Selected Publications

(2024). Using plausible values when fitting multilevel models with large-scale assessment data using R. Large-scale Assessments in Education.

PDF Code

(2023). Using robust standard errors for the analysis of binary outcomes with a small number of clusters. JREE.

PDF Code Dataset

(2023). Accounting for heteroskedasticity resulting from between-group differences in multilevel models. MBR.

Code

(2023). Alternatives to logistic regression models when analyzing cluster randomized trials with binary outcomes. Prev Sci.

PDF

(2023). Seasonality of school climate. School Psychology Review.

(2023). The impact of restorative practices on the use of out-of-school suspensions: Results from a cluster randomized controlled trial. Prevention Science.

PDF

(2023). Practical Multilevel Modeling Using R. Sage Publishing.

Preprint PDF

(2023). Advanced categorical data analysis in prevention science. Prevention Science.

PDF

(2022). Evaluation of the Whole School Restorative Practices Project: One-year impact on discipline incidents. JSP.

PDF

(2022). Using cluster robust standard errors when analyzing cluster randomized trials with few clusters. BRM.

Preprint PDF Code Dataset

(2022). Alternatives to logistic regression models in experimental studies. JXE, 90(1).

PDF Code Dataset

(2022). Analyzing cross-sectionally clustered data using generalized estimating equations. JEBS.

Preprint PDF Code Dataset

Recent Publications

See CV for complete list

More Publications

(2024). Steadfast restorative and supportive approaches to student behavior during the COVID-19 pandemic. JEPC.

(2024). Positive student-teacher relationships and exclusionary discipline practices. Journal of School Psychology.

(2023). Usability and social consequences of the early identification system as a universal screener for social, emotional, and behavioral risks. In SP.

PDF

Recent & Upcoming Talks

More Talks

Multilevel Modeling with Large-Scale International Databases Using HLM (Philadelphia)
Apr 10, 2024 9:00 AM
American Educational Research Association
Using cluster robust standard errors to analyze nested data with a few clusters (Korea)
Oct 12, 2023 2:00 PM
23rd International Conference on Education Research
PDF
Using Cluster Robust Standard Errors to Analyze Cross-Classified Data with a Small Number of Clusters
Sep 29, 2023 2:00 PM
Society for Research on Educational Effectiveness
PDF
Multilevel Modeling with Large-Scale International Databases Using HLM (Chicago)
Apr 12, 2023 9:00 AM
American Educational Research Association
Alternatives to logistic regression models in the analysis of cluster randomized trials with binary outcomes
Jun 4, 2021 10:00 AM
Society for Prevention Research Annual Meeting
PDF Slides
Multilevel Modeling with Large-Scale International Databases Using HLM
Apr 15, 2021 1:00 PM
American Educational Research Association
Multilevel Modeling with Large-Scale International Databases Using HLM
Apr 16, 2020 9:00 AM
American Educational Research Association
Alternatives to logistic regression models in experimental studies
Feb 21, 2020 10:00 AM
PDF Slides
Instrumental Variables for Educational Researchers
Sep 20, 2019 9:00 AM
PDF Slides
Summer of R Workshop
Aug 12, 2019 9:00 AM
Summer of R Workshop
Slides

Recent Posts

More Posts

Last updated on 8 min read large scale datasets, multiple imputation

Years ago I had written a post on using multiple imputation, weights, and accounting for clustering using R. However, the process was actually quite cumbersome and now in 2024, there are more straightforward ways of handling this. 1. Load in the required packges library(dplyr) #for basic data management library(tidyr) #converting wide to tall library(estimatr) #estimating models with robust SEs library(mitml) #for imputation and analyzing MI datasets library(MLMusingR) #contains the sample dataset library(mice) #for carrying out the analysis with MI data library(modelsummary) #outputting the results nicely library(survey) #alternative (classic) way 2.

CONTINUE READING

Last updated on 12 min read wemix, plausible values

Syntax to accompany the article: Huang, F. (2024). Using plausible values when fitting multilevel models with large-scale assessment data using R. Large-scale Assessments in Education. When fitting multilevel models using large scale assessments such as PISA or TIMSS, it is important to account for: the use of weights at different levels and the presence of multiple plausible values. I am often asked how do you run this analysis in R.

CONTINUE READING

Last updated on 3 min read jobs4

List of Research, Statistics, and Evaluation job postings (that I’ve seen) as of 2024-04-05. Postings for (2023-2024): As of 2024.04 Visiting Assistant Professor- Educational Research, University of Southern Mississippi, The Hattiesburg, MS As of 2024.03 Assistant Professor in Evaluation (tenure track). Louisville, KY. University of Louisville. Postdoc, Research and Evaluation Methodology. Gainesville, FL. University of Florida. As of 2023.12 Associate/Full Professor of Quantitative Research Methods, University of Texas at Arlington in Arlington, TX As of 2023.

CONTINUE READING

Last updated on 4 min read regression

Random notes. Regression based techniques often involve finding a maximum (e.g., the maximum likelihood) or a minimum (e.g., least squares or mean square error) value. Gradient descent is an iterative optimization algorithm used to find the minimum of a function (or gradient ascent to find the maximum). The algorithm for solving for \(\theta_j\) looks like: \[\theta_j = \theta_j - \alpha\frac{\partial{}}{\partial{\theta_j} }J(\theta)\] \(\alpha\) is the learning rate (smaller step size takes more iterations)

CONTINUE READING

Last updated on 8 min read residuals

(MLM notes). Residuals are often used for model diagnostics or for spotting outliers in the data. For single-level models, these are merely the observed - predicted data (i.e., \(y_i - \hat{y}_i\)). However, for multilevel models, these are a bit more complicated to compute. Since we have two error terms (in this example), we will have two sets of residuals. For example, a multilevel model with a single predictor at level one can be written as:

CONTINUE READING

Projects

All School climate Youth violence project Grants

Ongoing Grant Funded Work

Various ongoing grants funded by the National Institute of Justice, Department of Education (i3), and Institute of Education Sciences.

Missouri Prevention Science Institute

The Missouri Prevention Science Institute (MPSI) brings community members and researchers together to help schools and families apply techniques that promote social and academic success. Through community outreach, the institute’s staff provides parent training and teacher consultation services.

Youth Violence Project @ UVA

Our team of faculty and graduate students conducts research on effective methods and policies for youth violence prevention and school safety.

Teaching

I teach the following courses at the University of Missouri:

Multilevel Modeling

The goal of the course is to provide students with the necessary skills needed to review/critique, analyze, interpret, and write-up studies involving nested (clustered) data using multilevel modeling (MLM). Clustered data (e.g., students within schools, patients within clinics) occur quite naturally in the social sciences and being able to understand and conduct their own analyses using nested data is an important skill. Alternatives are discussed as well.

Applied Multivariate Statistics

This course is designed to provide students with both a theoretical and applied understanding of useful multivariate statistical procedures (e.g., factor analysis, principal components analysis, discriminant function analysis, cluster analysis, MANOVA) in education sciences.

Program (Impact) Evaluation

Evaluating the quantifiable impact of social programs is a key task that policy makers, governments, and program funders perform. In education and the social sciences, a fundamental question asked is ‘How do we know our policy or program works?’

Quantitative Foundations in Educational Research

This course is designed to provide students the fundamental and necessary quantitative methods in educational research.

Data Management (using R)

Good data management is a prerequisite for successful research, needed for reproducibility of results, and essential when collaborating with others.

Tags

Contact