strategy and organization theory. The Academy of Management Annals, 9(1), 575–632.

44.

Ferris

G. R.

Hochwarter

W. A.

Buckley

M. R.

(2012). Theory in the organizational sciences: How will we know it when we see it? Organizational Psychology Review, 2(1), 94–106.

45.

Field

A. P.

Gillett

(2010). How to do a meta-analysis. British Journal of Mathematical and Statistical Psychology, 63(3), 665–694.

46.

Foss

N. J.

Saebi

(2017). Fifteen years of research on business model innovation: How far have we come, and where should we go? Journal of Management, 43(1), 200–227.

47.

Gatrell

Breslin

(2017). Editors’ statement. International Journal of Management Reviews, 19(1), 1–3.

48.

George

Chattopadhyay

Sitkin

S. B.

Barden

(2006). Cognitive underpinnings of institutional persistence and change: A framing perspective. Academy of Management Review, 31(2), 347–365.

49.

George

J. M.

Zhou

(2007). Dual tuning in a supportive context: Joint contributions of positive mood, negative mood, and supervisory behaviors to employee creativity. Academy of Management Journal, 50(3), 605–622.

50.

Geyskens

Krishnan

Steenkamp

J. E. M.

Cunha

P. V.

(2009). A review and evaluation of meta-analysis practices in management research. Journal of Management, 35(2), 393–419.

51.

Glick

W. H.

Miller

C. C.

Cardinal

L. B.

(2007). Making a life in the field of organization science. Journal of Organizational Behavior, 28(7), 817–835.

52.

Good

Grossklags

Thaw

Perzanowski

Mulligan

D. K.

Konstan

(2006). User choices and regret: Understanding users’ decision process about consensually acquired spyware. Faculty Publications, 621. https://scholarlycommons.law.case.edu/faculty_publications/621

53.

Greenhalgh

(1997). How to read a paper: Papers that summarise other papers (systematic reviews and meta-analyses). The BMJ, 315(7109), 672–675.

54.

Griffin

M. A.

Neal

Parker

S. K.

(2007). A new model of work role performance: Positive behavior in uncertain and interdependent contexts. Academy of Management Journal, 50(2), 327–347.

55.

Haslam

S. A.

Cornelissen

J. P.

Werner

M. D.

(2017). Metatheories and metaphors of organizational identity: Integrating social constructionist, social identity, and social actor perspectives within a social interactionist model. International Journal of Management Reviews, 19(3), 318–336.

56.

Hennessey

B. A.

Amabile

T. M.

(2010). Creativity. Annual Review of Psychology, 61, 569–598.

57.

Hodgkinson

G. P.

Healey

M. P.

(2008). Cognition in organizations. Annual Review of Psychology, 59, 387–417.

58.

Hogg

M. A.

Turner

J. C.

Davidson

(1990). Polarized norms and social frames of reference: A test of self-categorization theory of group polarization. Basic and Applied Social Psychology, 11, 77–100.

59.

Hong

Morris

M. W.

Chiu

Benet-Martinez

(2000). Multicultural minds: A dynamic constructivist approach to culture and cognition. American Psychologist, 55(7), 709–720.

60.

Huff

A. S.

(2008). Designing research for publication. Sage.

61.

Hunter

J. E.

Schmidt

F. L.

(2004). Methods of meta-analysis: Correcting error and bias in research findings. Sage.

62.

Jang

Elfenbein

H. A.

Bottom

W. P.

(2018). More than a phase: Form and features of a general theory of negotiation. Academy of Management Annals, 12(1), 318–356.

63.

Joshi

Roh

(2009). The role of context in work team diversity research: A meta-analytic review. Academy of Management Journal, 52(3), 599–627.

64.

Judge

T. A.

Piccolo

R. F.

(2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89(5), 755–768.

65.

Kahneman

Slovic

S. P.

Slovic

Tversky

(1982). Judgment under uncertainty: Heuristics and biases. Cambridge University Press.

66.

Kepes

Banks

G. C.

McDaniel

Whetzel

D. L.

(2012). Publication bias in the organizational sciences. Organizational Research Methods, 15(4), 624–662.

67.

Kluckhohn

(1951). Values and value-orientations in the theory of action: An exploration in definition and classification. In Parsons

Shils

(Eds.), Toward a general theory of action (pp. 388–433). Harvard University Press.

68.

Kudesia

R. S.

(2019). Mindfulness as metacognitive practice. Academy of Management Review, 44(2), 405–423.

69.

Laudan

. (1978). Progress and its problems: Towards a theory of scientific growth (Vol. 282). University of California Press.

70.

Laughlin

P. R

. (1980). Social combination processes of cooperative problem-solving groups on verbal intellective tasks. In Fishbein

(Ed.), Progress in social psychology (Vol. 1, pp. 127–155). Psychology Press.

71.

Marsh

H. W.

Jayasinghe

U. W.

Bond

N. W.

(2008). Improving the peer-review process for grant applications: Reliability, validity, bias, and generalizability. American Psychologist, 63(3), 160–168.

72.

Martin

S. R.

Côté

(2019). Social class transitioners: Their cultural abilities and organizational importance. Academy of Management Review, 44(3), 618–642.

73.

Mathieu

J. E.

Gallagher

P. T.

Domingo

M. A.

Klock

E. A.

(2019). Embracing complexity: Reviewing the past decade of team effectiveness research. Annual Review of Organizational Psychology and Organizational Behavior, 6, 17–46.

74.

Miller

(2007). Paradigm prison, or in praise of atheoretic research. Strategic Organization, 5(2), 177–184.

75.

Morgeson

F. P.

Hofmann

D. A.

(1999). The structure and function of collective constructs: Implications for multilevel research and theory development. Academy of Management Review, 24(2), 249–265.

76.

Newell

Simon

H. A

. (1972). Human problem solving. Prentice-Hall.

77.

Okhuysen

G. A.

Richardson

H. A

. (2007). Group conflict as an emergent state: Temporal issues in the conceptualization and measurement of disagreement. In Thompson

Behfar

K. J.

(Eds.), Conflict in teams (pp. 145–179). NU Press.

78.

Otaye-Ebede

(2019). Antecedents and outcomes of managing diversity in a UK context: Test of a mediation model. The International Journal of Human Resource Management, 30(18), 2605–2627.

79.

Paré

Trudel

M.-C.

Jaana

Kitsiou

(2015). Synthesizing information systems knowledge: A typology of literature reviews. Information & Management, 52(2), 183–199.

80.

Pfeffer

(1993). Barriers to the advance of organizational science: Paradigm development as a dependent variable. Academy of Management Review, 18(4), 599–620.

81.

Posen

H. E.

Keil

Kim

Meissner

F. D.

(2018). Renewing research on problemistic search—A review and research agenda. Academy of Management Annals, 12(1), 208–251.

82.

Ramanujam

Rousseau

D. M.

(2006). The challenges are organizational not just clinical. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior, 27(7), 811–827.

83.

Rip

Courtial

(1984). Co-word maps of biotechnology: An example of cognitive scientometrics. Scientometrics, 6(6), 381–400.

84.

Roberts

C. M

. (2010). The dissertation journey: A practical and comprehensive guide to planning, writing, and defending your dissertation. Corwin Press.

85.

Roberts

K. H.

(1990). Managing high reliability organizations. California Management Review, 32(4), 101–113.

86.

Salas

Cannon-Bowers

J. A.

(2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471–499.

87.

Serfaty

Entin

E. E.

Johnston

J. H

. (1998). Team coordination training. In Cannon-Bowers

J. A.

Salas

(Eds.), Making decisions under stress: Implications for individual and team training (pp. 221–245). American Psychological Association.

88.

Shapiro

(1994). Meta-analysis/shmeta-analysis. American Journal of Epidemiology, 140(9), 771–778.

89.

Simon

H. A

. (1957). Models of man; social and rational. Wiley.

90.

Stake

R. E

. (2005). Qualitative case studies. In Denzin

N. K.

Lincoln

Y. S.

(Eds.), The Sage handbook of qualitative research (pp. 443–466). Sage Publications.

91.

Sterman

J. D

. (2000). Business dynamics: Systems thinking and modeling for a complex world. McGraw-Hill Education.

92.

Stryker

. (1980). Symbolic interactionism: A social structural version. Benjamin-Cummings Publishing Company.

93.

Torraco

R. J.

(2005). Writing integrative literature reviews: Guidelines and examples. Human Resource Development Review, 4(3), 356–367.

94.

Tranfield

Denyer

Smart

(2003). Towards a methodology for developing evidence-informed management knowledge by means of systematic review. British Journal of Management, 14(3), 207–222.

95.

Tranfield

Starkey

(1998). The nature, social organization and promotion of management research: Towards policy. British Journal of Management, 9(4), 341–353.

96.

Turner

S. F.

Cardinal

L. B.

Burton

R. M.

(2017). Research design for mixed methods: A triangulation-based framework and roadmap. Organizational Research Methods, 20(2), 243–267.

97.

Vaara

Sonenshein

Boje

(2016). Narratives as sources of stability and change in organizations: Approaches and directions for future research. The Academy of Management Annals, 10(1), 495–560.

98.

van Eck

Waltman

(2009). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84(2), 523–538.

99.

Van Knippenberg

Sitkin

S. B.

(2013). A critical assessment of charismatic—Transformational leadership research: Back to the drawing board? The Academy of Management Annals, 7(1), 1–60.

100.

Wagner

D. G.

Berger

(1985). Do sociological theories grow? American Journal of Sociology, 90(4), 697–728.

101.

Waller

M. J.

Okhuysen

G. A.

Saghafian

(2016). Conceptualizing emergent states: A strategy to advance the study of group dynamics. The Academy of Management Annals, 10(1), 561–598.

102.

Wash

Rader

Vaniea

Rizor

(2014). Out of the loop: How automated software updates cause unintended security consequences. In Symposium on Usable Privacy and Security (pp. 89–104). USENIX Association. https://www.usenix.org/system/files/soups14-paper-wash.pdf

103.

Weick

K. E.

Putnam

(2006). Organizing for mindfulness: Eastern wisdom and Western knowledge. Journal of Management Inquiry, 15(3), 275–287.

104.

Weick

K. E.

Sutcliffe

K. M.

(2006). Mindfulness and the quality of organizational attention. Organization Science, 17(4), 514–524.

105.

Wenger

(1999). Communities of practice: Learning, meaning, and identity. Cambridge University Press.

106.

Whetten

D. A.

(1989). What constitutes a theoretical contribution? Academy of Management Review, 14(4), 490–495.

107.

Whitley

(1984). The fragmented state of management studies: Reasons and consequences. Journal of Management Studies, 21(3), 331–348.

108.

Williams

(1998). The pitfalls of narrative reviews in clinical medicine. Annals of Oncology, 9(6), 601–605.

109.

Zhao

E. Y.

Thatcher

S. M.

Jehn

K. A.

(2019). Instigating, engaging in, and managing group conflict: A review of the literature addressing the critical role of the leader in group conflict. Academy of Management Annals, 13(1), 112–147.

110.

Zhu

Wanberg

C. R.

Harrison

D. A.

Diehn

E. W

. (2016). Ups and downs of the expatriate experience? Understanding work adjustment trajectories and career outcomes. Journal of Applied Psychology, 101(4), 549–568.