Model comparisons

Описание: In this video, the focus is on the use of multiple regression analysis in research papers, specifically using Hekman's paper as an example. The paper utilizes hierarchical moderated regression analysis, where multiple models are estimated, starting with a simple one and progressively adding more variables for comparison.

The video delves into model comparisons, particularly between two models. Model 1 includes only control variables, while Model 2 incorporates both control and interesting variables. The comparison aims to determine if the combination of control and interesting variables better explains the dependent variable than control variables alone. This is assessed using the R-squared statistic, including both the adjusted and actual R-squared. The adjusted R-squared is preferred for small samples, but since it lacks a well-known test, the R-squared difference is tested instead. The video explains the use of an F test to evaluate the null hypothesis that all regression coefficients for variables added to the model are zero. This F test is a nested model comparison test, where one model is a special case of another. The video concludes by highlighting the importance of understanding these calculations and the logic behind them, even though software typically performs these computations in practice.

Slides: https://osf.io/mw35h