Individual Differences - Chapter 23 - Rasch Analysis (IRT)

Who developed Rasch analysis and when?

Georg Rasch, 1960.

What are the two key Rasch parameters?

Person ability and item difficulty.

How does Rasch differ from classical test theory?

Models item-level response probabilities, not just total scores.

If ability > item difficulty, what is predicted?

High probability of a correct response.

How is item difficulty estimated?

By the proportion (or logit transform) of people answering the item correctly.

What indicates poor model fit in Rasch?

Observed responses deviate from predicted probabilities → item or model needs review.

Why can Rasch improve fairness?

Flags biased/non-functioning items and enables comparisons across groups.

What is one limitation of Rasch analysis?

Complex maths/software; assumes unidimensionality.

What does “model fit” mean in Rasch?

Agreement between observed and predicted item response patterns.

Name one practical benefit of Rasch for test design.

Item-level diagnostics to refine the scale and improve fairness.