Tayebe Dehghan nayeri; Noorali Farrokhi; Ahmad Borjali; Ebrahim Alizadeh
Abstract
The purpose of the present study was to investigate the precision of the statistics based on the Rasch model (Up , Wp, lp &lzp) to detect response style of acquiescence in polytomouse data. Two sampels size (2000 and 500 data) was simulated according to Partial Credit model.The precision of ...
Read More
The purpose of the present study was to investigate the precision of the statistics based on the Rasch model (Up , Wp, lp &lzp) to detect response style of acquiescence in polytomouse data. Two sampels size (2000 and 500 data) was simulated according to Partial Credit model.The precision of person-fit statistics was compared under different testing conditions (Sample size, test length, percentages of aberrant-responding examinees and percentages of aberrant items) together. In order to investigate precision of the statistics used area under the ROC curve. The results showed that Up and Wp statistics had high precision to detect response style of acquiescence in all situations. About these statistics distinguished between the compatible and aberrant response patterns with the highest precision (>./9). lp and lzp statistics had low precision(./5<) in more situations and they could not distinguish response style of acquiescence. It is suggested that be used Up and Wp statistics to detect of the response style of acquiescence . Especially when persons want to be hired and they try to justify themselves.
