Goodman and Kruskal's λ measures what?

• A. Goodman and Kruskal's λ
• B. Pearson's r
• C. Spearman's rho
• D. Cronbach's alpha

Answer: (A)

This item is about a measure used with nominal data to assess how strongly two variables are associated. Goodman and Kruskal's lambda quantifies the proportional reduction in prediction error when you know the category of one variable to predict the category of the other. In other words, it looks at how much knowing one variable helps you predict the other, based on a contingency table. The value ranges from 0 to 1, where higher values indicate stronger predictive association; 0 means knowing one variable doesn’t improve prediction, and 1 means perfect predictive reduction.

This differs from Pearson's r and Spearman's rho, which are correlations used with interval/ratio (and sometimes ordinal) data to describe the strength and direction of linear or monotonic relationships. Cronbach's alpha, on the other hand, measures internal consistency reliability among items in a scale, not the association between two categorical variables.