Responses to Organizational Mandates: How Voice Attenuates Psychological Reactance and Dissent

Abstract

Organizations sometimes create policies that restrict the decision freedom of their members. When doing so, they might create psychological reactance and dissent. This study examines whether providing voice into a decision can reduce the likelihood that those affected by the decision perceive it as imposing on their rights, can have negative emotional reactions to its adoption, and can want to engage in organizational dissent. Undergraduate students were randomly assigned to read scenarios in which a university committee decided to recommend that a university-wide, mandatory, comprehensive exit examination requirement be adopted after a student group either supported the requirement or opposed it. The results confirmed an indirect path between voice and dissent that flowed through perceived imposition on students’ rights and negative emotional reactions to the adoption of the requirement. The limitations of the study and implications of the results for theory and research are discussed.

Keywords:

• Dissent
• Psychological Reactance
• Voice

Notes

Note. N = 195 for all correlations, except those involving the Hong Reactance Scale (N = 193) and Imposes on rights (N = 194). Reliability coefficients for multi-item scales are reported along the diagonal.

^a Voice: 0 = no voice, 1 = voice.

^b Message form: 0 = implicit, 1 = explicit.

*p < .05. **p < .01. ***p < .001.

Note. N = 193. Number of bootstrap samples = 5,000. All coefficients are unstandardized. LL = lower limit; UL = upper limit; CI = confidence interval; BootSE = bootstrap standard error.

^a Message form and Hong Reactance Scale are covaried from all relationships.

^b Bivariate relationship between voice and articulated dissent.

^c Relationship between voice and articulated dissent, controlling for indirect relationships.

^d Bootstrap percentile 95% confidence intervals and standard errors.

*p < .05. **p < .01. ***p < .001.

Note. N = 193. Number of bootstrap samples = 5,000. All coefficients are unstandardized. LL = lower limit; UL = upper limit; CI = confidence interval; BootSE = bootstrap standard error.

^a Message form and Hong Reactance Scale are covaried from all relationships.

^b Bivariate relationship between voice and latent dissent.

^c Relationship between voice and latent dissent, controlling for indirect relationships.

^d Bootstrap percentile 95% confidence intervals and standard errors.

*p < .05. **p < .01. ***p < .001.

Note. N = 193. Number of bootstrap samples = 5,000. All coefficients are unstandardized. LL = lower limit; UL = upper limit; CI = confidence interval; BootSE = bootstrap standard error.

^a Message form and Hong Reactance Scale are covaried from all relationships.

^b Bivariate relationship between voice and articulated dissent.

^c Relationship between voice and articulated dissent, controlling for indirect relationships.

^d Bootstrap percentile 95% confidence intervals and standard errors.

*p < .05. **p < .01. ***p < .001.

Kassing (1998) noted that dissent occurs when “employees combat psychological and political restraints imposed by modern organizations” (p. 184). Dissent may be openly expressed to decision makers, but may also be covert as organizational members criticize the policy among themselves or to outsiders.

This topic has been used in prior research and, although undergraduates typically oppose such a policy, under some circumstances, they can be persuaded when they find the topic to have personal consequences for them and a message contains strong arguments in favor of the policy (e.g., Petty, Cacioppo, & Goldman, 1981).

The macro provides a regression-based approach for conducting path analyses. It examines direct paths, as well as alternative indirect paths. There are three advantages to using this statistical method: (a) Multiple mediators can be simultaneously tested, (b) it does not rely on the assumption of a normal sampling distribution (see MacKinnon, Lockwood, & Williams, 2004; Preacher & Hayes, 2004; Shrout & Bolger, 2002), and (c) the number of inferential tests is minimized, thus reducing the likelihood of Type-1 errors. Using ordinary least squares regression, the macro tests possible indirect paths between an independent and dependent variable. We employed a bootstrapping method by which one can create a sample-based estimate of the indirect effect and bias-corrected, accelerated-confidence intervals. In this case, we created 5,000 possible samples. If the 95% confidence interval for an indirect path did not contain zero, there is evidence of mediation. For each analysis, message form and the Hong Reactance Scale were covaried from each path. A full explanation of the procedure can be found at http://www.afhayes.com/public/med3c.pdf.

As indicated in Table 2, the confidence interval associated with the hypothesized negative indirect path did not contain zero. Also, all of the confidence intervals for simpler versions of the indirect path contained zero. Table 3 contains the results for latent dissent. The confidence interval for the negative, indirect, hypothesized path did not contain zero, whereas those associated with the simpler versions did contain zero. Table 4 contains the results for displaced dissent. In this case, the results are a bit more complex. The confidence interval for the predicted negative path did not contain zero, but that was also true for one of the simpler paths. In this path, providing voice reduced perceived imposition, and this attenuated displaced dissent, regardless of negative feelings. Thus, the hypothesis was supported for all three forms of dissent, although for displaced dissent, another indirect path was also found.

Additional information

Notes on contributors

Willona O. Olison

Willona O. Olison (PhD, Northwestern University, 2006) is an Assistant Professor in the College of Communication at DePaul University.

Michael E. Roloff

Michael E. Roloff (PhD, Michigan State University, 1975) is an Assistant Professor in the Department of Communication Studies at Northwestern University.