HOME / LEARNING HUB

ALL 80 ARTICLES AVAILABLE IN ENGLISH · 中文版每篇一键切换

MODULE 1 · 01–07

What IRT even is

Explained with exams and bathroom scales; why total scores fall short; the model family.

MODULE 2 · 08–15

Reading the curves

ICC, information functions, the θ scale, Wright maps — every chart, demystified.

MODULE 3 · 16–25

Jargon, cleared out

Latent variables, local independence, likelihood, standard errors, DIF, reliability — plus a cheat sheet.

MODULE 4 · 26–31

Strengths & limits

What IRT does uniquely well, its assumptions, sample-size appetite — and when not to use it.

MODULE 5 · 32–38

Hardware, software, workflow

What computer you need, what R is, GUI options, the standard analysis pipeline, dirty data.

MODULE 6 · 39–45

Reading & using results

Parameter tables, spotting bad items, score reports, shortening questionnaires, item banks.

MODULE 7 · 46–49

When results look wrong

Negative discrimination, misfit checklists, results vs. common sense, runaway parameters.

MODULE 8 · 50

Epilogue

What an ordinary reader can now do after 50 articles.

A01What is IRT: from total scores to latent variablesIRT basics A02Cleaning large-scale questionnaire data: the full pipeline and common trapsData processing A03Professional analysis delivery where sensitive data never leavesWorking with us A04IRT in R: getting started with the IRTC packageEngineering A05CTT vs. IRT, compared thoroughly: when IRT becomes necessaryIRT basics A06A complete guide to sampling weights: design weights, nonresponse adjustment, calibrationSurvey statistics A07Reliability and validity: a complete evidence frameworkScales A08Scale development end to end: from construct definition to formal administrationScales A09The Rasch model: measurement philosophy and practiceIRT basics A10Polytomous models: choosing among PCM, RSM, and GPCMIRT basics A11Latent regression: how background variables enter IRT modelsIRT basics A12EAP, MLE, WLE: three ability estimators and how they differIRT basics A13Equating and linking: making scores from different forms comparableTesting practice A14DIF detection: quality inspection for test fairnessTesting practice A15Missing data: from deletion methods to multiple imputationData processing A16Detecting careless responding: a method inventory with practical adviceData processing A17Sample size planning and statistical powerSurvey statistics A18Million-scale IRT estimation: engineering answers to memory and speedEngineering A19Multidimensional IRT: when you need it and how to model itIRT basics A20Standard setting: Angoff, Bookmark, and other cut-score methodsTesting practice A21A quality-control system for survey data: from design to releaseData processing A22R package development from zero to release: the birth of a statistical toolEngineering A23A reproducible research workflow: R + Git + automated testsEngineering A24Calibration weighting: the mathematics of matching sample to populationSurvey statistics A25Merging sparse cells: robust tabulation of graded proportionsSurvey statistics A26Methodology for literacy-style assessments: ability is not "items answered correctly"Testing practice A27A client's guide to working with an external data-analysis teamWorking with us A28Why trust statistical software: automated tests and honest failureEngineering A29Computerized adaptive testing: principles, item banks, and prerequisitesTesting practice A30From questionnaire to report: the full journey of a large survey analysisWorking with us

COPYRIGHT

All articles on this site (the "IRT for Everyone" guide and the in-depth series) are original works of WEIAN Data Technology (Beijing) Co., Ltd., authored by Ma Kunxiang, and protected under the Copyright Law of the People's Republic of China. Reproduction, adaptation, mirroring, translation, or commercial use without written permission is prohibited. For personal sharing, please credit the author and link back to this site. Licensing: contact@weiandata.com

Questions after reading?
Need an expert on your project?

Email us →

Initial feasibility assessment is free.
See these methods as shipped software in Tools & Cases.

contact@weiandata.com