An Investigation Of Socially Desirable Responding Effect On16PF In Psychological Measurement Context

With the personality tests widely used in personnel assessment selectionand psychological measurement of recruitment, academicians and employershave to focus on the problem of response bias which may exist in the personalitytests. That socially desirable responding (SDR) is one of the most importantsources which influence the reliability and validity of measurements.SDR is a response of self-evaluation in a socially acceptable way, whichmakes oneself to look more suitable for the needs of the community. That ismainly the "fake good" phenomenon. According to the contents and degree ofSDR, SDR can be differentiated into self-deception enhancement (SDE) factorand impression management (IM) factor. SDR is one of the most important errorsources for personality tests, therefore it is important to investigate SDR how toinfluence construct validity, predictive validity and selection decisions ofpersonality tests in theory and practice. This study included two parts:Part I, based on the two-factor model which put forward by Paulhus in1998, we prepared to design a SDR scale “the social desirability scale forpsychological measurement context”, which would be used in local culture;Then to assess the scale’s reliability and validity according to psychometricprinciples.Part II, based on the“the social desirability scale for psychologicalmeasurement context”, we prepared to discuss the impact of socially desirableresponding on16PF measurement validity. First of all, we would explore16PFconstruct validity stability between the high-score group and low-score group ofthe new SDR scale, and under different test scenarios; Second, we would use alinear regression equation to correct the16PF results by using the new socialdesirability scale scores, then test whether16PF predictive validity and thepeople rate of rightly hired greatly improved.Here are the main conclusions of this study:(1) The social desirability scale for psychological measurement context hadbetter reliability and validity, and could be used as a tool for college or militaryschool students’ psychological measurement.(2) The correlation and exploratory factor analysis showed that the relationshipand construct validity of16PF factors had no significant differences between the“motivation context”(applicant setting) and “non-motivation context”(researchsetting). Similarly, multiple confirmatory factor analysis also showed that16PFsecondary factor theoretical structure had no significant differences between the“motivation context” and “non-motivation context”, the high-score group andlow-score group of the new SDR scale. SDR only affected some factor scores,but had no impact on the16PF structure validity. (3) After using a linear regression equation to correct the16PF results by the△SDR under the two contexts, the16PF factors’ predictive validity commonlyimproved. The three factors, Emotional stability, Apprehension and Tension,improved significantly at the0.01level.(4) SDR affected selection decision of the mental test context. The△SDRcorrection commonly improved the correct selection rate under “select-in”procedure. The correction effect added to maximum, about10%, when theselection ratio was from5%to10%. However, the correction did not improvethe selection accuracy under “select-out” procedure.