EJMT Abstract
Title |
A web-based application for the visual exploration of regression outliers, high leverage points, and influential points |
Author |
ARRAY(0x1705ff44cc8) |
Volume |
19 |
Number |
1 |
Regression outliers, high leverage points, and influential observations are commonly taught in statistics courses in which regression methods are covered – from the introductory to the advanced levels. When covering these topics, it is typical to discuss various methods that exist for determining whether a point is one of these types – such as Cook’s distance or DFFITS. Yet while these topics are routinely covered, it can be difficult for instructors to graphically illustrate whether observations are outliers, high leverage points, or influential observations based on the methods and their associated criteria. To address this shortcoming, in this article, a free, web-based app created by the author, which focuses on influential and related observations, is described, with intended usage by both faculty and students. Faculty feedback, which highlights the accessibility, ease of use, and usefulness of the app, is also discussed.