psychometrics

What problem does it solve?

This Skill helps you validate and refine a psychological questionnaire by quantifying item quality, reliability, factor structure, and latent trait behavior so your scale measures what it claims to measure.

Core Features & Use Cases

• Classical Test Theory diagnostics: computes item difficulty proxies, skewness/kurtosis, corrected item-total correlations, and alpha-if-deleted to identify weak or misfitting items.
• Reliability estimation: calculates Cronbach’s alpha with a bootstrapped confidence interval and common interpretive guidance.
• Factor analysis workflow: runs EFA with KMO/Bartlett tests and parallel-analysis factor retention, then performs CFA with semopy to check model fit.
• Latent trait modeling (IRT): estimates a 2PL item response model for discrimination and difficulty parameters and supports DIF-oriented interpretation.
• Differential Item Functioning (DIF): flags potential DIF using Mantel–Haenszel testing with Bonferroni control.
• Measurement invariance testing: tests configural/metric/scalar invariance across groups using CFA comparisons.

Use Case: You have a draft 20-item scale administered to 500 respondents, and you need to (1) drop problematic items, (2) confirm the intended factor structure, (3) check whether reliability is acceptable, (4) model item difficulty/discrimination, and (5) test whether the scale behaves similarly across demographic groups.

Quick Start

Ask the AI to run a complete psychometric validation pipeline on your item responses by computing CTT item statistics, Cronbach’s alpha, EFA factor retention, CFA fit indices, 2PL IRT parameters, and Mantel–Haenszel DIF flags from a pandas DataFrame.