Median Distance Model for Likert-Type Items in Contingency Table Analysis

Altunay S. A., YILMAZ A. E.

Revstat Statistical Journal, vol.21, no.2, pp.267-281, 2023 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 21 Issue: 2
  • Publication Date: 2023
  • Doi Number: 10.57805/revstat.v21i2.401
  • Journal Name: Revstat Statistical Journal
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, zbMATH, Directory of Open Access Journals
  • Page Numbers: pp.267-281
  • Keywords: association models, contingency table, Likert-type items, log-linear models, ordinal variables
  • Hacettepe University Affiliated: Yes

Abstract

Likert-type items (questions) are a widely used scale in questionnaire design. The “neutral” or “undecided” option may lead to misinterpretation and confusion about the results. This paper proposes two novel log-linear models to measure how much accumulation of the neutral option over the contingency tables at any question levels. These models also test the odds that a respondent’s level how far from the median. These models will help the researchers how to incorporate the neutral option in conceptual frameworks.