The mum effect results when employees in an organization

Management Communication Quarterly 27(2) 155 –183

© The Author(s) 2013 Reprints and permission:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0893318912469770

mcq.sagepub.com

469770MCQ27210.1177/0893318912469770Mana gement Communication QuarterlyPloeger and Bisel © The Author(s) 2013

Reprints and permission: sagepub.com/journalsPermissions.nav

1University of Arkansas at Little Rock, Little Rock, AR, USA 2University of Oklahoma, Norman, OK, USA

Corresponding Author: Nicole A. Ploeger, Department of Speech Communication, University of Arkansas at Little Rock, Speech Communication Building 230, Little Rock, AR 72204, USA. Email: naploeger@ualr.edu

The Role of Identification in Giving Sense to Unethical Organizational Behavior: Defending the Organization

Nicole A. Ploeger1 and Ryan S. Bisel2

Abstract

This language production experiment investigates communication’s role in defending, and therefore giving sense to, organizational wrongdoing. The study suggests identification may possibly reduce organizations’ moral learn- ing capacity by encouraging highly identified members to engage in ethical sensegiving of their organizations’ wrongdoing in defensive ways. Work- ing adults (N = 318) responded to an organizational outsider regarding a gender discrimination lawsuit filed against their organization in one of two scenarios, which presented the organization’s guilt as either ambiguous or certain. Highly identified members used more linguistic defense mechanisms and reported more intense feelings. Additionally, participants in the ambigu- ous condition used more linguistic defense mechanisms than those in the certain condition. Veteran members reported higher levels of organizational identification and used more linguistic defense mechanisms than newcomers.

Keywords

organizational communication, organizational identification, organizational ethics, sensegiving, sensemaking

Article

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
156 Management Communication Quarterly 27(2)

Unethical organizational behaviors persist, despite a growing awareness of organizational ethics. Repercussions of such behaviors affect members, orga- nizations, stakeholders, and societies. Consider the recent gender discrimina- tion lawsuit: Dukes v. Wal-Mart Stores, Inc. (10-277). The lawsuit was filed on behalf of a half million female Wal-Mart employees, with “billions of dollars at stake” (Associated Press, 2011, March 29). As noted by Miceli, Near, and Dworkin (2008), “If wrongdoing was rare or inexpensive, we would not be concerned with it” (pp. 18-19). Clearly, unethical organizational behavior occurs with devastating consequences for many (e.g., financial and reputational loss, injury, and death).

The purpose of this investigation is to understand the communicative resources that form the substance of organizational members’ sensegiving. Rather than seeking to define what constitutes unethical behavior or question whether unethical organizational behavior occurs (for excellent reviews of organizational ethics research in communication see May, 2006; Meisenbach, 2006), this study reports a communicative investigation of organizational wrongdoing. This article contributes to the sensegiving literature (e.g., Gioia & Chittipeddi, 1991) the notion of ethical sensegiving in organizational con- texts. Here, ethical sensegiving refers to communicative attempts to influ- ence how others assign moral meaning to events or actions. Specifically, this study catalogues a range of maladaptive identity defense mechanisms that can be used by organizational members as discursive strategies to protect or maintain favorable individual and collective identities in light of accusations against one’s organization. The following paragraphs describe (a) how orga- nizational identification may manifest itself communicatively as ego defen- siveness, (b) how situational ambiguity may manifest itself as an interpretive resource for giving sense to collective wrongdoing in ways that foster ego defensiveness, and (c) how organizational tenure may encourage feelings of oneness with and a willingness to defend one’s organization.

Framing Organizational Identification Organizational identification (OI), or “the degree to which a member defines him- or herself by the same attributes that he or she believes define the orga- nization” (Dutton, Dukerich, & Harquail, 1994, p. 293), is a popular concept in communication and management (e.g., Ashforth & Mael, 1989; Barker & Tompkins, 1994; Chaput, Brummans, & Cooren, 2011; Cheney, 1983; Elsbach, 1999; Pratt, 2000; Stephens & Dailey, 2012). Much OI research is rooted in social identity theory (SIT; Haslam, 2001; Tajfel & Turner, 1985). Fundamental to many permutations of SIT is the notion that two poles moti- vate social action: (a) idiosyncratic characteristics; and (b) social classification

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 157

characteristics, in which actions are determined by group memberships and not affected by interpersonal factors. Individuals can identify with many social categories and experience multiple, overlapping, and even conflicting identities. An individual’s social identity was first described by Tajfel (1978) as “that part of an individual’s self concept which derives from his [sic] knowledge of his membership of a social group (or groups) together with the value and emotional significance attached to that membership” (p. 68). Social identification, then, involves group level classifications such as orga- nizational affiliation. Organizational identification is a specific form of social classification (Cheney, 1983). Tajfel and Turner argue that individuals attempt to maintain a positive self-concept and uphold a positive social iden- tity and strive to belong to groups that are viewed favorably when compared to other groups.

According to SIT, when we identify with any social group we perceive oneness with or belongingness to that group (Mael & Ashforth, 1992). Thus members with high levels of OI feel an increased sense of oneness with or belongingness to their organization (Dainton & Zelley, 2005). Highly identi- fied members who define themselves in terms of their organizations (e.g., “We have been receiving a lot of media attention;” Mael & Ashforth) are more likely to perform duties with the perceived best interest of the organi- zation in mind (Tompkins & Cheney, 1985), and form a psychological bond to the organization, incorporating the organization’s attributes and values as their own and acting in ways that reflect those beliefs, values, and norms (van Knippenberg, 2000). Membership in the organization becomes a dis- tinct part of the identity of the individual (Ashforth & Mael, 1989). However, we argue such strong investment and alignment between one’s own and one’s organization’s values may not always be favorable, as is explored below. Essential to this argument is the attention Tajfel and Turner (1985) dedicate to the “pressures to evaluate one’s own group positively” (p. 16). These pressures can lead to in-group biases and—in the case of organiza- tional ethics—an increased likelihood to frame and give sense to potentially unethical situations in ways that do the least damage to one’s positive sense of self.

Scholarship on identity work (e.g., Sveningsson & Alvesson, 2003; Watson, 2008) emphasizes that identity and identification are processual, complex, and everchanging. Individuals may identify with multiple targets— individual, work group, organizational, professional (Scott, Corman, & Cheney, 1998)—which may be conflicting, converging, or combined (for a review, see Ashforth, Harrison, & Corley, 2008) and may give preference to one over another at a given time (Kuhn & Nelson, 2002). Intensity of identifications fluctuates. This fluctuation may be a result of a negative event associated

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
158 Management Communication Quarterly 27(2)

with an identification target. Another possibility when one’s identity is threat- ened is that once-identified members may lower their identification with the organization and experience disidentification (Scott, 2007). Elsbach and Bhattacharya (2001) define organizational disidentification as “a self-perception based on (1) a cognitive separation between one’s identity and one’s percep- tion of the identity of an organization, and (2) a negative relational categori- zation of oneself and the organization” (p. 397). The process of disidentification may be triggered when identified members perceive incongruences between an organization’s values and their own or when members feel that their own reputation is threatened by the organization’s reputation (Elsbach & Bhattacharya, 2001). Thus it is plausible that identified members of an orga- nization accused of wrongdoing may begin to distance or separate themselves from that organization, cognitively and communicatively, or be motivated to work internally to correct the perceived misalignment (i.e., issue selling; Dutton, Ashford, O’Neill, Hayes, & Wierba, 1997).

Organizational identification is worthy of extensive investigation; yet in the past, management researchers tended to investigate the positive effects of OI for organizational outcomes like goal achievement, performance quality, and job satisfaction (e.g., Likert, 1967; McGregor, 1967), task involvement, and satisfaction with work, pay, supervision, and promotion (e.g., Efraty & Wolfe, 1988). Ashforth et al. (2008) advocate a multifaceted understanding of identification when they state, “. . . It is important to understand the dynam- ics, risks, and potential of identification in today’s organizations” (p. 360). Yet explorations of the negative effects or risks of OI for organizational out- comes (e.g., Gossett, 2002) remain less common in communication and management.

Consider, for example, communication scholars’ explication of unob- trusive control (Barker, 1993; Tompkins & Cheney, 1985) and its potential to highlight a dark side of identification for organizational members. Bisel, Ford, and Keyton (2007) define unobtrusive control as “the process by which members of an organization are guided [by organizations] in making organi- zationally relevant decisions” (p. 137). The researchers point out how mem- bers can be both “controlled by and resistant to the influence of their identifications” (p. 155). Barker (1993) described the related dynamic of con- certive control as control that is developed by members and occurring by “members . . . reaching a negotiated consensus on how to shape their behav- ior according to a set of core values” (p. 411). Barker found these values and the “value-based normative rules . . . controlled [members’] actions more powerfully and completely than the formal system” (p. 408). Similarly, Zoller (2003) demonstrated how members—driven by identity needs—may even

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 159

consent to dangerous working conditions by accepting and enacting organi- zational and community norms, and in doing so, control themselves in ways that appear to cause bodily danger and harm. These sophisticated communi- cation studies have in common the notion that OI can be detrimental for mem- bers. This study, however, demonstrates how high OI may be potentially detrimental for organizations’ capacities to engage in ethical learning because highly identified members’ communication tends to give sense to organiza- tional wrongdoing in insular and defensive ways instead of ways that readily admit fault and accept responsibility.

If a given organizational culture values practices that may be deemed unethical by stakeholders, it is argued here that members’ identification with that organization is detrimental not only to members but also for the organi- zation’s moral learning capacity. Being highly identified with an unethical organization could alter members’ ability and willingness to notice and acknowledge the unethical nature of the organization because doing so is akin to noticing and acknowledging their own unethical nature. As noted by Dukerich, Kramer, and Parks (1998), highly identified members may engage in unethical acts, turn a blind eye to evidence of such behaviors, or even make cover-up attempts as means of protecting their own positive senses of self. Similarly, we describe ethical sensegiving as communicative attempts to influence how moral meaning is assigned to events or actions. We argue that, for highly identified members of organizations accused of wrongdoing, this process tends to unfold as a defensiveness that invites others to assign moral meaning favorably (or, at least less negatively) on their organization and themselves. Next, we describe ego defense and its connection to ethical sensegiving.

Defense Mechanisms In organizational science, ego defense mechanisms refer to those automatic strategies used by individuals and organizations to maintain and protect con- cepts of self (Brown & Starkey, 2000). Defense mechanisms range from maladaptive and immature to adaptive and mature (Segal, Coolidge, & Mizuno, 2007). If functioning at an appropriate balance, defenses can aid individual growth and maturity (Laughlin, 1970). However, in an organiza- tional setting, many ego defenses are maladaptive, harmful, and destructive— to both members and their collectives (Brown & Starkey). Strategies such as denial and rationalization not only prevent organizational learning (e.g., Brown & Starkey)—their usage may be key to understanding the persis- tence of unethical behavior, particularly when utilized by highly identified

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
160 Management Communication Quarterly 27(2)

members. If a highly identified member’s organization is labeled “unethical,” the member—by nature of identification—is also defined as “unethical.” The potential to be deemed unethical may drive members to engage in defensive sensegiving to protect their self-concepts.

Like organizational identification, defense mechanisms are often thought of as psychological phenomena. The Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) defines defense mechanisms as “automatic psychological processes that protect the individual against anxiety and from the awareness of internal or external dangers or stressors” (American Psychiatric Association, 2000). Laughlin (1970) uses the label of ego defenses since these defensive strategies “are evolved automatically by the psyche in order to avoid psychic pain and discomfort through the sought-after resolu- tion of emotional conflicts” (p. 4).

Davidson and MacGregor (1998) remark that “defense mechanisms oper- ate to protect self-esteem and, in more extreme cases, to protect the integra- tion of the self” (p. 967). Similar to defense mechanisms’ protection of self-esteem, identifications also contribute to one’s self-concept (Tajfel & Turner, 1985). In other words, while identifications serve as resources for how individuals define their self-concept, defense mechanisms describe how individuals defend those self-definitions from critical assessments. Thus a relationship between OI and defense mechanisms seems likely.

The Communication of Ethical Sensegiving While the defense mechanism itself is cognitive, that cognition can be manifested communicatively. Outward portrayals of defense mechanisms are known as defensive behaviors or those behaviors that decrease threat and reduce anxiety. Cramer (2000) asserts that defense mechanisms are generally employed in reaction to anxiety and distress. In the case of being asked about an organization’s unethical behavior, the threat of being defined as unethical- by-association through organizational membership is one such source of anxiety. This process could result in defense mechanisms in members’ mes- saging about their organization’s guilt.

Sensegiving refers to attempts at influencing how the self and others inter- pret the meaning of action (Gioia & Chittipeddi, 1991). Studies of sensegiving emerged from studies of sensemaking (e.g., Weick, Sutcliffe, & Obstfeld, 2005). Broadly defined, sensemaking is akin to interpreting and understanding— processes often initiated in efforts to reduce equivocal inputs (Weick, 1995). Research indicates sensemaking is an important function of organizational leaders who grapple with ambiguous, complex, and weak environmental

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 161

signals. Leaders must make sense of these inputs to plan and learn (Hill & Levenhagen, 1995; Weick & Sutcliffe, 2007). Gioia and Chittipeddi observed that organizational leaders not only seek to make sense of their environments but they also must engage in influence to move members to action (i.e., sensegiving). Subsequent studies of sensegiving demonstrate middle manag- ers and workers also engage in sensegiving with other members and impor- tant constituents external to the organization in carrying out the will of management (Rouleau, 2005). Likewise, Maitlis and Lawrence (2007) explained how sensegiving attempts were often triggered by situations per- ceived as important and uncertain. Matters involving group-level wrongdo- ing are likely important and, at times, uncertain for organizational members. Thus these moments likely present highly identified members with a percep- tion that a gap exists in their understanding of the meaning of their member- ship to the organization and the related identity implications. We show later that denying or avoiding recognition of wrongdoing are communicative sensegiving strategies that may delay admissions of error. These sensegiv- ing attempts are ethics-based because they attempt to influence how others assign moral meaning to events or actions and are thus, moments of ethical sensegiving.

Organizational Identification and Defense Mechanisms As explained above, it stands to reason that employees’ organizational iden- tification may be related to their defensiveness in cases of unethical orga- nizational behavior. Brown and Starkey (2000) provide a psychodynamic analysis of organizational identity and learning, describing the ego defenses that organizations employ to maintain collective self-esteem. Such ego defense practices may create a barrier for organizations’ moral learning capacity (or, their ability to recognize, learn, and grow from [un]ethical situ- ations in which they find themselves). Highly identified members may find it difficult to interpret their organization—and by extension, themselves—as guilty of wrongdoing and, instead, engage in defensive sensegiving in an attempt to maintain their own and others’ positive impressions of the organi- zation. While Umpress, Bingham, and Mitchell (2010) explored the possibility that highly identified members engage in unethical proorganizational behav- iors, this study proposes an investigation of the specific communicative features and defense mechanisms members use when discussing and giving sense to their organization’s wrongdoing.

Thus, rather than a psychodynamic perspective of ego defenses (i.e., Brown & Starkey, 2000), this investigation is indicative of a communicodynamic

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
162 Management Communication Quarterly 27(2)

approach to the interpretation of and response to organizational wrongdoing and sensegiving (Bisel, Kelley, Ploeger, & Messersmith, 2011). It is posited here that those who are highly identified will interpret an attack on the orga- nization (i.e., questioning the organization’s innocence) as akin to an attack on the self, in which case motivation to protect one’s self-identity, and in turn, organizational identity is activated. The vocalization of thoughts on one’s organization’s wrongdoing will likely contain intense and frequent defense mechanisms, operating to reduce associated guilt and uncertainty (Menzies, 1970) to maintain a favorable public organizational (and self) image. Thus the following hypotheses are proposed:

Hypothesis 1a (H1a): Organizational identification is positively associ- ated with the felt intensity of members’ linguistic defensiveness on behalf of their organization, after controlling for certainty of organi- zational wrongdoing.

Hypothesis 1b (H1b): Organizational identification is positively asso- ciated with the frequency of members’ usage of linguistic defense mechanisms on behalf of their organization, after controlling for certainty of organizational wrongdoing.

Certainty of Wrongdoing and Defense Mechanisms As is the case in all investigations of communication and of ethics, the importance of context cannot be underestimated. The degree of certainty with which members can say that their organization did (or did not) engage in unethical behavior is one such contextual issue. Ambiguity acts here as an interpretive resource from which highly identified individuals can draw to make sense of wrongdoing in ways that protect their self-concepts from threat. Menzies (1970) described defense mechanisms as occurring to avoid or reduce uncertainty. Thus it stands to reason that if the unethical situation is ambiguous and few details are known, members will have more discursive resources to manage the ambiguity strategically and defend their organiza- tion and, in turn, themselves. Again, ambiguous environments can actually serve as triggers for sensegiving (Maitlis & Lawrence, 2007). Presumably then, ambiguity affords the opportunity for individuals to evoke defense mechanisms to counteract the threat of being associated with the unethical organization. On the other hand, if details are known and organizational guilt is certain, members may find it more difficult to defend the organization’s behavior and image and thus display less intense and frequent defensiveness than a member whose sensegiving can draw on the resource of ambiguity. Thus the following hypotheses are posited:

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 163

Hypothesis 2a (H2a): Participants assigned to the ambiguous organiza- tional wrongdoing condition report higher levels of felt intensity of linguistic defensiveness on behalf of their organization than partici- pants assigned to the certain organizational wrongdoing condition, after controlling for participants’ levels of organizational identifica- tion.

Hypothesis 2b (H2b): Participants assigned to the ambiguous orga- nizational wrongdoing condition produce more frequent usage of linguistic defense mechanisms on behalf of their organization than participants assigned to the certain organizational wrongdoing con- dition, after controlling for participants’ levels of organizational identification.

Tenure, Organizational Identification, and Linguistic Defensiveness In alignment with previous research on tenure—or membership length (Sass & Canary, 1991)—three final hypotheses are proposed. Veteran members’ organizational identification is likely to be stronger than newcomers’ organi- zational identification in that psychological associations in veterans have more opportunities to be reinforced. If a member remains with an organiza- tion, it is also plausible to predict that an individual’s membership in the organization will be increasingly likely to represent a defining feature of the individual’s self-concept. Organizational identification has been shown to be related positively to length of employment (e.g., Hall, Schneider, & Nygren, 1970; Mowday, Porter, & Steers, 1982). A similar finding seems likely in the context of this study as well. Thus it is proposed as follows:

Hypothesis 3a (H3a): Organizational tenure is positively associated with reported levels of organizational identification.

Furthermore, it stands to reason that veteran members, because of the increas- ing likelihood the organization will represent a defining feature of their self- concepts, likely also show a higher propensity to defend their organization intensely and with frequent defense mechanisms in instances of organiza- tional wrongdoing. Thus the following hypotheses are explored:

Hypothesis 3b (H3b): Organizational tenure is positively associated with the felt intensity of members’ linguistic defensiveness on behalf of their organization.

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
164 Management Communication Quarterly 27(2)

Hypothesis 3c (H3c): Organizational tenure is positively associated with the frequency of members’ usage of linguistic defense mecha- nisms on behalf of their organization.

Method Participants

Full-time working adults (N = 318) participated in this language production experiment. The sample consisted of 161 males and 153 females (4 did not answer), ranging in age from 21 to 70 (M = 42.37, SD = 12.35). Participants resided in 29 of the United States; 1 participant resided in Australia. Education levels ranged from a high school diploma to a doctorate; bachelor’s degrees were the most common educational level obtained (34%). Participants’ total working experience ranged from 6 months to 48 years (M = 21.27, SD = 13.20). Participants’ work experience with their current organization ranged from less than 1 month to 33.25 years (M = 12.79, SD = 12.30). The average amount of supervisory experience was 8.36 years (SD = 10.24), ranging from none to 43 years. Participants worked in a variety of organizational sizes: (a) small/home/micro organization (< 10 members, 6.6%); (b) small organiza- tion (< 100 members, 15.7%); (c) medium-sized organization (< 500 mem- bers, 12.6%); and (d) large organization (> 500 members, 64.2%; 3 did not answer).

Procedures and Design The researchers recruited participants from their professional and social networks. All solicited participants were asked to forward a solicitation email to five other working adults, who were asked to forward the email to five more working adults and so on. Participants were directed to an online survey hosted by Qualtrics®. Each participant read a consent form before participating and provided basic demographic information. Then, partici- pants were randomly assigned to one of two scenario conditions (i.e., ambiguous or certain organizational wrongdoing) and responded to a prompt (see description below). Lastly, participants completed a measure of organizational identification (Mael & Ashforth, 1992). Each section of the survey is described below, presented in the order in which participants completed the survey. Scenarios. Organizational wrongdoing was operationalized by gender dis- crimination. Gender discrimination is an appropriate operationalization of

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 165

wrongdoing in that this sort of unethical organizational behavior is applicable to a wide-variety of organizations and industries. The two scenario condi- tions in this experiment vary by the degree of ambiguity regarding the orga- nization’s guilt in committing gender discrimination (see Appendix A). The ambiguous organizational wrongdoing condition described the organization’s role ambiguously and the situation as uncertain—no official determinations of guilt or legal sanctions had been taken against the organization. Con- versely, the certain organizational wrongdoing scenario condition indicated that the participant’s organization was found guilty at the end of a class action lawsuit. Participants were then asked to respond to one open-ended question from a family member who inquired about the situation (i.e., What do you think about these accusations?). Details about the manipulation check are available from the corresponding author. Measure of Organizational Identification. Participants completed a brief, six-item organizational identification questionnaire (Mael & Ashforth, 1992). Questionnaire items are designed to measure participants’ perceptions of oneness with or belongingness to an organization (e.g., When I talk about my organization, I usually say “we” rather than “they.”). Responses ranged from 1 (strongly agree) to 5 (strongly disagree). Each item was recoded so higher scores on the measure indicated greater identification. Coefficient alphas for the shortened, six-item identification questionnaire have been reported as .81 (Mael, 1988) and .87 (Mael & Ashforth). Cronbach’s α for this study’s use of the measure was .83. Linguistic Defensiveness. Defense mechanisms are most often measured using self-report questionnaires (e.g., the Defense Mechanism Inventory [Gleser & Ihilevich, 1969]), which attempt to capture the psychological real- ity of ego-defensiveness. However, self-reports are unlikely to capture the communicative reality of ego-defensiveness. By contrast, analysis of mes- sage defensiveness better facilitates an understanding of communicative manifestations of ego-defensiveness. As such, linguistic defensiveness—a concept original to this investigation—was used to describe the communica- tive strategies employed by individuals as they defended and gave sense to their organization’s wrongdoing. For the purposes of this study, linguistic defensiveness is measured as two dependent variables: response intensity and defense mechanism frequency. While related, these variables differ both conceptually and pragmatically; response intensity measures the degree of affective arousal respondents experienced psychologically as they responded—which may or may not be reflected communicatively. Defense mechanism frequency measures the number of defense mecha- nisms invoked by respondents in their messaging. This distinction highlights

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
166 Management Communication Quarterly 27(2)

the possibility that internal feelings of defensiveness may not necessarily be expressed outwardly.

Response intensity. To measure a first aspect of linguistic defensiveness— response intensity—participants responded to a series of four, 7-point seman- tic differential scales after reading and responding to the scenario (cf. Waldron & Krone [1991]). Items to measure participants’ felt intensity included very unintense/very intense, very unforceful/very forceful, very unemotional/very emotional, and very unpassionate/very passionate. Scale reliability was suf- ficient, Cronbach’s α = .88.

Content Analysis Defense Mechanism Frequency. The frequency of defense mecha- nisms in participant responses was a second method of measuring linguis- tic defensiveness. To this end, an inductive coding scheme was created. The researchers relied partly on the identity and ego defense strategies described by Laughlin (1970), Brown and Starkey (2000), Benoit (1995), and Benoit and Hanczor (1994), such as denial and rationalization, among others.

Across two stages, two coders reviewed and analyzed open-ended responses to the scenarios to create a codebook of defense mechanisms. In the first stage of codebook development, coders read and reread 15 responses gathered in a pilot study to identify recurrent defensive communication strat- egies. Using a technique similar to open-coding in constant comparative analysis (Glaser & Strauss, 1967), coders identified similar groups of defen- sive strategies. In the second stage of codebook development, coders selected a random subset of responses from the full study data (n = 35), first to refine the coding scheme by seeking out response strategies that could not be easily coded and then to incorporate those responses by modifying the coding scheme. At the conclusion of the second stage of codebook development, categorical saturation was achieved in the sense that no new categories indic- ative of linguistic defensiveness could be located. Only one response of the total data set of 319 responses was deemed uncodable, which indicates cod- ing scheme exhaustiveness. Krippendorff’s α was computed on yet another randomly selected subset of the data (n = 36). Intercoder reliability was suf- ficiently high at α = .86. Intercoder reliability was again calculated at the completion of coding and was also acceptable, α = .73 (Hayes & Krippendorff, 2007). Coding Scheme. The first step in the coding process was to assess whether a response readily admitted guilt. If so, the response was assigned a code of

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
Ploeger and Bisel 167

0. If the response did not readily admit guilt and was, as a whole, defensive, the number of defense mechanisms within the response was summed. Defense mechanisms were counted according to each time participants’ language performed one of the five following functions. Multiple defense mechanisms performing the same function could appear in one response. First, some defense mechanisms attempted to (a) bolster the organization (e.g., I have had a great experience with the organization or I am proud to be a member of its team). Second, some defense mechanisms attempted to (b) create discursive closure (e.g., I am not at liberty to comment. Now, go have another beer or I don’t feel comfortable discussing the accusations since they don’t pertain directly to me). Third, some responses were defen- sive by attempting outright (c) denial (e.g., No, it is not our organization). Fourth, some defense mechanisms attempted to (d) undermine the accusa- tion or claim (e.g., Don’t believe everything you read until the facts come out or There are two sides to every story). Finally, and similarly, some par- ticipants attempted to defend their organization by (e) minimizing the situa- tion (e.g., I think at this point they are just rumors or There is not a company out there that is perfect). It is important to note that for minimizing the situ- ation to be coded, responses had to admit an awareness of the situation explicitly and characterize the accusation as illegitimate. For example, an utterance like, “I have only heard hearsay” did not make it apparent the par- ticipant was minimizing the situation because it could have been a statement of fact, drawn from the ambiguous scenario condition. Thus, such utterances were not coded as minimizing the situation. However, utterances in which participants characterized and demonstrated awareness of the situation explicitly were coded as minimizing the situation. To further illustrate the coding scheme, the following participant response received a code of 4 because it contained four defense mechanisms:

I haven’t personally experienced any problems [bolster the organiza- tion], I think there can often be extenuating circumstances [undermining the accusation or claim] when litigation is involved as well as policy change. ABC Corp is a good company [bolster the organization] and will do the right thing [bolster the organization].

Likewise, the following response received a code of 1 for containing a single defense mechanism: “I can’t really comment on whether I think that the accu- sations are true or false, but there is not a company out there that is perfect [minimizing the situation].”

at UNIV OF THE INCARNATE WORD on February 25, 2016mcq.sagepub.comDownloaded from

http://mcq.sagepub.com/
168 Management Communication Quarterly 27(2)

Coding Scheme Validation. Krippendorff (2004) recommended coding schemes be assessed for three categories of validation: face, social, and empir- ical validity. As face validity is associated with a coding scheme’s plausibility, we believe the scheme, based on existing literature and induction, meets this criterion. Krippendorff describes social validity as whether a coding scheme allows content analysts to “address important social issues” (p. 319). The his- tory of organizational wrongdoing meets this criterion well. Empirical validity refers to “the degree to which available evidence . . . support . . . [the] research process and its results” (Krippendorff, 2004, p. 319). An important means of establishing the empirical validity of a coding scheme is through semantic vali- dation, “the degree to which analytic categories accurately describe meanings and uses in the chosen contexts” (Krippendorff, 2004, p. 319). The scheme was subjected to a test of semantic validity by asking 10 working adults to read and rank order three responses from least to most defensive. Krippendorff’s α was computed across participants’ rankings to measure how consistently these untrained coders assigned meanings of linguistic defensiveness in alignment with the coding scheme. The untrained coders’ rankings aligned well with the coding scheme, α = .80. Lastly, the coding scheme achieved functional valid- ity (another form of empirical validity) through its success in detecting differ- ences described in the results section.