Abstract
We introduce a new nonverbal and unobtrusive measure to assess power motive activation, the Spatial Power Motivation Scale (SPMS). The unique features of this instrument are that it is (a) very simple and economical, (b) reliable and valid, and (c) sensitive to situational changes. Study 1 demonstrates the instrument's convergent and discriminant validity with explicit measures. Study 2 demonstrates the instrument's responsiveness to situational power motive salience: anticipating and winning competition versus losing competition and watching television. Studies 3 and 4 demonstrate that thoughts of competition result in higher power motivation specifically for individuals with a high dispositional power motive.
Acknowledgments
We are grateful to Patrick Mueller and Sven Sauer for their ideas and insightful comments on the development of the SPMS. We also thank Susanne Weigl for her useful theoretical input and her help in data collection.
Notes
Subsamples A and B did not differ significantly with regard to gender, age, highest level of education, student status, mean professional experience, or occupancy of current leadership position (all ps >.11). With regard to occupancy of a prior leadership position, the difference between subsamples was marginally significant (p =.08): In Subsample B slightly more participants indicated having been in a leadership position before than in Subsample A.
Five participants with values > 3 SD above the sample mean were excluded from this analysis, as it was apparent from the data that they had paused and later resumed filling out the SPMS.
Because the number of participants who indicated having occupied a leadership position in the past differed marginally significantly between subsamples (see Footnote 1), we also calculated the correlations with the SPMS separately for the two subsamples. Results showed that the correlation in Subsample A, r(117) =.09, p =.36, failed level of significance and was smaller than the correlation in Subsample B, r(116) =.19, p =.04, which was significant. However, because both correlation coefficients were positive and did not significantly differ from each other according to a Fisher's Z test, Z = –0.08, p =.44, we report the correlation for the whole sample in .