Effects of different type of covariates and sample size on parameter estimation for multinomial logistic regression model

Hamzah Abdul Hamid, Yap Bee Wah, Xian Jin Xie

Research output: Contribution to journalArticle

Abstract

The sample size and distributions of covariate may affect many statistical modeling techniques. This paper investigates the effects of sample size and data distribution on parameter estimates for multinomial logistic regression. A simulation study was conducted for different distributions (symmetric normal, positively skewed, negatively skewed) for the continuous covariates. In addition, we simulate categorical covariates to investigate their effects on parameter estimation for the multinomial logistic regression model. The simulation results show that the effect of skewed and categorical covariate reduces as sample size increases. The parameter estimates for normal distribution covariate apparently are less affected by sample size. For multinomial logistic regression model with a single covariate study, a sample size of at least 300 is required to obtain unbiased estimates when the covariate is positively skewed or is a categorical covariate. A much larger sample size is required when covariates are negatively skewed.

Original languageEnglish (US)
Pages (from-to)155-161
Number of pages7
JournalJurnal Teknologi
Volume78
Issue number12-3
DOIs
StatePublished - 2016

Fingerprint

Parameter estimation
Logistics
Normal distribution

Keywords

  • Multinomial logistic regression
  • Parameter estimation
  • Simulation
  • Skewed covariate

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Effects of different type of covariates and sample size on parameter estimation for multinomial logistic regression model. / Hamid, Hamzah Abdul; Wah, Yap Bee; Xie, Xian Jin.

In: Jurnal Teknologi, Vol. 78, No. 12-3, 2016, p. 155-161.

Research output: Contribution to journalArticle

@article{c5a5eaeba98f433897c30409f74c3010,
title = "Effects of different type of covariates and sample size on parameter estimation for multinomial logistic regression model",
abstract = "The sample size and distributions of covariate may affect many statistical modeling techniques. This paper investigates the effects of sample size and data distribution on parameter estimates for multinomial logistic regression. A simulation study was conducted for different distributions (symmetric normal, positively skewed, negatively skewed) for the continuous covariates. In addition, we simulate categorical covariates to investigate their effects on parameter estimation for the multinomial logistic regression model. The simulation results show that the effect of skewed and categorical covariate reduces as sample size increases. The parameter estimates for normal distribution covariate apparently are less affected by sample size. For multinomial logistic regression model with a single covariate study, a sample size of at least 300 is required to obtain unbiased estimates when the covariate is positively skewed or is a categorical covariate. A much larger sample size is required when covariates are negatively skewed.",
keywords = "Multinomial logistic regression, Parameter estimation, Simulation, Skewed covariate",
author = "Hamid, {Hamzah Abdul} and Wah, {Yap Bee} and Xie, {Xian Jin}",
year = "2016",
doi = "10.11113/jt.v78.10036",
language = "English (US)",
volume = "78",
pages = "155--161",
journal = "Jurnal Teknologi (Sciences and Engineering)",
issn = "0127-9696",
publisher = "Penerbit Universiti Teknologi Malaysia",
number = "12-3",

}

TY - JOUR

T1 - Effects of different type of covariates and sample size on parameter estimation for multinomial logistic regression model

AU - Hamid, Hamzah Abdul

AU - Wah, Yap Bee

AU - Xie, Xian Jin

PY - 2016

Y1 - 2016

N2 - The sample size and distributions of covariate may affect many statistical modeling techniques. This paper investigates the effects of sample size and data distribution on parameter estimates for multinomial logistic regression. A simulation study was conducted for different distributions (symmetric normal, positively skewed, negatively skewed) for the continuous covariates. In addition, we simulate categorical covariates to investigate their effects on parameter estimation for the multinomial logistic regression model. The simulation results show that the effect of skewed and categorical covariate reduces as sample size increases. The parameter estimates for normal distribution covariate apparently are less affected by sample size. For multinomial logistic regression model with a single covariate study, a sample size of at least 300 is required to obtain unbiased estimates when the covariate is positively skewed or is a categorical covariate. A much larger sample size is required when covariates are negatively skewed.

AB - The sample size and distributions of covariate may affect many statistical modeling techniques. This paper investigates the effects of sample size and data distribution on parameter estimates for multinomial logistic regression. A simulation study was conducted for different distributions (symmetric normal, positively skewed, negatively skewed) for the continuous covariates. In addition, we simulate categorical covariates to investigate their effects on parameter estimation for the multinomial logistic regression model. The simulation results show that the effect of skewed and categorical covariate reduces as sample size increases. The parameter estimates for normal distribution covariate apparently are less affected by sample size. For multinomial logistic regression model with a single covariate study, a sample size of at least 300 is required to obtain unbiased estimates when the covariate is positively skewed or is a categorical covariate. A much larger sample size is required when covariates are negatively skewed.

KW - Multinomial logistic regression

KW - Parameter estimation

KW - Simulation

KW - Skewed covariate

UR - http://www.scopus.com/inward/record.url?scp=85006364838&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006364838&partnerID=8YFLogxK

U2 - 10.11113/jt.v78.10036

DO - 10.11113/jt.v78.10036

M3 - Article

VL - 78

SP - 155

EP - 161

JO - Jurnal Teknologi (Sciences and Engineering)

JF - Jurnal Teknologi (Sciences and Engineering)

SN - 0127-9696

IS - 12-3

ER -