TRENDS IN THEORY BUILDING AND THEORY TESTING: A FIVE

TRENDS IN THEORY BUILDING AND THEORY TESTING: AFIVE-DECADE STUDY OF THE ACADEMY OF

MANAGEMENT JOURNAL

JASON A. COLQUITTCINDY P. ZAPATA-PHELAN

University of Florida

We introduce a taxonomy that reflects the theoretical contribution of empirical articlesalong two dimensions: theory building and theory testing. We used that taxonomy totrack trends in the theoretical contributions offered by articles over the past fivedecades. Results based on data from a sample of 74 issues of the Academy of Man-agement Journal reveal upward trends in theory building and testing over time. Inaddition, the levels of theory building and testing within articles are significantpredictors of citation rates. In particular, articles rated moderate to high on bothdimensions enjoyed the highest levels of citations.

It is difficult to overstate the importance of the-ory to the scientific endeavor. Theory allows scien-tists to understand and predict outcomes of inter-est, even if only probabilistically (Cook &Campbell, 1979; Kerlinger & Lee, 2000). Theoryalso allows scientists to describe and explain aprocess or sequence of events (DiMaggio, 1995;Mohr, 1982). Bacharach (1989) suggested that the-ory prevents scholars from being dazzled by thecomplexity of the empirical world by providing alinguistic tool for organizing it (see also Dubin,1976; Hall & Lindzey, 1957). In Brief and Duk-erich’s (1991) terms, theory acts as an educationaldevice that can raise consciousness about a specificset of concepts. Finally, Kerlinger and Lee (2000:11) went so far as to describe theory as the basicaim of science.

Many scholars define theory in terms of relation-ships between independent and dependent vari-ables. For example, Campbell defined theory as “acollection of assertions, both verbal and symbolic,that identifies what variables are important and forwhat reasons, specifies how they are interrelatedand why, and identifies the conditions underwhich they should be related or not related” (1990:65). From this perspective, a theory is evaluatedprimarily by its ability to explain variance in acriterion of interest (Bacharach, 1989). Other schol-ars have defined theory in terms of narratives andaccounts. For example, DiMaggio defined theory as

“an account of a social process, with emphasis onempirical tests of the plausibility of the narrative aswell as careful attention to the scope conditions ofthe account” (1995: 391). From this perspective, atheory is evaluated primarily by the richness of itsaccount, the degree to which it provides a close fitto empirical data, and the degree to which it resultsin novel insights (Eisenhardt, 1989b).

Although the discussion above reveals multipledefinitions of “theory,” even less agreement existsregarding the meaning of a “theoretical contribu-tion.” Many of the top journals in the managementfield demand that empirical articles make a contri-bution to management theory (Rynes, 2005; Sutton& Staw, 1995; Zedeck, 2003). However, many of thebest-regarded theories in management originated in(and were initially tested in) books, book chapters,or theory outlets such as the Academy of Manage-ment Review. Although a variety of factors couldexplain that trend, one likely reason is that empir-ical articles lack the space needed to fully describethe elements of a theory (Barley, 2006). Given thatlimitation, what exactly does it mean for an empir-ical article to make a theoretical contribution?

The purpose of our study was threefold. First, wecreated a taxonomy that can be used to capturemany of the facets of an empirical article’s theoret-ical contribution. That taxonomy includes two di-mensions: the extent to which an article builds newtheory and the extent to which an article tests ex-isting theory. Second, we used that taxonomy toexamine trends in theoretical contributions overtime, to see if the contributions offered by contem-porary management articles differ from the contri-butions offered by management articles from de-cades past. Third, we examined whether an

We would like to thank Sara Rynes, John Hollenbeck,Bradley Kirkman, Chet Miller, Debra Shapiro, and AmyHillman for their helpful comments on a draft of thiswork. We would also like to thank Jamie Bigayer for herassistance in the gathering of articles.

� Academy of Management Journal2007, Vol. 50, No. 6, 1281–1303.

1281

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download or email articles for individual use only.

article’s position on our taxonomy is predictive ofthe article’s impact on the literature, as judged us-ing citation rates.

The Academy of Management Journal was anideal venue for examining such issues, for fourprimary reasons. First, the five-decade life span ofthe journal (from 1958 to 2007) provides an ex-tended time frame for examining trends in theorytesting and theory building. Second, AMJ is a “bigtent” journal that publishes articles relevant to alldivisions of the Academy of Management, and itpublishes approximately equal numbers of microand macro articles (Biehl, Kim, & Wade, 2006;Schminke & Mitchell, 2003; Wiseman & Skilton,1999). Third, AMJ is one of the most influentialjournals in management, per recent studies of jour-nal citations (Podsakoff, MacKenzie, Bachrach, &Podsakoff, 2005). Fourth, AMJ has emphasized is-sues of theoretical contribution throughout its ex-istence, with its editors frequently using the “Infor-mation for Contributors” and “From the Editors”sections to describe their expectations for theory(e.g., Beyer, 1985; Eden, 2004; Tsui, 1999).

THE THEORETICAL CONTRIBUTION OFEMPIRICAL ARTICLES

One way that empirical articles can make theo-retical contributions is to test theory. The authorsof empirical articles that follow the hypothetico-deductive model use theory to formulate hypothe-ses before testing those hypotheses with observa-tions (Hempel, 1966; Popper, 1965). Platt (1964: 46)described the importance of theory testing in quot-ing the noted biologist W. A. H. Rushton, whowrote, “A theory which cannot be mortally endan-gered cannot be alive.” Indeed, one could arguethat theory testing is particularly important in man-agement because some of the most intuitive theo-ries introduced in the literature wind up being un-supported by empirical research. Building on anearlier review (Miner, 1984), Miner (2003) rated theestimated scientific validity of 73 theories found inthe management literature. A set of organizationalbehavior and strategic management scholars ratedthe perceived importance of the theories to themanagement literature, with the author rating theestimated scientific validity of each theory. Of the73 theories identified in the review, only 25 wererated as high in scientific validity. Such resultsillustrate the importance of theory testing, as suchtesting can temper enthusiasm for appealing butinvalid models.

Another way that empirical articles make a the-oretical contribution is by building theory. Empir-ical articles that follow the inductive model begin

with observations that the authors use to generatetheory through inductive reasoning (Chalmers,1999). Inductive studies can come in a number offorms. For example, theory building from cases in-volves using empirical evidence from one or morecases to create theoretical constructs and proposi-tions (Eisenhardt, 1989b; Eisenhardt & Graebner,2007). Grounded theory involves an iterative pro-cess of collecting and analyzing data in order tobuild a theory about how actors interpret theirdaily realities (Glaser & Strauss, 1967; Locke, 2002;Suddaby, 2006). Ethnography involves gainingfirst-hand experience with a research setting in or-der to build a theory that describes the views ofthose under study (Atkinson, Coffey, Delamont,Lofland, & Lofland, 2002). Regardless of the spe-cific methods used, inductive empirical articlestypically conclude with a set of propositions thatsummarize the resulting theory.

Of course, hypothetico-deductive empirical arti-cles can also build theory, though typically in adifferent fashion. Early tests of a theory are typi-cally concentrated on establishing the validity ofthe theory’s core propositions. In subsequent tests,researchers begin exploring the mediators that ex-plain those core relationships or the moderatorsthat reflect the theory’s boundary conditions. Even-tually, in yet further tests they begin expanding thetheory by incorporating antecedents or conse-quences that were not part of the original formula-tion. Weick (1995) described how empirical articlescan provide “interim struggles” that can help inchscholars forward toward stronger theories. In thisway, the findings, hypotheses, and diagrams foundin a given empirical article might not comprise truetheory (Sutton & Staw, 1995), but they may movethe theorizing in a literature toward maturity(Weick, 1995). Over time, a stream of such studiescan provide the inputs for more comprehensivetheorizing. For example, the first full explication ofgoal setting theory was based on 22 years of empir-ical studies (Locke & Latham, 2004).

Figure 1 introduces a taxonomy that combinesthe dual components of an empirical article’s the-oretical contribution: theory building and theorytesting. As the arc in the figure shows, we suggestthat an empirical article can offer a strong theoret-ical contribution by being strong in theory build-ing, strong in theory testing, or strong in both. Wealso suggest that the two components can be usedto classify empirical articles into five discrete cat-egories, which we refer to as reporters, testers, qual-ifiers, builders, and expanders. Builders, testers,and expanders tend to be higher in their theoreticalcontribution, whereas reporters and qualifiers tendto be lower in their theoretical contribution.

1282 DecemberAcademy of Management Journal

Before describing our taxonomy in more detail, itis important to note two of its limitations. First, aswith any taxonomy, ours can be accused of collaps-ing meaningful distinctions in the interest of parsi-mony. After all, taxonomies—like theories—are at-tempts to eliminate some of the complexity foundin the real world (Bacharach, 1989). Our intentionwas not to capture every nuance of theory buildingand theory testing, but rather to create a tool thatcould be used to chart trends in theoretical contri-butions over time. Second, Figure 1 only captureswhat empirical articles are intended to do—it doesnot capture how well they actually do it. One couldconceive of a third axis that captures how interest-ing a new construct is, how much a new relation-ship adds to a literature, how rigorously a theory istested, or the degree to which the mere writing of apaper makes a contribution in and of itself, apart

from the actual findings presented. These sorts ofissues are clearly critical to the quality of an arti-cle’s theoretical contribution and are likely to besignificant predictors of scholarly impact. Unfortu-nately, coding such issues requires an in-depthcontent expertise that is lacking in a journalwidereview of this type.

Theory Building

The vertical axis of Figure 1 describes levels oftheory building. Our conceptualization of theorybuilding captures the degree to which an empiricalarticle clarifies or supplements existing theory orintroduces relationships and constructs that serveas the foundations for new theory. Many of thearguments used to describe the degrees of theorybuilding on the vertical axis were inspired by

FIGURE 1A Taxonomy of Theoretical Contributions for Empirical Articles

2007 1283Colquitt and Zapata-Phelan

Whetten’s (1989) discussion of what constitutes atheoretical contribution. Although Whetten (1989)was focusing specifically on the qualities of astrong AMR submission, many of his arguments arealso applicable to theory building in empiricalarticles.

The first two points on our theory building axisrepresent relatively low levels of theory building.Replications are attempts to cross-validate the find-ings of earlier empirical studies. Lykken (1968) dis-tinguished between operational replication, inwhich a researcher attempts to duplicate all thedetails of another published study’s methods, andconstructive replication, in which a researcher de-liberately avoids imitation of the earlier study’smethods to create a more stringent test of the rep-licability of the findings (see also Eden, 2002;Tsang & Kwan, 1999). Constructive replications areclearly vital for establishing the external validity ofa study’s findings (Cook & Campbell, 1979; Hen-drick, 1991; Rosenthal, 1991) and key to the accu-mulation of scientific knowledge (Amir & Sharon,1991). However, they offer neither new conceptsnor original relationships (Tsang & Kwan, 1999).

The next point on our theory building axis rep-resents studies that examine effects that have beenthe subject of prior theorizing but not of prior em-pirical study. Like replications, these studies donot add to the ideas present in existing theory, nordo they introduce new relationships or constructs.However, they do open important new avenues fortheory-driven research. As Whetten (1989) argued,a theoretical model is most useful for guiding re-search when the relationships it describes have notyet been tested. Unfortunately, many of the theoriesthat are built are never formally tested. Kacmar andWhitfield (2000) reviewed articles that cited 70AMJ and AMR articles. The results indicated thatthe 70 articles had been cited over 1,500 times butthat the theoretical propositions they offered hadrarely been tested in the reviewed work. Indeed,only 9 percent of the AMR articles that were citedin a given paper were actually tested in that paper.These results reveal the importance of conductingan initial test of a previously built theory, so thatthe theory is more likely to become the focus offuture research.

The third point on our theory building axis rep-resents a moderate level of theory building—arti-cles that introduce a new substantive mediator ormoderator of an existing relationship or process. InWhetten’s (1989) terms, these articles involve add-ing a new “what” (i.e., a construct or variable) to anexisting theory in order to describe “how” a rela-tionship or process unfolds or “where,” “when,” or“for whom” that relationship or process is likely to

be manifested. Such articles represent a moderatelevel of theory building because they do clarify orsupplement existing theory. However, Whetten(1989) suggested that adding one or two variables toan existing model may not fundamentally alter thecore logic of an existing theory.

The next two points on our axis represent highlevels of theory building. Articles that examine apreviously unexplored relationship or process canserve as the foundation for brand new theory. Indescribing AMR submissions that make strong the-oretical contributions, Whetten (1989) noted thateditors ask, What’s new?—specifically gauging thedegree to which a submission changes currentthinking. The more a manuscript represents a rad-ical departure from the extant literature, the morethe field is impacted by the ideas presented withinit. Research suggests that this emphasis on noveltyextends to reviews of empirical articles. For exam-ple, Beyer, Chanove, and Fox’s (1995) analysis ofAMJ review process decisions between 1984 and1987 revealed that articles were more likely to beaccepted by reviewers and editors when the au-thors claimed that their content was novel.

Articles that introduce a completely new con-struct (or significantly reconceptualize an existingone) have the potential to be even more novel. Theintroduction of a new construct creates a radicaldeparture from existing work by generating a num-ber of new research directions that can shape futurethinking. New constructs also represent an originaland unique contribution on the part of authors, asopposed to new relationships between conceptsalready described, though not necessarily linked,in past research. Of course, a critical issue withsuch studies is whether the construct in question isreally new or whether it represents “old wine innew bottles” (Spell, 2001). As in other areas ofscience, there is an ebb and flow to the life cycle ofareas of inquiry in management, with previouslydormant ideas being recycled and repackaged asnew ones.

Theory Testing

The horizontal axis of Figure 1 describes levels oftheory testing. Our conceptualization of theory test-ing captures the degree to which existing theory isapplied in an empirical study as a means of ground-ing a specific set of a priori hypotheses. Many of thearguments used to describe the degrees of theorytesting on the horizontal axis were inspired by Sut-ton and Staw’s (1995) discussion of “what theory isnot.” Specifically, the intermediate points on theaxis represent circumstances in which an articleuses something other than theory to ground hy-


potheses. Although these “theory substitutes” fallshort of supplying all the ingredients described inCampbell’s (1990) definition, they can be arrayedon a continuum ranging from “furthest from the-ory” to “closest to theory” (Weick, 1995).

The first two points on our theory-testing axisrepresent low levels of theory testing. Empiricalarticles that follow the inductive model do not in-clude a priori hypotheses as a starting point, in-stead emphasizing the creation of propositions thatcan be tested in future studies. Such articles maydraw on existing theory to trigger research ques-tions or guide the categorizing of observations (Gla-ser & Strauss, 1967; Locke, 2002; Suddaby, 2006),as Weick observed when noting that his theorizingon wildland firefighting was done “with a head fullof theories” (2007: 16). However, the data that aregathered are not used to explicitly test those theo-ries (Eisenhardt, 1989b; Glaser & Strauss, 1967;Locke, 2002; Suddaby, 2006). Empirical articlesthat follow the hypothetico-deductive model mayalso be low in theory testing when they advance apriori hypotheses that are rooted only in logicalspeculation. Theory testing may be absent fromsuch articles simply because no existing theory isrelevant to the relationships of interest—giving thearticles somewhat of an exploratory character.

The second point on our theory-testing axis rep-resents empirical articles in which predictions aregrounded with reference to past findings. Unlikethe articles described in our preceding paragraph,these articles rely on the extant literature to grounda priori hypotheses. However, that grounding con-sists solely of lists of references to past findings,without explication of all the causal logic thatmight explain those findings. In Sutton and Staw’swords, “References are sometimes used like asmoke screen to hide the absence of theory” (1995:373). A paragraph reciting the findings of past stud-ies can convince the reader that the same sort ofrelationships should be observed in the currentarticle, though an understanding of why those re-lationships might exist would still be lacking (Sut-ton & Staw, 1995).

Articles in which predictions are grounded inpast conceptual arguments offer a moderate level oftheory testing. Here authors attempt to explain whya given relationship or process should exist by de-scribing the logic supplied by scholars in past re-search. However, those conceptual arguments havenot been developed or refined enough to constitutetrue theory, nor do they paint a comprehensivepicture of the phenomenon of interest. Neverthe-less, describing some of the causal logic behind agiven prediction supplies a critical ingredient thatreferences to past findings do not (Sutton & Staw,

1995). A reader is able to understand the justifica-tion for a prediction while connecting that justifi-cation to the existing literature.

The next two points on our axis represent highlevels of theory testing. Empirical articles in whichpredictions are grounded with existing models, di-agrams, and figures come very close to testing ac-tual theory (Weick, 1995). Sutton and Staw (1995)noted that diagrams or figures can explicitly delin-eate the causal connections among a set of vari-ables, though the logical nuances behind the boxesand arrows is often lacking. Still, models, diagrams,and figures provide the symbolic representation oftheory that Campbell (1990) described, and theyoften explicitly indicate the critical mediators andmoderators that govern particular relationships orprocesses.

Finally, the furthest point on our axis representsarticles that ground predictions with existing the-ory. In Sutton and Staw’s (1995) terms, true theorygoes beyond models and diagrams by delving intothe underlying processes that explain relation-ships, touching on neighboring concepts or broadersocial phenomena, and describing convincing andlogically interconnected arguments. Although Sut-ton and Staw (1995) focused on the degree to whichan empirical article contained such discussionwithin its pages, we emphasized the degree towhich such discussion could be found in existingdescriptions of a theory. Those existing descrip-tions may be found in prior empirical articles, the-oretical articles, or books and book chapters thatprovide the space needed to fully explicate a theory(Barley, 2006). For example, an author who usespopulation ecology to ground the predictions setforth in an article can find some elements of thetheory described in seminal journal articles (Han-nan & Freeman, 1977) and will find deeper andmore comprehensive treatments provided in books(Carroll & Hannan, 2000; Hannan & Freeman,1989).

Five Discrete Article Types

If we consider the theory-building and theory-testing axes of our taxonomy simultaneously, fivedistinct types of articles are evident (see Figure 1).We define reporters as empirical articles that pos-sess relatively low levels of both theory buildingand theory testing. For example, Martinson andWilkening (1984) conducted an examination of ru-ral-urban differences in job satisfaction, attemptingto replicate research that failed to uncover a signif-icant effect for that background variable. Theirstudy served as a constructive replication of pastresearch, and their three hypotheses were


grounded in references to the findings of the 12prior studies on the topic. Similarly, Cochran andWood (1984) reexamined the relationship betweencorporate social responsibility and financial perfor-mance using statistical tools and methods that weremore advanced than those used in prior studies.Their study therefore represented a constructivereplication that was based largely in references tothe conflicting findings of past research.

Testers are defined as empirical articles that con-tain high levels of theory testing but low levels oftheory building. An exemplar of this category isLee, Mitchell, Wise, and Fireman’s (1996) test ofthe unfolding model of voluntary turnover, a modelthat was introduced in AMR two years earlier (Lee& Mitchell, 1994). Lee et al. (1996) operationalizedthe processes described in the unfolding model toconduct its first direct test. Another representativeexample of the tester category is Stroh, Brett, Bau-mann, and Reilly’s (1996) investigation of the ef-fects of agency-theory-based variables on the com-pensation of middle managers. The authors notedthat agency theory, which is focused on responsesto risk on the part of an organization’s principalsand agents (Eisenhardt, 1989a; Jensen & Meckling,1976), had rarely been tested with middle managersas the sample. They therefore drew on the theory’spropositions to derive hypotheses about the pro-portion of an employee’s compensation that wasvariable.

Qualifiers are defined as empirical articles thatcontain moderate levels of both theory testing andtheory building. Such articles qualify previouslyestablished relationships or processes using con-ceptual arguments rooted in the extant literature.An exemplar of this category is Skarlicki, Folger,and Tesluk’s (1999) examination of personality as amoderator of the relationship between organization-al justice and counterproductive behaviors. The au-thors showed that the justice-counterproductivebehavior link could be qualified by negative affec-tivity and agreeableness and supported those find-ings using conceptual arguments from the person-ality literature. Another article falling into thequalifier category is Nohria and Gulati (1996); theseauthors further examined the relationship betweenslack resources and organizational innovation. Re-sponding to the conflicting findings of past re-search, they showed that the slack-innovation rela-tionship was actually curvilinear, with innovationbeing hindered by either too little slack or too muchslack.

We define builders as articles that are relativelyhigh in theory building but relatively low in theorytesting. Builders include inductive studies that fo-cus on new constructs, relationships, or processes.

For example, Butterfield, Trevino, and Ball (1996)noted that, despite the vast literature on punish-ment, scholars had neglected to examine whatmanagers actually thought and felt about punishingtheir employees. Using a series of interviews, theauthors identified a number of concepts thathelped capture how managers viewed punishment.Builders may also include hypothetico-deductivestudies that examine a relationship that has notbeen the subject of prior theorizing or empiricalresearch. For example, Oldham (1975) conductedthe first study linking supervisor characteristics(e.g., attractiveness, power, trustworthiness) to sub-ordinate acceptance of assigned goals. Because therelationship had not been explored previously, hegrounded his predictions by extrapolating fromstudies linking supervisor characteristics to subor-dinate performance.

Finally, expanders are articles that are relativelyhigh in both theory building and theory testing.Like builders, expanders focus on constructs, rela-tionships, or processes that have not been the sub-ject of prior theorizing, but they conduct that ex-amination while testing some existing theory. In sodoing, they expand a given literature by taking it ina new and different direction. For example, Bate-man and Organ (1983) introduced the citizenshipbehavior construct in a reexamination of the “sat-isfaction causes performance” hypothesis. Drawingon social exchange theory (Blau, 1964) to groundtheir hypotheses, the authors showed that severalfacets of job satisfaction were significantly corre-lated with their measure of citizenship behavior.Similarly, Klassen and Whybark (1999) introduceda new construct, environmental technology portfo-lio, to reflect a firm’s observable pattern of invest-ment used to improve its environmental perfor-mance. Drawing on the resource-based view(Barney, 1991) to justify their predictions, the au-thors showed that the composition of furnituremanufacturing plants’ environmental technologyportfolios was significantly related to the plants’performance.

THEORETICAL CONTRIBUTION:TRENDS AND IMPACT

Having described a tool for capturing the theoret-ical contribution made by empirical articles, wenow focus our attention on how that contributionmight evolve over time and how it might shape thescholarly impact of an article. Turning first totrends over time, we asked, How have the theoret-ical contributions offered by AMJ articles changedover the past five decades? Some predictions can bederived from the literatures on scientific paradigms


and schools of thought (e.g., Cole, 1983; Glick,Miller, & Cardinal, 2007; Kuhn, 1963; Lodahl &Gordon, 1972; McKinley, Mone, & Moon, 1999;Pfeffer, 1993). Scholars in this area have noted thatscientific disciplines vary in their levels of para-digm development, as reflected in the degree ofagreement about research questions, theory, andmethodology in a given discipline (Lodahl & Gor-don, 1972). Disciplines in which consensus existsenjoy a number of potential benefits, includingmore efficient communication among scholars anddecreased barriers to collaboration (Pfeffer, 1993).Both communication and collaboration are criticalingredients for testing a theory, as such efforts re-quire learning about existing theory and framingtests within the larger stream of work on thattheory.

Disciplines with consensus on paradigms arealso more likely to allow editorial teams to empha-size conceptual and methodological rigor over au-thor characteristics when judging journal submis-sions (Pfeffer, 1993). That emphasis on rigorsuggests that such disciplines will generate higherexpectations regarding the theory present in empir-ical articles. Indeed, Cole (1983) suggested thatconsensus on paradigms results in an increasedlevel of theory testing in a literature, and an in-creased rate of obsolescence as new theories re-place flawed predecessors. Popper emphasized theimportance of such obsolescence in writing: “It isnot the accumulation of observations which I havein mind when I speak of the growth of scientificknowledge, but the repeated overthrow of scientifictheories and their replacement by better or moresatisfactory ones” (1965: 215). DiMaggio (1995)echoed such sentiments in noting that the primarycontribution of a particular theory may be servingas a place holder until it inspires a more valid oruseful one. In Kuhn’s (1963) terms, shared para-digms provide the context for “convergent think-ing.” Such thinking is demonstrated when scien-tists conduct incremental research that tests andextends existing theory.

The arguments described above suggest that lev-els of theory testing should increase as manage-ment research attains strong consensus in its theo-retical paradigms. Although scholars acknowledgethat management has much more dissensus in par-adigms than the hard sciences (Glick et al., 2007;Pfeffer, 1993), partially because management is in-terdisciplinary in nature (Rousseau, 2007), the crit-ical question for our purposes concerns whetherthat fragmentation has decreased over the past fivedecades. Clearly the fact that Miner’s (2003) reviewof scientific validity included 73 theories repre-senting the management domain reveals a certain

lack of consensus on paradigms. However, the pastfew decades have seemed to bring an increasedfocus on a smaller set of theories in the micro andmacro domains, including the theories judged byMiner (2003) to be highest in scientific validity.

For example, concepts and models rooted in goalsetting theory (Locke, 1968), expectancy theory(Vroom, 1964), job characteristics theory (Hackman& Oldham, 1976), social exchange theory (Blau,1964), and social learning theory (Bandura, 1977)can be found in many different areas of organiza-tional behavior. Similarly, research driven byagency theory (Jensen & Meckling, 1976), resourcedependence theory (Pfeffer & Salancik, 1978), theresource-based view (Barney, 1991), populationecology (Hannan & Freeman, 1977), and institu-tional theory (DiMaggio & Powell, 1983) can befound in many areas of organization theory andstrategic management. As research efforts on thesetheories have progressed, they seem to have createdmore agreement on the theoretical paradigms thatare represented in the pages of top managementjournals. At the same time, methodological trends,such as the increased use of structural equationmodeling (e.g., Henley, Shook, & Peterson, 2006;James, Mulaik, & Brett, 2006), seem to have broughtmore consistency in the approaches used to testthose theories. As a result, our first prediction wasthat theory-testing levels would exhibit an upwardtrend over the past five decades, a trend partiallysymptomatic of increasing agreement about keytheories and appropriate methods.

What does that theory testing trend suggest aboutlevels of theory building over time? Kuhn’s (1963)classic term “essential tension” describes how theconvergent thinking created by theoretical consen-sus actually fosters, rather than inhibits, the “diver-gent thinking” needed for path-breaking research.Specifically, Kuhn (1963) argued that scholars needto be well versed in a current way of thinkingbefore they can recognize the gaps in scientificunderstanding that trigger the building of new the-ories. Similarly, McKinley et al. (1999) describedhow scientific schools of thought are establishedand maintained by a mix of continuity, reflectingtheory testing, and novelty, reflecting new con-structs, relationships, and research directions. Con-tinuity is needed for scholars to understand how towork within a given school of thought, whereasnovelty is needed to attract attention to and interestin the school. Taken together, these arguments sug-gest that theory building will also rise as the para-digms in the management literature become moremature.

The literatures on scientific paradigms andschools also describe the importance of theory test-


ing and theory building to the scholarly impact ofjournal articles. McKinley et al. (1999) drew onhuman information processing arguments to de-scribe when information is more likely to be salientand memorable to scholars. Specifically, they sug-gested that novel information separates a given ar-ticle from the multitude of articles competing for ascholar’s attention. That suggestion echoes Davis’s(1971) emphasis on generating theories that are notjust “true” but “interesting.” McKinley et al. (1999)further emphasized that an article must also in-clude a bridge to a scholar’s existing knowledge tobe included in the “to be read” pile (see also Davis,1971). Creating that bridge supplies the communi-cation efficiency and decreased barriers to collabo-ration that can make a given article impactful to astream of research (Pfeffer, 1993). We therefore ex-pect that the theory building and theory testingassociated with a given empirical article will bepositively associated with its scholarly impact, asjudged by citation rates.

To summarize, in our study we used the taxon-omy shown in Figure 1 to examine three specificquestions: (1) How has the level of theory build-ing in empirical articles changed over time? (2)How has the level of theory testing in empiricalarticles changed over time? and (3) What are theimplications of those trends for the scholarly im-pact of empirical articles? As noted at the outset,AMJ is an appropriate outlet for examining suchquestions because it has long emphasized theo-retical contribution. In fact, one could argue thatAMJ possesses an “organizational culture” thatgives special priority to theoretical concerns.Schein (1985) described an organization’s cultureas existing simultaneously at three levels: as-sumptions (i.e., taken-for-granted beliefs), values(i.e., principles or standards with intrinsicworth), and artifacts (i.e., visible and tangiblemanifestations of those assumptions and values).

Journals, like organizations, can have culturesthat impact the kinds of manuscripts that are sub-mitted to them, how those manuscripts are writtenand framed, and how editors and reviewers receiveand critique them. A secondary focus of our studywas therefore to examine how trends in theorybuilding and theory testing within AMJ corre-sponded to changes in the artifacts that might rep-resent the journal’s culture. The most salient tangi-ble manifestations of AMJ’s values likely include its“Information for Contributors,” which instructs au-thors on how to prepare journal submissions, andits “From the Editors,” which provides a forum foreditors to speak directly to the journal’s readership(Beyer, 1987; Vance, 1967). Given the limitations ofour data, we were unable to directly attribute any

shift in theory-building or theory-testing levels tochanges in these journal artifacts. Still, we presentthis analysis in a descriptive and historical fashionin an attempt to provide some context for the trendsobserved in our data.

METHODS

Data and Sample

The data for our study were taken from AMJarticles published between 1963 and 2007. AMJ’sfirst issue was actually published in April 1958 (forhistorical reviews, see Adams and Davis [1986];Kirkman and Law [2005]; Mowday [1997];Schminke and Mitchell [2003]). However, the firstfive volumes of the journal contained few empiricalarticles, with many issues instead including essays,reviews, and discussions of management educationissues. We therefore began our review in 1963, sothat our coding of theory building and theory test-ing would be based on a larger set of articles. Ourcoding covered all issues of every third volume ofthe journal. As each AMJ editor serves a three-yearterm, our coding therefore included issues fromevery editorial term, beginning with Dalton McFar-land (1961–63) and ending with Sara Rynes (2005–07). We included the first five issues of 2007 in ourreview to be as current as possible. In all, the sam-ple included 16 volumes, 75 issues, and 770 arti-cles (AMJ volumes included four issues until 1991,moved to five in 1992, then began including sixissues in 1993). Of those 770 articles, 667 wereempirical articles that could be coded on theorybuilding and testing. The remaining 103 articleswere methods pieces, introductions to special re-search forums, and conceptual articles written be-fore the launch of AMR in 1976.

Procedures

We used the theory-building and theory-testingaxes shown in Figure 1 to code the articles. Bothaxes were conceptualized as “nearly interval”scales (Schwab, 2005), with the anchor descrip-tions in the figure used to reduce ambiguity, as in abehaviorally anchored rating scale (Smith & Ken-dall, 1963). The first step in data collection in-volved ensuring that the scales in Figure 1 wouldallow us to code the AMJ articles in a reliablemanner. To check reliability, both authors codedarticles from the 1983 volume—a volume that wasnot included in our review. This volume included50 empirical articles. We checked interrater reli-ability using the ICC(1) form of the intraclass cor-relation (James, 1982; Shrout & Fleiss, 1979). The


magnitude of the ICC(1) can be interpreted as thereliability associated with a single assessment of anarticle’s building or testing rating, with high valuesbeing around .30 (Bliese, 2000). The ICC(1) for ourtheory building rating was .51, and the ICC(1) forour theory testing rating was .59. Having estab-lished adequate reliability, the first author codedhalf of each issue included in our review, and thesecond author coded the other half of thosesame issues.

We coded using paper copies of all articles sothat notes could be recorded. When coding theorybuilding, we noted when new or reconceptualizedconstructs were being introduced. This was mostoften evident when a sentence introduced and de-fined a new term but lacked citations to past arti-cles. Citing existing and still-relevant definitions ofa concept in the management literature, even ifthese definitions were only in a second-tier journalarticle or a book chapter, typically prevented anarticle from being coded as introducing a new con-struct. Two other points should be noted about ourcoding of theory building. First, in coding an articleas forwarding a new moderator of an existing rela-tionship or process, we used a broad definition ofmoderation, including identifying curvilinear ef-fects or exploring variations in effects over time.Second, if a mediator or moderator of an existingeffect had been the subject of prior research, thatarticle was coded as an attempt to replicate previ-ously examined findings.

When coding theory testing, we noted when a par-ticular model or theory was being applied to groundpredictions. Two points should be noted about ourcoding of theory testing. First, we focused our codingprimarily on the sections used to ground an article’sa priori hypotheses. This focus was meant to ensurethat an author was using a given theory in a substan-tive fashion rather than merely including some cita-tions to theory in the opening of his or her paper.Second, articles that followed an inductive modeland lacked explicit a priori hypotheses were rated aslow on theory testing. As noted earlier, in writingsuch articles authors may have drawn on existingtheory to guide their research, but the data that weregathered were not used to explicitly test those theo-ries (Eisenhardt, 1989b; Glaser & Strauss, 1967;Locke, 2002; Suddaby, 2006).

In many cases, empirical articles actually repre-sented blends of different theory-building and the-ory-testing components. For example, an articlemight describe an examination of a previously un-explored relationship as a relatively minor facet ofa study while focusing more attention on replicat-ing previously examined effects. Alternatively, anarticle might present some theory-grounded hy-

potheses and describe others as based on past find-ings. In such cases, in coding we averaged ratingson the multiple components while giving moreweight to the ones that were more central to a focalarticle. As a result, we allowed our coding to in-clude half-points in addition to integers (1.5, 2.5,3.5, 4.5).

For analyses involving the five discrete catego-ries in Figure 1, we used the following computa-tions: Reporters were articles that received a ratingof 1 or 2 on theory building and a 1 or 2 on theorytesting. Testers were articles that received a 1 or 2on theory building and a 4 or 5 on theory testing.Articles that received from 2.5 to 3.5 on theorybuilding and from 2.5 to 3.5 on theory testing werecategorized as qualifiers. Articles that received a 4or 5 on theory building and a 1 or 2 on theorytesting were categorized as builders. Finally, ex-panders were articles that received a 4 or 5 ontheory building and a 4 or 5 on theory testing.

We also printed all of AMJ’s “From the Editors”sections (originally called “Editorial Comments”)and “Information for Contributors” sections (origi-nally called “Suggestions to Authors”). We notedcases in which these sections articulated the jour-nal’s policy on theoretical contributions. Some ofthe cases concerned a change or adjustment to thepolicy, whereas others were efforts to clarify ordeepen potential authors’ understanding of the ex-isting policy. Appendix A summarizes the mostsignificant changes in AMJ’s “Information for Con-tributors” over the past five decades. The Appendixincludes the editor who originally crafted eachstatement, the years in which the statement was ineffect, and an excerpt relevant to the issue of mak-ing a theoretical contribution.

Finally, we assessed the impact of empirical arti-cles using citation counts from the Institute for Sci-entific Information’s (ISI) Social Sciences Citation In-dex (SSCI). The SSCI provides citation counts forarticles published in thousands of journals since1954. It therefore includes data on all AMJ volumes.Citation counts are a commonly used metric for as-sessing the impact, quality, and scientific merit ofjournal articles and have been used to gauge the rel-ative prominence of scholarly journals (Podsakoff etal., 2005; Tahai & Meyer, 1999) and the effects ofarticle, author, and journal-based variables on articleinfluence (Judge, Cable, Colbert, & Rynes, 2007).

RESULTS

Descriptive Statistics

Table 1 presents the means, standard deviations,and zero-order correlations among our theory-


building and theory-testing numbers, along witharticle citations and coded years (ranging from 0 for1963 to 44 for 2007). The theory-building mean was2.52 (s.d. � 1.13), indicating that the typical articlepublished in AMJ during our five-decade span ei-ther examined effects that had been the subject ofprior theorizing or introduced a new mediator ormoderator of an existing relationship or process.The theory-testing mean was 3.10 (s.d. � 1.19),indicating that in the typical article past conceptualarguments were utilized as a means of groundinghypotheses. Table 1 also reveals a weak positivecorrelation between theory building and theorytesting (r � .15), suggesting that levels of the twofacets of a theoretical contribution are largely inde-pendent. The strong positive correlations between

coded year and theory building and theory testingsuggest that more recent articles included higherlevels of the two facets. We explore those trendsmore fully below.

Trends in Theory Building and Theory Testingover Time

Figure 2 presents the trends in theory-buildingand theory-testing levels from 1963 to 2007. Itshould be noted that the 1963 and 1966 averagesare based on only 3 and 5 empirical articles, respec-tively. The 1969 and 1972 averages are based on 17and 16 studies, with the remaining 12 coded yearsaveraging 52 studies. As expected, the trend movesupward for theory building after an initial declineduring the first few editorial terms. A one-wayanalysis of variance (ANOVA) revealed a signifi-cant effect of publication year on theory-buildinglevels (F[15, 666] � 18.83, p � .001). Figure 2 alsoillustrates a fairly steady upward trend for theorytesting. A one-way ANOVA showed that the theorytesting trend was also statistically significant (F[15,666] � 9.98, p � .001).

Another way of examining these trends is to ex-plore the relative frequencies of the reporter, qual-ifier, builder, tester, and expander categories overtime. Figure 3 summarizes these category trends.The graph reveals a decline in the frequency of

TABLE 1Descriptive Statistics and Zero-Order

Correlationsa

Variable Mean s.d. 1 2 3

1. Theory building 2.52 1.132. Theory testing 3.10 1.19 .15*3. Citations 31.45 40.70 .00 .09*4. Year 26.92 11.33 .51* .41* �.06

a n � 667 empirical articles. Year ranges from 0 for 1968 to 44for 2007.

* p � .05

FIGURE 2Trends in Theory Building and Testing from 1963 to 2007


reporters over time, with those articles fillingaround 70 percent of early volumes in the 1970sand early 1980s before declining to a handful in the1990s and none in the 2000s. A chi-square testshowed that this trend was statistically significant(�2[df � 15, n � 667] � 166.86, p � .001). Asignificant increase in qualifiers was observed:only a handful of these were published in the late1970s and 1980s before they rose to 40 percent of athe typical volume in the late 1990s and 2000s(�2[df � 15, n � 667] � 45.32, p � .001). The figurealso reveals an increase in expanders from the late1990s to the 2000s, with their representation reach-ing a peak of 30–40 percent of a typical volume inthe 2000s (�2[df � 15, n � 667] � 56.45, p � .001).In contrast, there was no significant change in therepresentation of either builders (�2[df � 15, n �667] � 18.63, n.s.) or testers (�2[df � 15, n � 667] �19.63, n.s.) over time. Builders hovered around amean of 6 percent of a volume, with no detectabletrend taking place. Testers oscillated around amean of 14 percent of a volume, though theyseemed to be declining from the mid 1990s into the2000s.

Our analyses of AMJ’s “Information for Contrib-utors” and “From the Editors” showed that changesin the journal’s communication about theoryseemed to coincide with shifts in building and test-

ing levels. During the 1960s, the diversity in arti-cles likely made it difficult for the journal to de-velop an identity, as many types of articles wereincluded that are not part of contemporary volumes(e.g., essays, conceptual articles, management edu-cation pieces). Vance provided one of the first ex-plicit articulations of the journal’s “big tent” phi-losophy, noting in a “From the Editors” that “asspace permits, we will try to include the researchedendeavors of classicists and iconoclasts, quantifiersand verbalizers, eclectics and functionalists, empir-icists and conceptualists, behaviorists and noncon-formists” (1967: 7). Although no explicit mentionwas made of theory, that profile clearly allowed forseveral different mixes of theory building and the-ory testing. The “Information for Contributors”statements authored by Dauten in 1958 and Scott in1970 did explicitly mention theory, though theyincluded no details on what constituted a theoret-ical contribution (see the Appendix).

The first significant evolution in the “Informa-tion for Contributors” occurred during Miner’sterm, and the version of the statement Miner au-thored stayed in effect from 1973 to 1984. Theupdated statement now explicitly referenced thetesting of theoretical propositions, while notingthat exploratory research and replications were stillwelcome. Two other historical points about Min-

FIGURE 3Trends in Article Types from 1963 to 2007


er’s term seem relevant. First, the ”research note”category was introduced in 1973 and would con-tinue to exist until 2005. As might be expected,given the description of research notes in the “In-formation for Contributors” in the Appendix, thosearticles tended to have lower levels of theory build-ing. Specifically, the 193 research notes in our dataaveraged a 2.19 for theory building as compared toan average of 2.65 for the 474 articles (t[665] � 4.82,p � .001). No significant differences were found fortheory testing, however, with 3.01 for researchnotes as compared to 3.14 for articles (t[665] �1.34, n.s.). Second, Miner’s term led into the launchof AMR, a companion outlet for theory buildingarticles. Taken together, these events may havecontributed to the high levels of reporters andtesters observed in the 1970s and 1980s.

The next major evolution in the “Information forContributors” occurred during Beyer’s term, andthe statement she wrote stayed in effect from 1985to 1996. For the first time, the statement explicitlyreferred to both theory building and theory testing,using words such as “develops” and “tests.” In herinaugural “From the Editors” essay, Beyer (1985)noted hearing criticisms that AMJ was “dull” and“uninteresting,” and she promised to consider in-terest to readership as a criterion for acceptingmanuscripts. Elsewhere she described a lack ofconvincing theoretical grounding as a major reasonfor manuscripts being rejected, noting, “Even stud-ies that do not intend to advance theory must beplaced within some body of theory to make themscientifically meaningful” (Beyer, 1987: 624). Suchsentiments coincided with a sharp decline in thenumber of reporters and a general increase in the-ory-building and theory-testing levels.

The transition from Tsui’s term to Northcraft’sterm brought a now familiar element to AMJ’s “In-formation for Contributors”: the statement that ar-ticles must make a strong theoretical contribution.The revised statement, in effect from 1999 to 2004,also departed from previous volumes in two otherrespects: Research notes were now expected tomake theoretical contributions (albeit smaller inscope), and replications and incremental researchwere more explicitly discouraged. In addition tothis shift, no fewer than six “From the Editors”essays under Tsui’s, Northcraft’s, and Lee’s termswere used to flesh out the theory requirement inmore detail (Bergh, 2003; Eden, 2002, 2004; Lee,2001; Northcraft, 2000; Rynes, 2002; Schminke,2004; Tsui, 1999). This emphasis correspondedwith increases in theory building and the first clearrise in articles in the expander category.

The last major revision to the “Information forContributors” occurred early in Rynes’s term, in

2005. Although the bolded theoretical contributionstatement remained, the potential avenues for mak-ing that contribution were expanded to includetheory building using inductive or qualitativemethods, the first empirical test of an existing the-ory, meta-analysis with theoretical implications,and constructive replications that clarified theboundaries of a theory. This more multifacetedconceptualization of theoretical contribution ech-oed earlier critiques noting that AMJ had empha-sized novelty and originality at the cost of tests ofexisting theory (Eden, 2004). It also coincided witheditorial board member perceptions that the jour-nal should loosen the theory requirement whilestriving to publish more interesting and innovativeresearch (Rynes, 2005). Comparisons of 2007 and2005 indicate that these changes have been associ-ated with rises in both theory building and theorytesting, along with peak levels of expander articles.

Theory Building, Theory Testing, andArticle Impact

Table 2 presents the results of regression analy-ses in which we assessed the relationship betweentheory building, theory testing, and article cita-tions. Our analyses revealed an inverted U-shapedrelationship between coded year and citations,with articles in the 1950s and 1960s garnering fewcitations and articles in the late 1990s and 2000salso receiving few citations. We therefore con-trolled for squared and cubic versions of coded yearin the first step of our regressions, in which thosevariables explain 24 percent of the variance in ci-tations. The effects of theory testing and theorybuilding are modeled in step 2, where those ratingsexplain an incremental 1 percent of the variance.The unstandardized regression coefficients revealthat a one-unit increase in theory building or theory

TABLE 2Theory Building, Theory Testing, and

Article Impacta

Regression Step and Variable

Citations

R2 �R2 b

1. Year .24* .24* 0.24Year squared 0.20*Year cubed �0.01*

2. Theory building .25* .01* 3.72*Theory testing 3.34*

a n � 667 empirical articles. Year ranges from 0 for 1968 to 44for 2007.

* p � .05


testing ratings was associated with approximately3.5 more citations per empirical article. Note thatthe articles in our database averaged 31.45 citations(s.d. � 40.70), making an increment of 3.5 citationssignificant from a practical perspective. We ex-plored the effects of a theory testing by theorybuilding product term, but that analysis revealedno significant interaction effect.

Another way of examining these trends was toexplore the citations associated with the reporter,qualifier, builder, tester, and expander categories.To test these relationships, we selected only thearticles that we had coded into those five catego-ries, omitting articles that earned low-moderate,moderate-low, moderate-high, or high-moderateclassifications on the theory-building and -testingaxes in Figure 1. Limiting our analyses to the fivenamed categories resulted in a sample of 392 em-pirical articles.

Table 3 shows the results of our regression anal-yses. The three coded year terms explained 26 per-cent of the variance in citations. The four dummycodes representing article categories explained anadditional 3 percent, with reporters as the referentgroup. The unstandardized regression coefficientfor testers shows that articles in that category gar-nered 16 more citations on average than reporters,with builders receiving 13 more citations on aver-age. The citation advantage for qualifiers and ex-panders was larger, with articles in those categoriesreceiving an average of 23 more citations thanreporters.

DISCUSSION

What stands out most from the results of ourstudy is the increase in both theory building and

theory testing in management research, as repre-sented by the 16 volumes of AMJ included in ourreview. The level of theory building in 2007 wasmore than one and a half standard deviationshigher than its lowest level (in 1975), and the levelof theory testing was almost two standard devia-tions higher than its lowest level (in 1963). Thesetrends have impacted the kinds of articles that findtheir way into the management literature. For ex-ample, the reporters that were so common in the1970s and 1980s have become largely extinct in thepages of AMJ, replaced by articles that make a moresignificant theoretical contribution.

In particular, reporters have been replaced byarticles that blend theory building and theory test-ing. These include qualifiers, which have moderatelevels of both, and expanders, which have highlevels of both. Our citation analyses revealed thatqualifiers and expanders are the two most impact-ful kinds of articles published in AMJ, garnering anaverage of 23 more citations than reporters. Sucharticles enjoy additive combinations of the typicalcitation advantages associated with increased the-ory building and testing: 3–4 additional citations,in our data. We suspect that the building-testingbalance explains that impact, as it represents a bal-ance between novelty and continuity. As McKinleyet al. (1999) described, novelty—in the form of anew construct or relationship or a new mediator ormoderator—attracts attention to a given article.Continuity, in turn, provides a bridge to scholars’current understanding, increasing the likelihoodthat an article will be read.

Of course, expanders and qualifiers were not theonly types of articles that were significantly moreimpactful than reporters. Testers enjoyed around16 more citations on average than reporters andlikely benefited from continuity with establishedliteratures and paradigms (McKinley et al., 1999).However, the level at which testers are publishedhas not changed significantly over the past fivedecades in AMJ, and it even seems to have de-creased in recent years. Is that problematic, giventhe critical role that early tests play in the estab-lishment of a new theory, and the importance of aseries of constructive replications to the accumula-tion of knowledge (Amir & Sharon, 1991; Hendrick,1991; Rosenthal, 1991; Tsang & Kwan, 1999)? Toexplore that question, we examined AMJ’s relativepresence in the literatures of 12 major micro theo-ries and 7 major macro theories. Tables 4 and 5present these results. We drew the theories from acombination of sources, including Miner’s (2003)review, the list of theoretical areas that AMJ pro-vides for authors to utilize during the online sub-mission process, and the theories represented in

TABLE 3Five Article Types and Article Impacta

Regression Step and Variable

Citations

R2 �R2 b

1. Year .26* .26* �1.21Year squared 0.29*Year cubed �0.01*

2. Qualifiers .29* .03* 23.30*Builders 12.64†

Testers 15.89*Expanders 22.16*

a n � 392 empirical articles. Year ranges from 0 for 1968 to 44for 2007; b’s are judged with reporters as the reference group.

† p � .10* p � .05


the 16 volumes that we coded. Although our list iscertainly not exhaustive, the theories representmany of the major schools of thought in the microand macro areas of management.

To examine AMJ’s relative presence in these lit-eratures, we searched ISI’s database using thebolded keywords in the tables, restricting the out-put to AMJ, three top micro journals (Journal of

TABLE 4Citations to Micro Theories in AMJ Articlesa

Micro Theories

Academy ofManagement

Journal

Journal ofApplied

PsychologyPersonnel

Psychology

OrganizationalBehavior

and HumanDecisionProcesses

1. Job characteristics theory (Hackman& Oldham, 1976)

.0172 (32/1,858) .0300 (89/2,966) .0086 (34/3,932) .0087 (15/1,729)

2. Expectancy theory (Vroom, 1964) .0064 (16/2,490) .0122 (52/4,259) .0016 ( 8/4,927) .0188 (40/2,124)3. Equity theory (Adams, 1963) .0057 (14/2,464) .0072 (30/4,185) .0020 (11/4,903) .0099 (36/2,124)4. Goal setting theory (Locke, 1968) .0089 (21/2,369) .0195 (76/3,907) .0021 (10/4,699) .0129 (27/2,089)5. Social exchange theory (Blau, 1964) .0104 (26/2,490) .0099 (42/4,259) .0014 ( 7/4,927) .0075 (16/2,124)6. Social identity theory (Tajfel &

Turner, 1979).0113 (19/1,679) .0073 (19/2,620) .0011 ( 4/3,577) .0157 (24/1,529)

7. Social learning theory (Bandura,1977)

.0083 (15/1,798) .0137 (39/2,847) .0031 (12/3,888) .0073 (12/1,651)

8. Cognitive evaluation theory (Deci,1972)

.0033 ( 7/2,135) .0048 (17/3,531) .0000 ( 0/4,304) .0092 (18/1,961)

9. Path goal theory (House, 1971) .0146 (32/2,192) .0058 (21/3,630) .0027 (12/4,504) .0140 (28/1,998)10. Transformational leadership (Burns,

1978).0075 (13/1,741) .0099 (27/2,719) .0011 ( 4/3,734) .0019 ( 3/1,585)

11. Prospect theory (Kahneman &Tversky, 1979)

.0054 ( 9/1,679) .0015 ( 4/2,620) .0000 ( 0/3,577) .0366 (56/1,529)

12. Social information processing theory(Salancik & Pfeffer, 1978)

.0190 (33/1,741) .0162 (44/2,719) .0080 (30/3,734) .0145 (23/1,585)

a Numbers in boldface indicate the journal with the highest levels of testing for a given theory. Words in boldface indicate the exactkeywords used in the search.

TABLE 5Citations to Macro Theories in AMJ Articlesa

Macro Theories

Academy ofManagement

Journal

AdministrativeScience

Quarterly

StrategicManagement

JournalOrganization

Studies

1. Agency theory (Jensen &Meckling, 1976)

.0258 (48/1,858) .0087 (17/1,945) .0415 ( 63/1,518) .0222 (15/676)

2. Resource dependencetheory (Pfeffer & Salancik,1978)

.0057 (10/1,741) .0067 (12/1,798) .0059 ( 9/1,518) .0148 (10/676)

3. Transaction costeconomics (Williamson,1975)

.0088 (17/1,942) .0030 ( 6/2,025) .0389 ( 59/1,518) .0547 (37/676)

4. Resource-based view(Barney, 1991)

.0292 (30/1,027) .0011 ( 1/935) .1537 (164/1,067) .0429 (29/676)

5. Population ecology(Hannan & Freeman, 1977)

.0061 (11/1,798) .0070 (13/1,864) .0033 ( 5/1,518) .0089 ( 6/676)

6. Institutional theory(DiMaggio & Powell, 1983)

.0181 (26/1,439) .0134 (20/1,494) .0147 ( 21/1,428) .0488 (33/676)

7. Contingency theory(Lawrence & Lorsch, 1967)

.0083 (20/2,407) .0018 ( 5/2,740) .0138 ( 21/1,518) .0178 (12/676)

8. Upper echelons theory(Hambrick & Mason, 1984)

.0508 (70/1,378) .0174 (25/1,436) .0642 ( 88/1,370) .0296 (20/676)

a Numbers in boldface indicate the journal with the highest levels of testing for a given theory. Words in boldface indicate the exactkeywords used in the search.


Applied Psychology, Personnel Psychology, and Or-ganizational Behavior and Human Decision Pro-cesses), and three top macro journals (Administra-tive Science Quarterly, Strategic ManagementJournal, and Organization Science). If the name of atheory was a viable search term, we used that name;otherwise the author string for the article introduc-ing the theory was used (note that ISI’s databasedoes not include books or book chapters, so thosesources cannot be used as search terms). The tablepresents the total number of “hits” in the databasefor a search term in each journal, relative to thetotal number of articles that had appeared in thejournal since the first publication of the theory. Forexample, a search for “agency theory” yielded 48hits in AMJ, out of 1,858 articles appearing in thejournal from 1976 on.

Although the differences between journals aresometimes minor, the results reveal that AMJ tendsto have the second most visible presence in microtheory literatures, typically trailing JAP and some-times OBHDP. With respect to macro theories, AMJtends to have either the second or third most visiblepresence; the journal leaders vary considerablyacross the theories. Of the 20 theories included inthe tables, AMJ has the most visible presence foronly three (social exchange theory, path goal the-ory, and social information processing). Of course,these results are largely dictated by AMJ’s “big tent”status, as it strives to maintain a balance of microand macro articles (Schminke & Mitchell, 2003)while appealing to the membership of multipleAcademy of Management divisions (Wiseman &Skilton, 1999). However, it may also be that thejournal’s culture, as represented in its “Informationfor Contributors” and “From the Editors,” has dis-couraged the submission or acceptance of articlesin the tester category. It may be that Rynes’s revi-sion of the “Information for Contributors,” whichencourages the submission of first empirical tests ofa theory (Rynes, 2005; see also Eden, 2004), couldultimately increase the presence of testers. Thisdoes not appear to have occurred as of 2007,however.

Like testers, builders have remained a steadypresence over the past five decades, though appear-ing recently at a lower mean level. They also enjoya citation advantage relative to reporters, garnering13 more citations on average. A number of calls formore qualitative submissions to AMJ seem to haveincreased the number of builders in the journal(Lee, 2001; Rynes, 2005; Suddaby, 2006), to thepoint where they outpaced testers in the last threevolumes we coded. The increase in builders mayalso reflect the increased attention focused on pub-lishing interesting research (Barley, 2006; Bar-

tunek, Rynes, & Ireland, 2006; Rynes, 2005). Forexample, in 2005 the journal added “interesting-ness, innovativeness, and novelty” as an explicitrating category on its reviewer rating form (Rynes,2005), an action with a great deal of culturalsignificance.

One potential concern raised by an increasednumber of builders and expanders is construct pro-liferation—an already fragmented literature becom-ing even more so through the addition of new con-structs (Barley, 2006; Pfeffer, 1993). A relatedconcern is construct redundancy, whereby “new”constructs actually represent older concepts withnew labels (Spell, 2001). Given the increase in ex-panders over time, we wondered whether the newconstructs introduced in AMJ are vulnerable tosuch criticisms or whether they have truly ex-panded theoretical areas in meaningful ways. Toexplore this issue, we included the new and recon-ceptualized constructs uncovered in our review inTable 6 and supplemented that list with an inspec-tion of the other volumes of the journal. AlthoughTable 6 may not offer an exhaustive list, we believeit does reflect the vast majority of constructs intro-duced in AMJ.

The table includes many cases in which an ex-isting concept was reconceptualized or redefinedin the interest of guiding future research (e.g.,Finkelstein, 1992; McAllister, 1995; Rindova, Wil-liamson, Petkova, & Sever, 2005; Washington &Zajac, 2005). The table also includes cases in whicha more specific version of a broader construct wasintroduced (e.g., Pierce, Gardner, Cummings, &Dunham 1989; Tierney & Farmer, 2002). Two note-worthy trends are evident in the table. First, someof the constructs have been very impactful to themanagement literature, most notably citizenshipbehavior (Bateman & Organ, 1983), affect- and cog-nition-based trust (McAllister, 1995), employee de-viance (Robinson & Bennett, 1995), and relationaldemography (Tsui & O’Reilly, 1989). However,other constructs have had less impact, as judgedusing citations, and appear somewhat similar toexisting constructs in the management literature.For example, Kahn’s (1990) personal engagement issimilar to a number of constructs, including jobinvolvement, intrinsic motivation, and organiza-tional commitment. Frese, Kring, Soose, and Zem-pel’s (1996) personal initiative is similar to manyexisting forms of organizational citizenship behav-ior, as is Morrison and Phelps’s (1999) takingcharge construct. Finally, Duffy, Ganster, andPagon’s (2002) social undermining is similar tomany aspects of employee deviance (Robinson &Bennett, 1995). Second, the rate of new and recon-ceptualized constructs being introduced appears to


TABLE 6New Constructs Introduced in AMJ Articles

Construct Definition Citations

Competitive tension (Chen, Su, & Tsai, 2007) The strain between a focal firm and a given rival that islikely to result in the firm taking action against the rival.

0

Cutthroat cooperation (Johnson, Hollenbeck,Humphrey, Ilgen, Jundt, & Meyer, 2006)

The type of cooperation seen among past competitors, asopposed to the type of cooperation seen among those whohave always cooperated.

0

Expertness diversity (Van der Vegt, Bunderson, &Oosterhoff, 2006)

Differences in the level of expertise (i.e., “expertness”) ofteam members.

0

Organizational reputation (Rindova, Williamson,Petkova, & Sever, 2005)

The degree to which stakeholders evaluate an organizationpositively on an attribute (perceived quality) and thedegree to which an organization receives recognition in itsfield (prominence).

0

Servant leadership (Schneider, Ehrhart, Mayer, Saltz, &Miles-Jolly, 2005)

Leadership that communicates a commitment to high levelsof service quality.

0

Organizational status (Washington & Zajac, 2005) A socially constructed, intersubjectively agreed-upon andaccepted ordering or ranking of individuals, groups, organ-izations, or activities in a social system.

2

Technological dynamism (Wu, Levitas, & Priem, 2005) The rate of change in and the unpredictability of newtechnologies.

0

Transient slack (George, 2005) Excess resources available after resource demands foroperations have been met.

2

Creative self-efficacy (Tierney & Farmer, 2002) The belief that one has the ability to produce creativeoutcomes.

15

Cultural competitiveness (Hult, Ketchen, & Nichols,2002)

The degree to which supply chains are predisposed to detectand fill gaps between what the market desires and what iscurrently offered.

16

Intergenerational reciprocity (Wade-Benzoni, 2002) Passing on benefits (or burdens) to future generations as amatter of retaliation for the good (or bad) received frompast generations.

2

Relational job learning (Lankau & Scandura, 2002) Increased understanding about the interdependence orconnectedness of one’s job to others.

18

Social undermining (Duffy, Ganster, & Pagon, 2002) Behaviors intended to hinder, over time, the ability toestablish and maintain positive interpersonalrelationships, work-related success, and favorablereputation.

24

Symbolic isomorphism (Glynn & Abzug, 2002) The resemblance of an organization’s symbolic attributes tothose of others within its institutional field.

13

Job embeddedness (Mitchell, Holtom, Lee,Sablynski, & Erez, 2001)

The extent to which people have links to other people, theextent to which their jobs and communities fit with theirlives, and the ease with which links can be broken.

40

Ecological embeddedness (Whiteman & Cooper, 2000) The extent to which a manager is rooted in the land. 9

Environmental technological portfolio (Klassen &Whybark, 1999)

An observable pattern of investment designed to improve afirm’s environmental performance.

42

Taking charge (Morrison & Phelps, 1999) Voluntary and constructive efforts to effect organizationallyfunctional change with respect to how work is executed.

44

Personal initiative (Frese, Kring, Soose, & Zempel,1996)

A behavior syndrome resulting in an individual’s taking anactive and self-starting approach to work and goingbeyond what is formally required in a given job.

70


be increasing in recent years, which is partiallysymptomatic of the increase in expanders in the2000s. If interestingness, innovativeness, and nov-elty continue to be emphasized in management re-search, then it will be critical to ensure that newand reconceptualized constructs actually add valueto the literature (Pfeffer, 1993; Spell, 2001).

Limitations

Our article has some limitations that should benoted. First, we reiterate our earlier points aboutthe taxonomy in Figure 1. Not only does it collapsepotentially meaningful distinctions in the interestof parsimony, but also, it captures only what au-

thors intended to do in their studies. Certainlysome articles ground predictions in better theorythan others, and some articles examine new con-structs and relationships that are more important(and less redundant) than others. Indeed, we sus-pect that, if such “quality of execution” could bereliably coded, it would explain more variance inarticle citations than our theory-building and theo-ry-testing categorizations. Our taxonomy also ne-glects the quality of authors’ explication. Suttonand Staw (1995) argued that authors should de-scribe exactly why a theory predicts what it does soreaders do not need to consult other sources on thetheory. Some authors clearly have a talent for writ-ing that elevates the contributions of their articles

TABLE 6(Continued)

Construct Definition Citations

Affect- and cognition-based trust (McAllister,1995)

Trust grounded in reciprocated interpersonal care andconcern (affect-based) and individual beliefs about peerreliability and dependability (cognition-based).

278

Change schema (Lau & Woodman, 1995) A mental map representing the knowledge structures ofchange attributes and relationships among differentchange events.

27

Employee deviance (Robinson & Bennett, 1995) Voluntary behavior that violates significant organizationalnorms and in so doing threatens the well-being of anorganization.

133

Archetypes (Greenwood & Hinings, 1993) A set of structures and systems that reflects a singleinterpretive scheme.

81

Top manager power (Finkelstein, 1992) The capacity of individual actors to exert their will as afunction of structures, ownership, expertise, andprestige.

137

Personal engagement (Kahn, 1990) The harnessing of organization members’ selves to theirwork roles such that people employ and expressthemselves physically, cognitively, and emotionallyduring role performances.

75

Task revision (Staw & Boettger, 1990) Taking action to correct a faulty procedure, inaccurate jobdescription, or dysfunctional role expectation.

55

Organization-based self-esteem (Pierce, Garder,Cummings, & Dunham, 1989)

The degree to which organizational members believe thatthey can satisfy their needs by participating in roleswithin the context of an organization.

99

Relational demography (Tsui & O’Reilly, 1989) The comparative demographic characteristics of membersof dyads or work groups who are in a position to engagein regular interactions.

302

Citizenship behavior (Bateman & Organ, 1983) Those gestures (often taken for granted) that lubricate thesocial machinery of the organization but that do notdirectly inhere in the usual notion of task performance.

226

Alienation (Korman, Wittig-Berman, & Lang,1981)

Seeing a discrepancy between one’s everyday behaviorand one’s self-image (personal alienation) and seeing theself as separated from others (social alienation).

45


beyond their particular objectives. That writing,which is often most evident in Discussion sections,may create contributions not reflected inour coding.

Three other limitations of this study lay outsideour taxonomy. First, our coding was limited toevery third volume of AMJ. It remains an empiricalquestion whether the trends observed in our datawould hold with all volumes coded. Second, it maybe that the trends observed in our data would havediffered if other top management journals had beencoded. As noted previously, journals develop theirown particular cultures, which may alter the levels(and impact) of theory testing and building overtime. Third, we utilized citation rates as a means ofcapturing the impact of empirical articles. A keylimitation of citation counts is that they weigh eachcitation equally, regardless of the importance of thecited article to the citing manuscript (Kacmar &Whitfield, 2000). Citation counts are also driven bya number of factors that were not captured in ourstudy, including specific methodological and arti-cle characteristics (Judge et al., 2007).

Conclusion

In their discussion of “what theory is not,” Sut-ton and Staw (1995: 380) raised the following ques-tion: Should management journals strive to publishinnovative theory building and rigorous theorytesting, or is trying for such a balance “a quixoticventure?” Our results suggest that it need not be aquixotic venture. The trends revealed in AMJ overthe past five decades show that theory testing andtheory building are not zero-sum ideals. Both rep-resent key components of theoretical contributionthat can coexist within a given empirical articleand within a given stream of research. Moreover,both have their own unique impacts on the cita-tions of scholarly works, and thus on the accumu-lation and sharing of knowledge.

REFERENCES

Adams, J. S. 1963. Toward an understanding of inequity.Journal of Abnormal and Social Psychology, 67:422–436.

Adams, S. A., & Davis, K. 1986. Academy of ManagementJournal: The first decade. In D. A. Wren & J. A.Pearce III (Eds.), Papers dedicated to the develop-ment of modern management: 89–94. BriarcliffManor, NY: Academy of Management.

Amir, Y., & Sharon, I. 1991. Replication research: A“must” for the scientific advancement of psychol-ogy. In J. W. Neuliep (Ed.), Replication research inthe social sciences: 51–69. Newbury Park, CA: Sage.

Atkinson, P. A., Coffey, A. J., Delamont, S., Lofland, J., &Lofland, L. H. 2002. Handbook of ethnography.Newbury Park, CA: Sage.

Bacharach, S. B. 1989. Organizational theories: Somecriteria for evaluation. Academy of ManagementReview, 14: 496–515.

Bandura, A. 1977. Social learning theory. EnglewoodCliffs, NJ: Prentice-Hall.

Barley, S. R. 2006. When I write my masterpiece:Thoughts on what makes a paper interesting. Acad-emy of Management Journal, 49: 16–20.

Barney, J. 1991. Firm resources and the theory of com-petitive advantage. Journal of Management, 17: 99–120.

Bartunek, J. M., Rynes, S. L., & Ireland, R. D. 2006. Whatmakes management research interesting, and whydoes it matter? Academy of Management Journal,49: 9–15.

Bateman, T. S., & Organ, D. W. 1983. Job satisfaction andthe good soldier: The relationship between affectand employee “citizenship.” Academy of Manage-ment Journal, 26: 587–595.

Bergh, D. 2003. Thinking strategically about contribu-tion. Academy of Management Journal, 46: 135–136.

Beyer, J. M. 1985. From the editor. Academy of Manage-ment Journal, 28: 5–8.

Beyer, J. M. 1987. From the editor. Academy of Manage-ment Journal, 30: 621–624.

Beyer, J. M., Chanove, R. G., & Fox, W. B. 1995. Thereview process and the fates of manuscripts submit-ted to AMJ. Academy of Management Journal, 38:1219–1260.

Biehl, M., Kim, H., & Wade, M. 2006. Relationshipsamong the academic business disciplines: A multi-method citation analysis. Omega, 34: 359–371.

Blau, P. 1964. Exchange and power in social life. NewYork: Wiley.

Bliese, P. D. 2000. Within-group agreement, non-inde-pendence, and reliability: Implications for data ag-gregation and analysis. In K. J. Klein & S. W. J.Kozlowski (Eds.), Multilevel theory, research, andmethods in organizations: Foundations, exten-sions, and new directions: 349–381. San Francisco:Jossey-Bass.

Brief, A. P., & Dukerich, J. M. 1991. Theory in organiza-tional behavior: Can it be useful? In B. M. Staw &L. L. Cummings (Eds.), Research in organizationalbehavior, vol. 13: 327–352. Greenwich, CT: JAIPress.

Burns, J. M. 1978. Leadership. San Francisco: Harper &Row.

Butterfield, K. D., Trevino, L. K. & Ball, G. A. 1996.Punishment from the manager’s perspective: A


grounded investigation and inductive model. Acad-emy of Management Journal, 39: 1479–1512.

Campbell, J. P. 1990. The role of theory in industrial andorganizational psychology. In M. D. Dunnette & L. M.Hough (Eds.), Handbook of industrial and organi-zational psychology, vol. 1: 39–74. Palo Alto, CA:Consulting Psychologists Press.

Carroll, G. R., & Hannan, M. T. 2000. The demography ofcorporations and industries. Princeton, NJ: Prince-ton University Press.

Chalmers, A. F. 1999. What is this thing called science?(3rd ed.). St. Lucia, Australia: University of Queens-land Press.

Chen, M. J., Su, K. H., & Tsai, W. 2007. Competitivetension: The awareness-motivation-capability per-spective. Academy of Management Journal, 50:101–118.

Cochran, P. L., & Wood, R. A. 1984. Corporate social-responsibility and financial performance. Academyof Management Journal, 27: 42–56.

Cole, S. 1983. The hierarchy of the sciences? AmericanJournal of Sociology, 89: 111–139.

Cook, T. D., & Campbell, D. T. 1979. Quasi-experimen-tation: Design and analysis issues for field set-tings: 9–36. Boston: Houghton-Mifflin.

Davis, M. S. 1971. That’s interesting! Toward a phenom-enology of sociology and a sociology of phenomenol-ogy. Philosophy of the Social Sciences, 1: 309–344.

Deci, E. L. 1972. The effects of contingent and non-contingent rewards and controls on intrinsic moti-vation. Organizational Behavior and Human Per-formance, 48: 217–229.

DiMaggio, P. J. 1995. Comments on “what theory is not.”Administrative Science Quarterly, 40: 391–397.

DiMaggio, P. J., & Powell, W. W. 1983. The iron cagerevisited: Institutional isomorphism and collectiverationality in organizational fields. American Socio-logical Review, 48: 147–160.

Dubin, R. 1976. Theory building in applied areas. InM. D. Dunnette (Ed.), Handbook of industrial andorganizational psychology: 17–40. Chicago: RandMcNally.

Duffy, M. K., Ganster, D. C., & Pagon, M. 2002. Socialundermining in the workplace. Academy of Man-agement Journal, 45: 331–351.

Eden, D. 2002. Replication, meta-analysis, scientificprogress, and AMJ’s publication policy. Academy ofManagement Journal, 45: 841–846.

Eden, D. 2004. Reflections on the AMJ associate editorrole. Academy of Management Journal, 47: 167–173.

Eisenhardt, K. M. 1989a. Agency theory: An assessmentand review. Academy of Management Review, 14:57–74.

Eisenhardt, K. M. 1989b. Building theories from casestudy research. Academy of Management Review,14: 532–550.

Eisenhardt, K. M., & Graebner, M. E. 2007. Theory build-ing from cases: Opportunities and challenges. Acad-emy of Management Journal, 50: 25–32.

Finkelstein, S. 1992. Power in top management teams:Dimensions, measurement, and validation. Acad-emy of Management Journal, 35: 505–538.

Frese, M., Kring, W., Soose, A., & Zempel, J. 1996. Per-sonal initiative at work: Differences between eastand west Germany. Academy of Management Jour-nal, 39: 37–63.

George, G. 2005. Slack resources and the performance ofprivately held firms. Academy of ManagementJournal, 48: 661–676.

Glaser, B. G., & Strauss, A. L. 1967. The discovery ofgrounded theory. Hawthorne, NY: Aldine deGruyter.

Glick, W. H., Miller, C. C., & Cardinal, L. B. 2007. Makinga life in the field of organization science. Journal ofOrganizational Behavior, 28: 817–835.

Glynn, M. A., & Abzug, R. 2002. Institutionalizing iden-tity: Symbolic isomorphism and organizationalnames. Academy of Management Journal, 45: 267–280.

Greenwood, R., & Hinings, C. R. 1993. Understandingstrategic change: The contribution of archetypes.Academy of Management Journal, 36: 1052–1081.

Hackman, J. R., & Oldham, G. R. 1976. Motivationthrough the design of work: Test of a theory. Orga-nizational Behavior and Human Performance, 16:250–279.

Hall, C. S., & Lindzey, G. 1957. Theories of personality.New York: Wiley.

Hambrick, D. C., & Mason, P. A. 1984. Upper echelons:The organization as a reflection of its top managers.Academy of Management Review, 9: 193–206.

Hannan, M. T., & Freeman, J. 1977. The population ecol-ogy of organizations. American Journal of Sociol-ogy, 82: 929–964.

Hannan, M. T., & Freeman, J. 1989. Organizational ecol-ogy. Cambridge, MA: Harvard University Press.

Hempel, C. 1966. Philosophy of natural science. Engle-wood Cliffs, NJ: Prentice-Hall.

Hendrick, C. 1991. Replications, strict replications, andconceptual replications: Are they important? In J. W.Neuliep (Ed.), Replication research in the socialsciences: 41–49. Newbury Park, CA: Sage.

Henley, A. B., Shook, C. L., & Peterson, M. 2006. Thepresence of equivalent models in strategic manage-ment research using structural equation modeling:Assessing and addressing the problem. Organiza-tional Research Methods, 9: 516–535.


House, R. J. 1971. A path goal theory of leader effective-ness. Administrative Science Quarterly, 16: 321–328.

Hult, G. T. M., Ketchen, D. J., & Nichols, E. L. 2002. Anexamination of cultural competitiveness and orderfulfillment cycle time within supply chains. Acad-emy of Management Journal, 45: 577–586.

James, L. R. 1982. Aggregation bias in estimates of per-ceptual agreement. Journal of Applied Psychology,67: 219–229.

James, L. R., Mulaik, S. A., & Brett, J. M. 2006. A tale oftwo methods. Organizational Research Methods,9: 233–244.

Jensen, M. C. & Meckling, W. H. 1976. Theory of firm-managerial behavior, agency costs and ownershipstructure. Journal of Financial Economics, 3: 305–360.

Johnson, M. D., Hollenbeck, J. R., Humphrey, S. E., Ilgen,D. R., Jundt, D., & Meyer, C. J. 2006. Cutthroat coop-eration: Asymmetrical adaptation to changes in teamreward structures. Academy of Management Jour-nal, 49: 103–119.

Judge, T. A., Cable, D. M., Colbert, A. E., & Rynes, S. L.2007. What causes a management article to be cited:Article, author, or journal? Academy of Manage-ment Journal, 50: 491–506.

Kacmar, K. M., & Whitfield, J. M. 2000. An additionalrating method for journal articles in the field of man-agement. Organizational Research Methods, 3:392–406.

Kahn, W. A. 1990. Psychological conditions of personalengagement and disengagement at work. Academyof Management Journal, 33: 692–724.

Kahneman, D., & Tversky, A. 1979. Prospect theory: Ananalysis of decisions under risk. Econometrica, 47:313–327.

Kerlinger, F. N., & Lee, H. B. 2000. Foundations of be-havioral research. Forth Worth, TX: Harcourt.

Kirkman, B. L., & Law, K. S. 2005. International manage-ment research in AMJ: Our past, present, and future.Academy of Management Journal, 48: 377–386.

Klassen, R. D., & Whybark, D. C. 1999. The impact ofenvironmental technologies on manufacturing per-formance. Academy of Management Journal, 42:599–615.

Korman, A. K., Wittig-Berman, U., & Lang, D. 1981. Ca-reer success and personal failure: Alienation in pro-fessionals and managers. Academy of ManagementJournal, 24: 342–360.

Kuhn, T. S. 1963. The essential tension: Tradition andinnovation in scientific research. In C. W. Taylor & F.Barron (Eds.), Scientific creativity: Its recognitionand development. New York: Wiley.

Lankau, M. J., & Scandura, T. A. 2002. An investigation ofpersonal learning in mentoring relationships: Con-

tent, antecedents, and consequences. Academy ofManagement Journal, 45: 779–790.

Lau, C. M., & Woodman, R. W. 1995. Understandingorganizational-change: A schematic perspective.Academy of Management Journal, 38: 537–554.

Lawrence, P., & Lorsch, J. 1967. Differentiation and inte-gration in complex organizations. AdministrativeScience Quarterly, 12: 1–30.

Lee, T. W. 2001. On qualitative research in AMJ. Acad-emy of Management Journal, 44: 215–216.

Lee, T. W., & Mitchell, T. R. 1994. An alternative ap-proach: The unfolding model of employee turnover.Academy of Management Review, 19: 51–89.

Lee, T. W., Mitchell, T. R., Wise, L., & Fireman, S. 1996.An unfolding model of voluntary employee turn-over. Academy of Management Journal, 39: 5–36.

Locke. A., & Latham, G. P. 2004. What should we doabout motivation theory? Six recommendations forthe twenty-first century. Academy of ManagementReview, 29: 388–403.

Locke, E. A. 1968. Toward a theory of task motivationand incentives. Organizational Behavior and Hu-man Performance, 3: 157–189.

Locke, K. 2002. The grounded theory approach to quali-tative research. In F. Drasgow & N. Schmitt (Eds.),Measuring and analyzing behavior in organiza-tions: 17–43. San Francisco: Jossey-Bass.

Lodahl, J. B., & Gordon, G. 1972. The structure of scien-tific fields and the functioning of university graduatedepartments. American Sociological Review, 37:57–72.

Lykken, D. T. 1968. Statistical significance in psycholog-ical research. Psychological Bulletin, 70: 151–159.

Martinson, O. B., & Wilkening, E. A. 1984. Rural-urbandifferences in job satisfaction: Further evidence.Academy of Management Journal, 27: 199–206.

McAllister, D. J. 1995. Affect-based and cognition-basedtrust as foundations for interpersonal cooperation inorganizations. Academy of Management Journal,38: 24–59.

McKinley, W., Mone, M. A., & Moon, G. 1999. Determi-nants and development of schools in organizationtheory. Academy of Management Review, 24: 634–648.

Miner, J. B. 1984. The validity and usefulness of theoriesin an emerging organizational science. Academy ofManagement Review, 9: 296–306.

Miner, J. B. 2003. The rated importance, scientific valid-ity, and practical usefulness of organizational behav-ior theories: A quantitative review. Academy ofManagement Learning and Education, 2: 250–268.

Mitchell, T. R., Holtom, B. C., Lee, T. W., Sablynski, C. J.,& Erez, M. 2001. Why people stay: Using job embed-dedness to predict voluntary turnover. Academy ofManagement Journal, 44: 1102–1121.


Mohr, J. 1982. Explaining organizational behavior. SanFrancisco: Jossey-Bass.

Morrison, E. W., & Phelps, C. C. 1999. Taking charge atwork: Extrarole efforts to initiate workplace change.Academy of Management Journal, 42: 403–419.

Mowday, R. T. 1997. Celebrating 40 years of the Acad-emy of Management Journal. Academy of Manage-ment Journal, 40: 1400–1413.

Nohria, N., & Gulati, R. 1996. Is slack good or bad forinnovation? Academy of Management Journal, 39:1245–1264.

Northcraft, G. B. 2000. From the editors. Academy ofManagement Journal, 43: 133–134.

Oldham, G. R. 1975. Impact of supervisory characteris-tics on goal acceptance. Academy of ManagementJournal, 18: 461–475.

Pfeffer, J. 1993. Barriers to the advance of organizationalscience: Paradigm development as a dependent vari-able. Academy of Management Review, 18: 599–620.

Pfeffer, J., & Salancik, G. R. 1978. The external control oforganizations: A resource dependence perspec-tive. New York: Harper & Row.

Pierce, J. L., Gardner, D. G., Cummings, L. L., & Dunham,R. B. 1989. Organization-based self-esteem: Con-struct definition, measurement, and validation.Academy of Management Journal, 32: 622–648.

Platt, J. R. 1964. Strong inference. Science, 146: 347–353.

Podsakoff, P. M., MacKenzie, S. B., Bachrach, D. G., &Podsakoff, N. P. 2005. The influence of managementjournals in the 1980s and 1990s. Strategic Manage-ment Journal, 26: 473–488.

Popper, K. R. 1965. Conjectures and refutations: Thegrowth of scientific knowledge. New York: Harper &Row.

Rindova, V. P., Williamson, I. O., Petkova, A. P., & Sever,J. M. 2005. Being good or being known: An empiricalexamination of the dimensions, antecedents, andconsequences of organizational reputation. Acad-emy of Management Journal, 48: 1033–1049.

Robinson, S. L., & Bennett, R. J. 1995. A typology ofdeviant workplace behaviors: A multidimensional-scaling study. Academy of Management Journal,38: 555–572.

Rosenthal, R. 1991. Replications in behavioral re-search. In J. W. Neuliep (Eds.), Replication re-search in the social sciences: 1–30. NewburyPark, CA: Sage.

Rousseau, D. M. 2007. Standing out in the fields of or-ganization science. Journal of Organizational Be-havior, 28: 849–857.

Rynes, S. L. 2002. From the editors. Academy of Man-agement Journal, 45: 311–313.

Rynes, S. L. 2005. Taking stock and looking ahead. Acad-emy of Management Journal, 48: 732–737.

Salancik, G. R., & Pfeffer, J. 1978. A social informationprocessing approach to job attitudes and taskdesign. Administrative Science Quarterly, 23: 224–253.

Schein, E. H. 1985. Organizational culture and leader-ship. San Francisco: Jossey-Bass.

Schminke, M. 2004. Raising the bamboo curtain. Acad-emy of Management Journal, 47: 310–314.

Schminke, M., & Mitchell, M. 2003. In the begin-ning . . . Academy of Management Journal, 46:279 –282.

Schneider, B., Ehrhart, M. G., Mayer, D. M., Saltz, J. L., &Niles-Jolly, K. 2005. Understanding organiza-tion-customer links in service settings. Academy ofManagement Journal, 48: 1017–1032.

Schwab, D. P. 2005. Research methods for organiza-tional studies. Mahwah, NJ: Erlbaum.

Shrout, P. E., & Fleiss, F. L. 1979. Intraclass correlations:Uses in assessing rater reliability. PsychologicalBulletin, 86: 420–428.

Skarlicki, D. P., Folger, R., & Tesluk, P. 1999. Personalityas a moderator in the relationship between fairnessand retaliation. Academy of Management Journal,42: 100–108.

Smith, P., & Kendall, L. 1963. Retranslation of expecta-tions: An approach to the construction of unambig-uous anchors for rating scales. Journal of AppliedPsychology, 47: 149–155.

Spell, C. S. 2001. Management fashions: Where do theycome from, and are they old wine in new bottles.Journal of Management Inquiry, 10: 358–373.

Staw, B. M., & Boettger, R. D. 1990. Task revision: Aneglected form of work performance. Academy ofManagement Journal, 33: 534–559.

Stroh, L. K., Brett, J. M., Baumann, J. P., & Reilly, A. H. 1996.Agency theory and variable pay compensation strate-gies. Academy of Management Journal, 39: 751–767.

Suddaby, R. 2006. From the editors: What grounded theoryis not. Academy of Management Journal, 49: 633–642.

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

Tahai, A., & Meyer, M. 1999. A revealed preference studyof management journals’ direct influences. StrategicManagement Journal, 20: 279–296.

Tajfel, H., & Turner, J. C. 1979. An integrative theory ofintergroup conflict. In W. G. Austin & S. Worchel(Eds.), The social psychology of intergroup rela-tions: 33–47. Monterey, CA: Brooks/Cole.

Tierney, P., & Farmer, S. M. 2002. Creative self-efficacy:Its potential antecedents and relationship to creative


performance. Academy of Management Journal,45: 1137–1148.

Tsang, E. W. K., & Kwan, K. M. 1999. Replication andtheory development in organizational science: Acritical realist perspective. Academy of Manage-ment Review, 24: 759–780.

Tsui, A. S. 1999. From the editor. Academy of Manage-ment Journal, 42: 237–238.

Tsui, A. S., & O’Reilly, C. A. 1989. Beyond simple demo-graphic effects: The importance of relational demog-raphy in superior-subordinate dyads. Academy ofManagement Journal, 32: 402–423.

Vance, S. C. 1967. Editorial comment. Academy of Man-agement Journal, 10: 7–7.

Van der Vegt, G. S., Bunderson, J. S., & Oosterhoff, A. 2006.Expertness diversity and interpersonal helping inteams: Why those who need the most help end upgetting the least. Academy of Management Journal,49: 877–893.

Vroom, V. H. 1964. Work and motivation. New York:Wiley.

Wade-Benzoni, K. A. 2002. A golden rule over time: Reci-procity in intergenerational allocation decisions.Academy of Management Journal, 45: 1011–1028.

Washington, M., & Zajac, E. J. 2005. Status evolution andcompetition: Theory and evidence. Academy of Man-agement Journal, 48: 282–296.

Weick, K. E. 1995. What theory is not, theorizing is.Administrative Science Quarterly, 40: 385–390.

Weick, K. E. 2007. The experience of theorizing: Sense-making as topic and resource. In K. G. Smith & M. A.Hitt (Eds.), Great minds in management: The pro-cess of theory development: 394–406. Oxford, U.K.:Oxford University Press.

Whetten, D. A. 1989. What constitutes a theoretical con-tribution? Academy of Management Review, 14:490–495.

Whiteman, G., & Cooper, W. H. 2000. Ecological embed-dedness. Academy of Management Journal, 43:1265–1282.

Williamson, O. E. 1975. Markets and hierarchies. NewYork: Free Press.

Wiseman, R. M., & Skilton, P. F. 1999. Divisions anddifferences: Exploring publication preferences andproductivity across management subfields. Journalof Management Inquiry, 8: 299–320.

Wu, S., Levitas, E., Priem, R. L. 2005. CEO tenure andcompany invention under differing levels of techno-logical dynamism. Academy of Management Jour-nal, 48: 859–873.

Zedeck, S. 2003. Editorial. Journal of Applied Psychol-ogy, 88: 3–5.

APPENDIXThe Evolution of Expectations about Theoretical Contribution in AMJ’s “Information for Contributors”

Editor Crafting theStatement Years in Effect Relevant Quotations about Theory Expectations

Paul M. Dauten 1958–69 The Academy is founded to foster the search for truth and the general advancement oflearning through free discussion and research in the field of management. The interest ofthe Academy lies in the theory and practice of management, both administrative andoperative.

William G. Scott 1970–72 The interest of the Academy of Management lies in management theory, research, teaching,and practice. To foster these interests, the Journal’s objectives are: (1) the developmentof management research and theory that will help achieve the economic and socialobjectives of industrial societies; (2) the advancement of understanding aboutadministrative leadership and behavior through research within the environment of suchsocieties; (3) the enlargement of scholarly communication and cooperation amongcolleagues engaged in management research and theory.

John B. Miner 1973–84 The Journal publishes original research of an empirical nature either in the form of articlesor as research notes. Although studies which serve to test either theoretical propositionsor hypotheses derived from practice are of particular interest, exploratory work andsurvey research findings are also included. . . . For consideration in the Research Notescategory, articles should not exceed ten double spaced typewritten pages in length,including tables. Replications, survey reports, and studies which fail to obtain significantresults that might have been expected on other grounds are especially appropriate.

Janice M. Beyer 1985–96 In its articles, the Journal seeks to publish reports of research that develops, tests, oradvances management theory and practice. All types of empirical methods-quantitative,qualitative, or combinations-are acceptable. . . . Atheoretical exploratory or surveyresearch, methodological studies, replications or extensions of past research, andcommentaries with new empirical content are also of interest for publication as researchnotes if they make an important contribution to knowledge relevant to management.


Jason A. Colquitt ([email protected]) is a professor at theUniversity of Florida’s Warrington College of Business.He received his Ph.D. from Michigan State University’sEli Broad Graduate School of Management. His researchinterests include organizational justice, trust, and per-sonality influences on task and learning performance.

Cindy P. Zapata-Phelan ([email protected])is a doctoral candidate at the University of Florida’sWarrington College of Business. She earned her B.S.in psychology at the University of Florida. Her re-search interests include personality, motivation, andorganizational justice.

APPENDIX(Continued)

Editor Crafting theStatement Years In Effect Exemplary Quotations about Theory Expectations

Anne S. Tsui andGregory B.Northcraft

1999–2004 All articles published in the Journal must make a strong theoretical contribution.Submissions should reflect a clear understanding of the position of the contribution inthe related organization and management literatures. Methodological articles arewelcome to the extent that they contain an accompanying theoretical contribution. Allarticles published in the Journal must be clearly relevant to management theory andpractice. The best submissions are those that identify both a compelling practicalmanagement issue and a strong theoretical framework for addressing it. . . . Manuscriptsthat offer an original theoretical and empirical contribution, but one that is small inscope, may be published as research notes. . . . Replications of previously publishedwork and very incremental research rarely offer enough of a contribution to warrantpublication. Authors should strive to be original, insightful, and theoretically bold;demonstration of a significant “value-added” advance to the field’s understanding of anissue or topic is critical to acceptance for publication.

Sara L. Rynes 2005–07 The mission of the Academy of Management Journal is to publish empirical research thattests, extends, or builds management theory and contributes to management practice. Allempirical methods—including, but not limited to, qualitative, quantitative, field,laboratory, meta-analytic, and combination methods—are welcome. To be published inAMJ, a manuscript must make strong empirical and theoretical contributions andhighlight the significance of those contributions to the management field. Thus,preference is given to submissions that test, extend, or build strong theoreticalframeworks while empirically examining issues with high importance for managementtheory and practice. . . . Authors should strive to produce original, insightful, interesting,important, and theoretically bold research. Demonstration of a significant “value-added”contribution to the field’s understanding of an issue or topic is crucial to acceptance forpublication. . . . All articles published in the Academy of Management Journal must alsomake strong theoretical contributions. Meaningful new implications or insights fortheory must be present in all AMJ articles, although such insights may be developed in avariety of ways (e.g., falsification of conventional understanding, theory building throughinductive or qualitative research, first empirical testing of a theory, meta-analysis withtheoretical implications, constructive replication that clarifies the boundaries or range ofa theory). Submissions should clearly communication the nature of their theoreticalcontribution in relation to the existing management and organizational literatures.Methodological articles are welcome, but they must contain accompanying theoreticaland empirical contributions.


Documents

TRENDS IN THEORY BUILDING AND THEORY TESTING: A FIVE