«by Andrew Schnackenberg WP-10-09 Copyright Department of Organizational Behavior Weatherhead School of Management Case Western Reserve University ...»
THE CONSTITUTIVE ROLE OF TRANSPARENCY IN
Department of Organizational Behavior
Weatherhead School of Management
Case Western Reserve University
Cleveland OH 44106-7235
THE CONSTITUTIVE ROLE OF TRANSPARENCY IN ORGANIZATIONS
ANDREW SCHNACKENBERGPhD Student Department of Organizational Behavior Weatherhead School of Management Case Western Reserve University Qualifying Paper Abstract Transparency is an increasingly important, yet poorly understood, construct. Drawing on relevant literatures, this study builds a comprehensive definition of transparency and empirically examines its relationship with trust, organizational buy-in, and information usefulness. A quasi- experimental analysis reveals transparency significantly predicts these constructs. Implications for managers and researchers are discussed.
Andrew Schnackenberg │1 In recent years, a great deal of interest has coalesced around the topic of transparency (Patel, Balic, & Bwakira, 2002). Mainstream periodicals cite transparency as essential to successful mergers and joint ventures (“Shades of grey,” 2009; “Changing course,” 2009), well- functioning financial markets (Ackermann, 2008), effective national governance (Theil, 2010), and the reestablishment of organizational trust (Fifield, 2010). Empirical studies, however, show mixed results as to the effects of transparency on institutional outcomes (Madhavan, Porter, & Weaver, 2005). Some scholars argue that transparency is an essential component to effective organizing (Fleischmann & Wallace, 2005) while others argue it is an “overrated” concept (Pirson & Malhotra, 2008: p.44).
Transparency is a complicated subject (Madhavan, 2000) and there is a general lack of knowledge about its effects on institutional behavior (Bloomfield & O’Hara, 1999). A fundamental problem in understanding how transparency relates to institutional outcomes stems from a lack of consensus about the basic definition of transparency (Bloomfield & O’Hara, 1999). The primary objective of this study is to develop a comprehensive and unified definition for transparency as the level of perceived understandability, completeness, and correctness in institutional communications.
Specifically, the purposes of this paper are threefold. First, I develop a comprehensive conceptual model for the construct of transparency. Second, I build the case for a multidimensional theory-based scale for transparency and offerevidence of its construct validity.
To accomplish this, I elaborate on the theoretical dimensions of transparency and describe the item development and validation processes performed to assess this derived structure.
Third, I use Structural Equation Modeling (SEM) to show the predictive power of
in, and information usefulness. This analysis reveals the extent to which higher levels of transparency contribute to perceptions of organizational trust. It further reveals that two dimensions of transparency (viz., understandability and completeness) uniquely account for company buy-in and information usefulness, controlling for perceptions of organizational trust.
This paper has implications for researchers interested in understanding the relationship between transparency and trust. Further, it offers thought provoking implications for practitioners looking to manage transparency in institutional environments.
The Latin etymology of the word transparency is bipartite, consisting of trāns – meaning “across” or “through” – and pāreō – meaning “be seen”. In the physical sciences, the MerriamWebster Dictionary defines a transparent object as having the property of transmitting light without appreciable scattering so that bodies lying beyond are seen clearly. Social scientists have metaphorically adopted this definition to connote the ability of interested parties to see through otherwise private information to understand the intentions of the sender.
The field of finance has done much to study the effects of transparency across a broad range of industrial settings. Such studies generally investigate transparency as the inherent quality of information in institutional communications. However, researchers in finance have not yet come to consensus on a single theoretical definition to describe the basic dimensions of transparency (Bloomfield & O’Hara, 1999). Be that as it may, scholars from a variety of research domains have noted the importance of understanding the critical role of institutional communications (e.g., employment contracts) on organizational behavior (e.g., Ashcraft, Kuhn,
comprehensive definition of transparency based on literature in finance and empirically investigates its constituent parts against important organizational constructs such as trust.
Empirical studies investigating transparency in finance have analyzed it as both the dependent variable (Hodge, Kennedy, & Maines, 2004; Patel et al., 2002) and independent variable (Board & Sutcliffe, 2000; Gemmill, 1996; Rosengren, 1999; Winkler, 2000) under question. Some studies define transparency as the timely disclosure of information (e.g., Bloomfield & O’Hara, 1999; Madhavan, Porter, & Weaver, 2005; Pagano & Roell, 1996;
Securities and Exchange Commission, 1995; Securities and Investment Board, 1995) while other studies define transparency as the level of clarity in information (Bushman, Piotroski, & Smith, 2004; Jordan, Peek, & Rosengren, 2000; Winkler, 2000). Still others studies define transparency as the level of accuracy in information (Flood, Huisman, Koedijk, & Mahieu, 1999; Granados, Gupta, & Kauffman, 2006). Overall, the literature in finance appears to ascribe three primary dimensions to the construct of transparency. Namely, transparency is the degree to which information is disclosed, clear, and accurate.
Van Dijk, Duysters, and Beulens (2003) argue that transparency is dynamically constructed between institutions and individuals through exchanges of information construction and interpretation. In essence, transparency is seen as both enacted and perceived in institutional communications. The literature in finance is primarily focused on understanding transparency as an enacted phenomenon (i.e., as the sending of accurate, disclosed, and clear information). Yet Van Dijk, Duysters, & Beulens (2003) argue that the level of transparency in systems is ultimately determined by its perception rather than its enactment. To date, no instrument has
initial steps towards development of a comprehensive psychometric instrument to measure perceptions of transparency.
To accomplish this, each enacted dimension of transparency identified in the literature is reframed to account for its perception. Specifically, transparency is defined as the level of perceived completeness (i.e., disclosure), understandability (i.e., clarity), and correctness (i.e., accuracy) in messages, documents, or other institutional communications (Table 1). Within this framework, completeness refers to the perceived quantity of information in messages or other communications as well as the availability of that information to interested parties, correctness is defined as the degree to which material claims are made truthfully and reflect truthful qualifications about their perceived validity, and understandability is defined as the extent to which representations are designed in ways that are understandable to focal audiences (Figure 1).
Recent literature has suggested that transparency and trust are intricately related concepts (Pirson & Malhotra, 2008). However, the relationship between trust and transparency has yet to be empirically examined. Hence, this paper reports the results of a quasi experiment conducted to investigate the relationship between perceptions of transparency and organizational trust.
The Dimensional Structure of Transparency: An Empirical Analysis The conceptual model of transparency outlined in Figure 1 can be understood as a multidimensional model consisting of three formative first-order factors (viz., accuracy, disclosure, and clarity), three reflective first-order factors (viz., perceived correctness, completeness, and understandability), and one second-order factor (viz., transparency) (Law,
development and validation of an instrument to measure the three reflective first-order factors of transparency by controlling the three formative first-order factors of transparency.
Item Development and Validation The following outlines the item development process used to construct the measure of transparency. I then outline the results of an Exploratory Factor Analysis (EFA) conducted to investigate the factor structure of transparency. Next, I offer the results of a Confirmatory Factor Analysis (CFA) carried out to test the higher-order factor structure of transparency. Results investigating the convergent and discriminant validity of transparency are then offered. Finally, factorial equivalence was tested to see if each dimension of transparency was group invariant.
To build a measure for transparency, I used both deductive and inductive approaches for item generation to assess how transparency was perceived in communications (Hinkin, 1995).
Initial content specifications were developed based on an extensive review of the literature in finance (e.g., Flood et al., 1999; Flood, Huisman, Koedijk, Mahieu, & Roell, 1997; Fons, 1999;
Geraats, 2002; Fry, Julius, Mahadeva, Roger, & Sterne, 2000; Jappelli & Pagano, 2000; Jordan, Peek, & Rosengren, 2000; Khanna, Palepu, & Srinivasan, 2004; Morris & Shin, 1997; O’Hara, 1999; Pagano & Roell, 1996; Rosengren, 1999; Tanzi, 1998; Winkler, 2000), and semistructured interviews with academic specialists in the areas of finance, organizational behavior, and marketing. Based on this comprehensive review, three initial domains were deemed appropriate to constitute the transparency measure (viz., accuracy, disclosure, and clarity).
The three domains associated with perceived completeness, correctness, and understandability were developed and an item pool generated based on an analysis of the
1998; Butler, 1991; Das & Teng, 2001; Duarte & Snyder, 1999; Ferrin, Bligh, & Kohles, 2008;
Gabarro, 1978; Jones & George, 1998; Lewicki, McAllister, & Bies, 1998; Mayer, Davis, & Schoorman, 1995; McAllister, 1995; Nelson & Cooprider, 1996; Rempel, Holmes, & Zanna, 1985; Ring, 1996; Rousseau, Sitkin, Burt, & Camerer, 1998; Sheppard & Sherman, 1998;
Sheppard & Tuckinsky, 1996; Sitkin & Roth, 1993), source credibility and faking (e.g., Ba & Pavlou, 2002; Cunningham, Wong, & Barbee, 1994; Doney & Cannon, 1997; Ganesan 1994;
Giffin, 1967; Hovland, & Weiss, 1951; Komar, Brown, & Robie, 2008; McFarland & Ryan, 2000, 2006), organizational justice (e.g., Bies & Shapiro, 1987; Colquitt, 2001; Greenberg, 1987, 1990; Leventhal, Karuza, & Fry, 1980; Mikula, Petrik, & Tanzer, 1990; Moore, 1978; Roch, & Shanock, 2006; Thibaut & Walker, 1975), and framing (e.g., Benford & Snow, 2000; Druckman, 2001; Fiss & Zajac, 2006; Kahneman & Tversky, 1984; Kaplan, 2008; Kuhberger, 1998; Levin, Schneider, & Gaeth, 1998; Tversky & Kahneman, 1981). Semi-structured interviews were then completed with industry professionals to refine the three domains of transparency. Based on these initial results, feedback from doctoral students and faculty, and discussions with other researchers at conferences, no new dimensions were added. From this review I theoretically derived 31 items which were later refined to 21 items that best captured the proposed content areas. These 21 items were then subjected to a content validity assessment by an expert panel of academic specialists from the related research areas listed above. 53 academic specialists were emailed the items and asked to sort them into their respective categories. A final category of “none of the above” was included for good measure. Respondents were then asked to remark on the item structure and content appropriateness of each question.
21 of the 53 specialists responded to the request. Based on the recommendations of the
category more than 80% of the time, indicating they should be retained for further analysis (MacKenzie, Podsakoff, & Fetter, 1991). Fleiss’ kappa was used as a measure of agreement between raters for the remaining 18 items. Fleiss' kappa is a statistical measure for assessing the reliability of agreement between multiple raters when classifying items into multiple categories (Fleiss, 1971). According to Landis & Koch (1977), kappa values above.6 are indicative of substantial agreement. Fleiss’ kappa was.62, indicating a sufficient level of agreement between raters. On the basis of the ratings, as well as comments regarding the appropriateness of the items for the specified categories, I revised some of the items accordingly and then tested them in a pilot study using PhD students and faculty from the institution where the measure was being developed. The items that were retained for further analysis were as follows: perceived correctness (6 items), perceived completeness (6 items), and perceived understandability (6 items).
EXPLORATORY FACTOR ANALYSIS AND CONFIRMATORY FACTOR ANALYSISParticipants and Procedures This study investigated the properties of transparency using a within-subjects quasiexperimental design. Participants were brought to a single location and instructed to work independently to analyze information from five different organizations. The information on each organization consisted of a company prospectus covering various aspects of the firm’s operations and a separate Audited Statement of Performance (ASP) used to verify the information in the company prospectus. Each prospectus included the following sections: company overview,
were asked to evaluate each organization based on the following criteria: financial performance, recent industry growth, management experience, future outlook, and employee morale.