This post describes an R function -- salientDIF -- that estimates the expectation of the proportion of evaluees having absolute differences in DIF-naive and DIF-adjusted latent trait estimates falling above a specified threshold. Used as an impact measure in DIF studies.

Rich Jones

17 Jun 2024

Topics: DIF, R, Mplus, MplusAutomation

This is a collection of pictures to use when trying to explain  or understand the design fo the Health and Retirement Study (HRS).

Rich Jones

10 November 2023

Topics: HRS

The Bayesian Information Criterion (BIC) has been described using two different formulas. One leads to positive numbers (usually) and among competing models the smaller is preferred. The other leads to negative numbers (usually) and among competing models the larger is preferred. Do you know what flavor you are using?

Rich Jones

6 November 2023

Topics: BIC, Mplus, Proc Traj, Stata traj

This is a work in progress but might be useful. Cohen's d family effect sizes are well covered.

Rich Jones

31 October 2023

Topics: Mplus, Effect Size, MplusAutomation (R)

Edited sections from a grant proposal using multilevel factor analysis and dynamic SEM (location scale modeling). 

Rich Jones

4 August 2022

Topics: Factor analysis, Multilevel data, Dynamic SEM

About the AAPOR (American Association for Public Opinion Research) response rate definition (see also this page from the AAPOR).

Rich Jones

July 28, 2022

Topics: Response Rate, AAPOR, SymPy

Use Mplus from R with MplusAutomation, and extract ordinal dependent variable item parameter estimates and use them to prepare graphical summaries (item characteristic curves, boundary response functions, etc.)

Rich Jones

May 19, 2022

Topics: Item Response Theory (IRT), Mplus, MplusAutomation, ggplot2

Relating the ETS MH D-DIF categorization into negligible, slight to moderate, and moderate-to-large DIF categorization to direct effect magnitudes using (Mplus) categorical data confirmatory factor analysis with covariates (so-called MIMIC modeling).

Rich Jones

November 10, 2021

Topics: Item Response Theory (IRT), Differential item functioning (DIF), Mplus

Special alert for folks using haven::write_dta to share data between R and Stata. The feature of haven::write_dta in exporting factor variables and labels extends to ordinal factors as well. 

Rich Jones

September 10, 2021

Topics: R, Stata, Factor variables, value labels, data management

This is very drafty. Not a lot of explanatory text. 

Rich Jones

August 20, 2021

Topics: Item Response Theory (IRT), Differential item functioning (DIF)

Rich Jones

July 19, 2021

Topics: Mplus, R, Item Response Theory

Rich Jones

May 9, 2021

Topics: R, R/Markdown

Rich Jones

May 8, 2021

Topics: Mplus, Item Response Theory, Factor analysis

Rich Jones

April 15, 2021

Topics: Mplus, Item Respose Theory, Factor Analysis

Jones, Glymour, Crane, Gibbons, and Manly

6 January 2019

Topics: Garden of Forking Paths

Rich Jones

May 19, 2019

Topics: SEM, fit 

This essay includes a checklist to make sure you are including all of the things reviewers look for when evaluating your statistical power analysis and sample size justification.

Rich Jones

February 12, 2018

Topics: Power, Sample size, Proposals

The QSP Blog and research contributions are supported in part by the Department of Psychiatry and Human Behavior (DPHB), Brown University Warren Alpert Medical School. The content is solely the responsibility of the authors and does not necessarily represent the official views of the DPHB, or Brown University.