Weber contends that the use of Rasch analysis, principal components analysis, and classical test analysis shows that an instrument designed to measure a "bilevel dimensionality" in probability achievement measures a single latent trait. That interpretation and the use of Rasch and classical analysis to establish unidimensionality are questioned. (Author/RL)