Faking On Personality Tests Under Personnel Selection Situation:An Eyemovement Study

In the21st century, in order to be more objective and quantitative of evaluating the candidate’s skills, values, personality, etc., personality tests are widely used in personnel selection. However, test items can be selected according to candidates’social expectations or operational needs, namely Faking in their personality tests. Faking on Personality Test will lead to the phenomenon of distortion, what’s worse, HR cannot know clearly about the candidates’personality types and working attitudes, which will greatly bring down the efficiency of people-post matching, let alone the effect of personality test. This paper try to explore the phenomenon of Faking in Personnel Test. Using two eye movement experiments to make an comparison of two conditions’(Faking good instruction and Respond honestly) response latencies and eye-movement indexes. We try to know whether these indexes (response latencies, dwell times, fixation counts, pupil sizes) can infer the candidates’ Faking, whereas effectively improve the reliability and availability in Personnel selection.In an experimental within-participants design, a Big Five personality test was administered to participants in2conditions: faking good instruction and respond honestly(time interval9months). Item responses, response latencies, and eye movements index were measured. Results showed that:(1) Differences between two conditions on item responses was significant, all personality dimensions were fakeable;(2) Differences between two conditions on response latencies was significant(p=0.00<0.05), response times were on average365.88ms slower in the fake good condition;(3) Differences between two conditions on dwell times (p=0.021<0.05) and fixation counts (p=0.007<0.05) in the interest of key words(items) were significant, participants had fewer dwell times(d=90ms) and less fixation counts(d=0.39) in faking good conditions;(4) Differences between two conditions on dwell times (p=0.00<0.05), fixation counts(p=0.00<0.05) and pupil sizes (p=0.00<0.05) in the intrest of extreme response options(thel and5-point answering scale) were significant, participants had more dwell times(d=118.06ms), more fixation counts(d=0.31) and more expanded pupil sizes(d=l78.81μm) in faking good conditions;(5) Differences between two conditions on dwell times(p=0.00<0.05), fixation counts(p=0.00<0.05) and pupil sizes (p=.005<0.05) in the interest of neutral response options(the2,3and4-point answering scale) were significant, participants had fewer dwell times(d=589.47ms), less fixation counts(d=1.96) and less expanded pupil sizes(d=18.63μm) in faking good conditions. Our research found that, participants’eye-movement indexes on two conditions(faking good and respond honestly) are significant, eye-movement indexes can be effectively infer the cognition process of the participants’ faking in the personality test. Our research provides a new train of thought for identifying faking in personality test.