A tutorial on generalized linear models

Raymond H. Myers, Douglas Montgomery

Research output: Contribution to journalArticle

80 Citations (Scopus)

Abstract

Situations in which the observations are not normally distributed arise frequently in the quality engineering field. The standard approach to the analysis of such responses is to transform the response into a new quantity that behaves more like a normal random variable. An alternative approach is to use an analysis procedure based on the generalized linear model (GLM), where a nonnormal error distribution and a function that links the predictor to the response may be specified. We present an introduction to the GLM, and show how such models may be fit. We present the GLM as an analog to the normal theory linear model. The usefulness of this approach is illustrated with examples.

Original languageEnglish (US)
Pages (from-to)274-291
Number of pages18
JournalJournal of Quality Technology
Volume29
Issue number3
StatePublished - Jul 1997

Fingerprint

Generalized Linear Model
Link Function
Predictors
Linear Model
Random variable
Transform
Engineering
Analogue
Random variables
Alternatives
Tutorial
Generalized linear model
Model

Keywords

  • Data Analysis
  • Design of Experiments
  • Generalized Linear Models
  • Least Squares
  • Maximum Likelihood
  • Regression

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Industrial and Manufacturing Engineering
  • Statistics and Probability

Cite this

A tutorial on generalized linear models. / Myers, Raymond H.; Montgomery, Douglas.

In: Journal of Quality Technology, Vol. 29, No. 3, 07.1997, p. 274-291.

Research output: Contribution to journalArticle

Myers, Raymond H. ; Montgomery, Douglas. / A tutorial on generalized linear models. In: Journal of Quality Technology. 1997 ; Vol. 29, No. 3. pp. 274-291.
@article{9b252a8ecf1a4069a6c7b2a0537a670f,
title = "A tutorial on generalized linear models",
abstract = "Situations in which the observations are not normally distributed arise frequently in the quality engineering field. The standard approach to the analysis of such responses is to transform the response into a new quantity that behaves more like a normal random variable. An alternative approach is to use an analysis procedure based on the generalized linear model (GLM), where a nonnormal error distribution and a function that links the predictor to the response may be specified. We present an introduction to the GLM, and show how such models may be fit. We present the GLM as an analog to the normal theory linear model. The usefulness of this approach is illustrated with examples.",
keywords = "Data Analysis, Design of Experiments, Generalized Linear Models, Least Squares, Maximum Likelihood, Regression",
author = "Myers, {Raymond H.} and Douglas Montgomery",
year = "1997",
month = "7",
language = "English (US)",
volume = "29",
pages = "274--291",
journal = "Journal of Quality Technology",
issn = "0022-4065",
publisher = "American Society for Quality",
number = "3",

}

TY - JOUR

T1 - A tutorial on generalized linear models

AU - Myers, Raymond H.

AU - Montgomery, Douglas

PY - 1997/7

Y1 - 1997/7

N2 - Situations in which the observations are not normally distributed arise frequently in the quality engineering field. The standard approach to the analysis of such responses is to transform the response into a new quantity that behaves more like a normal random variable. An alternative approach is to use an analysis procedure based on the generalized linear model (GLM), where a nonnormal error distribution and a function that links the predictor to the response may be specified. We present an introduction to the GLM, and show how such models may be fit. We present the GLM as an analog to the normal theory linear model. The usefulness of this approach is illustrated with examples.

AB - Situations in which the observations are not normally distributed arise frequently in the quality engineering field. The standard approach to the analysis of such responses is to transform the response into a new quantity that behaves more like a normal random variable. An alternative approach is to use an analysis procedure based on the generalized linear model (GLM), where a nonnormal error distribution and a function that links the predictor to the response may be specified. We present an introduction to the GLM, and show how such models may be fit. We present the GLM as an analog to the normal theory linear model. The usefulness of this approach is illustrated with examples.

KW - Data Analysis

KW - Design of Experiments

KW - Generalized Linear Models

KW - Least Squares

KW - Maximum Likelihood

KW - Regression

UR - http://www.scopus.com/inward/record.url?scp=0031190088&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031190088&partnerID=8YFLogxK

M3 - Article

VL - 29

SP - 274

EP - 291

JO - Journal of Quality Technology

JF - Journal of Quality Technology

SN - 0022-4065

IS - 3

ER -