Intertest Chances in Acquiescence

Abstract

Seventy Ss at Utah State University were administered the Gough ACL and the Taylor MAS in counterbalanced order. Separate true and false scores were obtained for each test and were converted into proportions. Differences between these true proportions and false proportions were analyzed in a disproportionate multiple covariance design with the effects of sex and order of testing treated as deviations from the general mean. Estimates of variance due to set and content within each test were treated as covariants and employed as partial regression coefficients in the design. Results seemed to indicate that true and false responding are not parallel strategies, and that the former seems more content bound than the latter.