panelr

Provides an object type and associated tools for storing and wrangling panel data. Implements several methods for creating regression models that take advantage of the unique aspects of panel data. Among other capabilities, automates the "within-between" (also known as "between-within" and "hybrid") panel regression specification that combines the desirable aspects of both fixed effects and random effects econometric models and fits them as multilevel models (Allison, 2009 doi:10.4135/9781412993869.d33; Bell & Jones, 2015 doi:10.1017/psrm.2014.7). These models can also be estimated via generalized estimating equations (GEE; McNeish, 2019 doi:10.1080/00273171.2019.1602504) and Bayesian estimation is (optionally) supported via 'Stan'. Supports estimation of asymmetric effects models via first differences (Allison, 2019 doi:10.1177/2378023119826441) as well as a generalized linear model extension thereof using GEE.