Designing Short Forms with IRT and Constrained Test Assembly
A methodological guide with concise findings and a companion technical report: reproducible simulations show why item selection should match the intended scoring goal.
INSIGHTS
Statistics, explained for decisions in the real world.
Practical guidance on designing stronger studies, evaluating programs and AI systems, and avoiding the statistical traps that make confident conclusions fragile.
FROM NEDIM
Notes from practice.
How to Choose a Reference Group in Regression: A Practical Guide
Four questions for choosing a regression reference group that makes the comparison decision-makers need easy to understand.
How to Score a Questionnaire: Total Score, Average, or Subscales?
A practical guide to choosing a questionnaire scoring rule that preserves the meaning of the measurement and the decision it supports.
How to Analyze Pre-Post Data: Change Scores vs. Baseline-Adjusted Models
A practical framework for deciding whether change scores or a baseline-adjusted model best answers the question your pre-post study is asking.
How to Choose Covariates for a Regression Model: A Practical Guide
Five questions for choosing regression covariates before analysis, so adjustment clarifies the comparison rather than changing it by accident.
How to Choose a Survey Scale for Research: A Practical Validity Checklist
Choosing a familiar questionnaire is not enough. A practical checklist for matching a survey scale to the claim, people, and decision in your study.
Do You Really Need a Multilevel Model? A Quick Diagnostic
Nested data doesn't always need a multilevel model. A quick diagnostic for when ignoring clustering quietly breaks your standard errors, and when it doesn't.
Your Power Analysis Is Probably Wrong — Here's Why
Where power analyses go wrong: effect sizes borrowed from inflated published studies, and other assumptions G*Power won't question for you.
What Grant Reviewers Actually Look For in Your Methods Section
What grant reviewers look for in a methods section, from a statistician who designs federally funded studies and reviews other people's proposals.
HAVE A QUESTION OF YOUR OWN?