شبکه های پژوهش حسابداری: تجزیه و تحلیل شبکه ای و ساختاری مبتنی بر استناد

This study measures journal influence using the principles of knowledge capital and social networks. The structural index measure is used to evaluate knowledge capital flows in a network of 22 accounting research journals over the years 2000–2006. The influence measure is a function of both the quantity and quality of citations resulting in journal influence rankings that reflect each journal's contribution to the broad context of accounting research. Findings show that although some journals are highly cited, they may be less influential in the accounting research community. Publishing research of high value is one key to increasing the influence of a journal in the network. Furthermore, the analysis of a well-defined journal network provides a view of the distinct contribution of individual journals. The network analysis diagram provides a visual perspective of journal relationships, emphasizes the strength of relational ties and suggests that influential journals may take on different roles. The contribution of other disciplines to the top-five influential accounting research journals is also presented.

Journals form the basis of a knowledge network that communicates scholarly research and advances a discipline. The creation of the accounting research network began with The Accounting Review as the leading North American journal from 1926 until 1963 when the Journal of Accounting Research debuted. In the U.K., Accounting Research (1948–1958) was a major research outlet later joined by three additional non-U.S. research journals—Abacus, Journal of Business Finance and Accounting, and Accounting and Business Research (see Appendix A). Seven more accounting research journals were established in the 1970s, nine in the 1980s, and two in the 1990s. In addition to research journals, numerous other accounting publications have appeared. Cabell lists more than 130 accounting-related publishing opportunities (see Cabell, 2001–2002) including news periodicals and applied professional publications with at least 77 classified as research-oriented (Zeff, 1996). Within the larger collection of accounting-related and research-oriented publications, a smaller network of scientific journals serves as the major forum for accounting research (Campbell et al., 1983 and Brown and Huefner, 1994). The objective of these journals is to disseminate original scholarly research or “discovery” research that utilizes a scientific basis. Accounting academics are often evaluated on their research productivity and the publication of discovery research is an important measure of scholarly performance. Because research productivity may impact promotion and tenure (Kirkpatrick and Locke, 1992), compensation (Gomez-Mejia and Balkin, 1992) and departmental or business school rankings (Saftner, 1988, Schultz et al., 1989, Hasselback and Reinstein, 1995, Read et al., 1998 and Everett et al., 2003), the publication outlet (i.e., journal) is a major focal point. University departments may specify a list of “quality” journals in which faculty/researchers are encouraged to publish (Richardson and Williams, 1990; Whittington, 1997, Van Fleet et al., 2000, Mylonopoulos and Theoharakis, 2001 and Mathieu and McConomy, 2003). Consequently, the factors comprising journal quality continue to be a source of scrutiny and debate among faculty, researchers, and university administrators. Numerous accounting studies attempt to measure journal quality using peer-ranking (Benjamin and Brenner, 1974, Howard and Nikolai, 1983, Hull and Wright, 1990, Brown and Huefner, 1994, Ballas and Theoharakis, 2003, Lowe and Locke, 2005 and Lowe and Locke, 2006) and citation methodologies (McRae, 1974, Dyckman and Zeff, 1984, Brown and Gardner, 1985, Smith and Krogstad, 1988 and Milne, 2001). Although self-serving and predisposition biases may be inherent in perception rankings (Morris et al., 1990 and Todorov and Glanzel, 1988), and scope limitations (e.g., Dyckman and Zeff, 1984 and Brown and Gardner, 1985) with issues of negative referencing (Croom, 1970) and halo effects (May, 1967) often plague the results of citation analyses, survey and citation analysis remain legitimate methods for journal evaluation. In fact, recent accounting journal studies use the perception approach to identify quality journals using a global survey of academics (Ballas and Theoharakis, 2003) and a sample of Australasian and British academics (Lowe and Locke, 2006). However, citation analyses have not been used to support the current rankings and increase the confidence placed in explicitly naming the ‘top’ accounting journals. When the two methodologies provide similar evidence despite their limitations, this ‘triangulation’ adds to our confidence in identifying influential journals across a broad and varied mix of accounting research. In contrast to past citation analyses in accounting, the present study calculates the relative influence of accounting journals in a specific research journal network using structural influence indexing (Salancik, 1986) as the methodological approach. The influence measure reflects the extent to which a journal is utilized within a specified network for the knowledge capital it contains. An additional contribution of this study is in presenting how influence indexing may be used to evaluate the unique contribution of some accounting journals. For example, a journal of limited scope (e.g., taxation) that tends to publish behavior-based research may hold greater influence among a network of journals with similar methodological preferences, and lesser influence within a broad journal network. On the one hand, the contribution of the journal is highlighted and on the other it is relatively obscure. Furthermore, when the influence index is applied to a network of journals that span disciplines (e.g., accounting, social science, economics) patterns of knowledge flows are revealed as well as the extent to which accounting research influences or is influenced by other disciplines. The underlying premise in applying the indexing method to the research journal context is that researchers identify and build upon the best research and a journal is influential to the degree it contributes to current and subsequent research. In essence, this study evaluates the perspective of the primary consumers of accounting research, active researchers. Through examining the network of bibliographic citations, the intellectual linkages among authors, journals, and fields are revealed. The structural influence metric traces the flows of researchers’ intellectual assets and provides an estimate of journal influence from the researchers’ viewpoint. Influence indexing has been successfully applied to a network of marketing journals (Baumgartner and Pieters, 2003). The unique contribution of this method includes a weighting scheme that evaluates both direct and indirect linkages among journals with the elimination of self-citations in the analysis. Consequently, journal influence is a function of both the quantity and quality of cites received, and the elimination of self-cites in the analysis may reduce the extent of journal bias.1 Results in this study show that the number of citations a journal receives from other network journals or the number of articles a journal publishes may not be indicative of higher journal status. Journals publishing fewer manuscripts of greater quality (i.e., contribution) may be more influential in the network than journals publishing many articles. Furthermore, specialized accounting journals such as the Journal of Management Accounting Research and the Journal of the American Taxation Association also demonstrate influence in the broad accounting research community despite a narrow research focus. Additional network analyses using UCINET VI software provides a visual depiction of relational ties among North American and Non-North American accounting research journals and shows the extent of knowledge sharing within the geographically dispersed accounting research community.

Limitations to the research include the biases discussed (i.e., halo effect, editorial emphases, negative referencing) as well as the period of data collection. However, the elimination of self-cites in the analysis curtails the full impact of the halo effect and may also reduce the effect of perfunctory citing of the journal in which the author is publishing. It is likely that the removal of all self-cites from the analyses also eliminates some legitimate citations. Nonetheless, the methodology clearly depicts where researchers publishing in other network journals derive knowledge capital. Although the possibility exists that ‘hot’ topics may alter a journal's relative influence over short time periods, the seven-year collection period in this study is likely to control for some fluctuation. The data collection period (2000–2006) limits interpretation of the findings apart from this time frame, but also represents an opportunity to extend the research. The influence indexing method provides a means to examine any number of journal networks to identify variations of influence that may occur with changes in editorial staff, journal mission, or the passage of time. Time period analyses of selected networks may reveal the evolutionary changes of journals in the network as research streams evolve and new methodologies emerge. Structural indexing has the potential to evaluate the contribution of specialized journals that may remain relatively obscure in peer-reviewed lists and rankings. Additionally, networks may be constructed that include influential journals from other disciplines to empirically assess the extent to which other fields contribute to accounting research as well as the degree to which accounting research influences other disciplines. Measuring research productivity is an important part of assessing academic scholarship. However, measuring the quality aspect of productivity is difficult as no one metric indisputably captures all aspects of journal status or research quality. This study takes the user (or researcher) perspective to judge accounting journal influence using a network methodology that attempts to control for biases. The calculation of influence shows that journals may increase their ranking within a network not necessarily by publishing more manuscripts, but by publishing papers utilized by researchers publishing in higher ranking journals. While more articles do increase the knowledge pool which benefits the publishing journal when researchers cite that knowledge, quality or high value research is a significant driver of journal influence. Furthermore, the value of research or the ‘quality’ of a journal may be better determined by evaluating networks of journals with specific commonalities rather than one network with a broad mix of accounting research. A comparison of this study's influence index results with those of a recent peer-ranking study (Ballas and Theoharakis, 2003) shows that the top-five journals remain the same with a slight variation in positions. However, after the 7th position, the journal rankings between the two studies are quite disparate leaving questions as to the veracity of survey-based findings and the contribution of individual journals. The objectivity provided by citations allows the influence index to impartially attribute value to a journal and the research it publishes beyond the limitations of individual perception. When the index is used to evaluate distinctive networks of journals, unique contributions of individual publications may be more readily apparent and not as easily dismissed. This paper also explores network links among accounting journals and relationships with other discipline journals they reference. It is anticipated that alternative measures of journal contribution will help create an enlarged view of accounting scholarship that considers trends and diversity in the discipline. The general tendency is for maturing markets such as accounting to become more differentiated and specialized (Zeff, 1996). This indicates the need to assess the value of accounting research using diverse but meaningful methods. As specialization continues, structural influence measures and network analysis can provide insights into changes in knowledge flows and identify emerging areas of knowledge capital. The belief that only a few journals in a field are indicative of academic excellence in an evolving and maturing discipline may be short-sighted. Research productivity measures may unduly restrict the advancement of the discipline by promoting sub-optimal criterion, and the extent to which capable researchers curtail diverse research agendas because of narrow benchmarks shackles the creative energy that is the basis of progress. Exploring growing diversity in the field using valid and meaningful measures of scholarly contribution will benefit both accounting research and the accounting discipline.