Fluckinger, Chris D. (2014, August 9). Big Five measurement via Q-sort: An alternative method for constraining socially desirable responding. SAGE Open, 4(3). (Accessible at http://journals.sagepub.com/doi/10.1177/2158244014547196)
Abstract: Socially desirable responding presents a difficult challenge in measuring personality. I tested whether a partially ipsative measure—a normatively scored Q-sort containing traditional Big Five items—would produce personality scores indicative of less socially desirable responding compared with Likert-based measures. Across both instructions to respond honestly and in the context of applying for a job, the Q-sort produced lower mean scores, lower intercorrelations between dimensions, and similar validity in predicting supervisor performance ratings to Likert. In addition, the Q-sort produced a more orthogonal structure (but not fully orthogonal) when modeled at the latent level. These results indicate that the Q-sort method did constrain socially desirable responding. Researchers and practitioners should consider Big Five measurement via Q-sort for contexts in which high socially desirable responding is expected.
Chris D Fluckinger <cdfluck@bgsu.edu> is an instructor of psychology in Firelands College, Bowling Green State University, Huron, OH (USA). His primary research interests include personality measurement in the workplace, applied health psychology, and aging issues in the workforce.
