ERIC Number: ED607605
Record Type: Non-Journal
Publication Date: 2020-Aug-24
Pages: 13
Abstractor: As Provided
ISBN: N/A
ISSN: EISSN-
EISSN: N/A
Available Date: N/A
Comparing Gamma and Log-Normal GLMs in R Using Simulation and Real Data Set
Mohammad, Nagham; McGivern, Lucinda
Online Submission
In regression analysis courses, there are many settings in which the response variable under study is continuous, strictly positive, and right skew. This type of response variable does not adhere to the normality assumptions underlying the traditional linear regression model, and accordingly may be analyzed using a generalized linear model assuming either a lognormal or gamma distribution. As such, students oftentimes become confused about which of these two distributions should be chosen to model a given dataset. In this article, we argue that the comparability of these two models should be taught through both simulation and real data analysis. Students will learn to identify the cases in which these two models can be used somewhat interchangeably through this teaching methodology.
Publication Type: Reports - Descriptive
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Grant or Contract Numbers: N/A
Author Affiliations: N/A