Software

Online Applications

Applications currently listed here are related to research, but some may be primarily teaching tools. Please also see the teaching page for applications that are not directly related to research.

Software Packages

Title Subtitle Author Description
detranli DEtection of RANdom LIkert-type responses Michael John Ilagan, Carl F. Falk, & Wenshuo Li R package for detecting full (i.e. all items) content-nonresponsivity (e.g., random and patterned responses) in Likert-type data. Initially based on work by Ilagan & Falk (2024).
EMgaussian Expectation-Maximization algorithm for multivariate normal (Gaussian) with missing data Carl F. Falk R package for estimating means and covariances or regularized network models using glasso under high rates of missing data. A result of Falk & Starr (2025) and much of the heavy lifting in C++.
multilevelmediation Utility functions for multilevel mediation analysis in R. Carl F. Falk, Todd Vogel, Sarah Hammami, & Milica Miočević R package for classic 1-1-1 multilevel mediation analysis model with random effects. A result of Falk, Vogel, Hammami, & Miočević (2024).
mpirt Functions for estimating monotonic polynomial item response models using rpf and OpenMx. Carl F. Falk R package developed to help making estimation of flexible item response models easier to do. A result of Falk & Fisher (2022) but based on prior work by Falk & Cai (2016), Falk (2020), and infrastructure by Pritikin & Falk (2020). Includes simulated annealing to select polynomial order from Falk (2019). Still a work in progress.
ockhamSEM Tools for studying fit propensity in structural equation modeling. Carl F. Falk & Michael Muthukrishna R package developed to make investigating fitting propensity (and model complexity) easier to do. Detailed in Falk & Muthurkishna (2023; equal first-authorship)
pln Polytomous logit-normit (graded logistic) model estimation. Carl F. Falk & Harry Joe R package developed based on C code from Harry Joe to estimate unidimensional item response models using a bivariate composite likelihood (or full information maximum likelihood) approach and compute a limited information test statistic.
No matching items

Code Contributions or Forks

Title Subtitle Description
lavaan Latent variable analysis. Forked version of the package on GitHub that can estimate (robust) likelihood-based confidence intervals for structural equation models. These intervals were initially studied by Falk (2018) and later with this package by Falk & Chen (2026).
difR Collection of Methods to Detect Dichotomous and Polytomous Differential Item Functioning (DIF). Authored some code for new functions for the 6.0 version of the R package.
mirt Multidimensional item response theory. Contributed C++ code providing second-order derivatives for the constrained version of the multidimensional nominal response model described in Falk & Cai (2016) and Falk & Ju (2020) using the Metropolis-Hastings Robbins-Monro algorithm of Cai (2010).
rpf Response probability functions. Contributed R and C++ code to allow estimation of monotonic polynomial models (Falk, 2020; Falk & Cai, 2016; Pritikin & Falk, 2020), and a tutorial explaining how to add additional item models to the package.
No matching items

Code Snippets

Many/most publications include custom implementation of some statistical method or estimated model. Some more recent publications, some trainee led, have such code available. These generally appear on the Publications page as links directly to the code or some OSF repositories.