The impact of method effects on structural parameters in validation research
通过假设数据说明，在选拔研究中，一个“讨人喜欢”的方法因子会显著改变预测变量与效标之间的结构参数估计，并比较了LISREL分析与回归分析的优劣。
Use of a structural equations analysis to estimate the degree of predictor-related criterion bias in selection research is illustrated. For the hypothetical data considered in this paper, a likability methods factor produced significant changes in estimates of the structural parameters relating the predictors to the criterion. Regression analyses using two methods factors as covariates yielded similar results. The advantages of employing the more complicated LISREL analysis versus more familiar regression analysis are discussed and some suggestions for future research using the structural equations approach are presented.