Abstract
In this study, we investigated how the perceptions of ethical climate as it relates to supervisor behavior and impression management are related to admission of past unethical behavior in a sample of Canadian Army personnel. We expected that among respondents who perceive less ethical supervisor-related climate, the relationship between Impression Management (IM) and the admission of past unethical behavior should be significant and positive. In contrast, among respondents who perceive a more ethical supervisor-related climate, the relationship between IM and self-report past unethical behavior should be weak or nonsignificant, and the frequency of admission of past unethical behavior should be low. Although moderation occurred, it was not as hypothesized. IM predicted admission of having obeyed, or witnessed others obey an unlawful command when perceptions of supervisor-related ethical climate were low, and not when perceptions of supervisor-related ethical climate were high. However, higher (vs. lower) impression managers admitted less past unethical behavior, regardless of their perceptions of supervisor-related ethical climate. These results suggest that high impression managers do not admit to having obeyed or witnessed others obey an unlawful command in the past, regardless of their perceptions of the ethical climate as it relates to supervisor behavior.
Notes
1 Information on the psychometric properties of the complete scale is available from the first author.
2 Although the factor loadings of the two items in this component (i.e., “Witnessed others obey an unlawful command and did not report it” and “Followed orders from a superior that you knew were against regulations”) justified inclusion in one scale, it could be argued that they may differ in how active respondents were in engaging in these two types of unethical behavior. As such, for exploratory purposes, we conducted separate regression analyses using each of the two items separately as the criterion. The results of the regression analysis for each item were the same as when the items were combined into one scale. For this reason, we combined the two items into one scale and labeled it Responses to Unlawful Commands.
3 Cohen’s f2 was used a measure of effect size in the regression analyses, with effect sizes of .02, .15, and .35 indicating small, medium, and large effects, respectively, and was calculated as follows:
where R2 is the squared multiple correlation.