Rasch-transformed raw scores and Two-way ANOVA: a simulation analysis.

摘要: This article demonstrates, through the exposition of underestimated variable effects or spurious interaction effects, inherent inadequacy untransformed 0-1 raw scores for analysis via Two-Way variance (ANOVA). The pioneering work in this field, conducted 1990's by Dr. Susan Embretson University Kansas, USA, is highlighted, and eminent suitability Rasch transformations ANOVA also demonstrated. In study Monte Carlo techniques simulations are utilized to determine precise psychometric conditions under which differences between those detectable ANOVA. partially replicates Embretson's studies, defines extent underestimation spuriousness ensue when uniform skewed distributions item difficulties used instead normal distributions, misfitting data fitting data.