Visualization of 'Influence' in Regression
  • Intro
  • Plots
  • About

\(\bullet\) How Do we Evaluate Outliers and Influential Cases?
\(\bullet\) An interactive visualization based on Fox, J. (2016), Applied Regression Analysis and Generalized Linear Models, Sage, chapter 11
\(\bullet\) Adapted from R code developed by by James Steiger (http://www.statpower.net/R2101/Leverage.R)


Shiny app by Bruce Dudek

Adjust Y and X values of the Red data point and Examine Effects on the Influence Statistics




This simulation is based on an interactive visualization approach and R code developed by James Steiger. His original approach used the manipulate package rather than shiny

Tools for Statistics Instruction using R and Shiny

========================================================

Author: Bruce Dudek at the University at Albany

Using a simple regression simulated data set, this app provides an introductory visualization for the concepts related to leverage and influence. It also permits distinction between univariate outliers and multidimensional outliers relative to the fit of the model.

The approach follows the initial logic laid out by Fox in chapter 11 of his regression text.
(Fox, J. (2016), Applied Regression Analysis and Generalized Linear Models, Sage)

The app is modeled on an approach to the visualizaton created by by James Steiger. His example uses a different approach to interactivity in R, but the core plotting code here is heavily based on his code. (http://www.statpower.net/R2101/Leverage.R)

Built using Shiny by Rstudio and R, the Statistical Programming Language.

Ver 1.0, Feb 4, 2016