Present study was designed to examine the effects of emotional loading upon the MMPI response latency. Scaled scores on emotional loading, social desirability and item ambiguity were obtained through successive interval method based on 100
undergraduate
students' rating for the first 270 MMPI items. Computerized MMPI were administered to other 100 undergraduate students. Computer-assisted instruction system automatically recorded response latency on each item for each subject. Multiple
regression
anlysis was done for the first 270 items. Predictors of the mean latency were item length(number of item character), and scale scores of emotional loading, of social desirability and of item ambiguity. Using same predictors, additional regression
analyses were done for two 40 item groups. Items of one group were selected among the items whose scale scores were below or above 27% on each scale. Items of the other were randomly sampled among remained 230 items.
@ES The results were as follows:
@EN 1) Multiple repression analysis using 270 items showed item's length explained 65.2% variance for response latency and emotional loading 3.5%, ambiguity and social desirability the less than 1% respectively.
2) Mulfiple regression analysis using characteristic 40 items showed item length explained 47.7% variance for response latency and ambiguity did 21.9%, and emotial loading did 9.6% respectively. On the others, randomy selected 40 items showed
only
the
item length explained 71.9% variance for response latency.
Clinically, this findings address the issues that analysis of response latencies in computer-administed MMPI may help identify the areas where patients are reluctant to respond or feel 'conflict'. It would seem that clinical applicability of
response
latency be very promising if items be carefully selected and procedures be elaboratly controlled.
|