Structural equation models (SEM and variables measured on the polychoric ordinal scale in the analysis of marketing data Cover Image

Modele równań strukturalnych a zmienne mierzone na skali porządkowej polichorycznej w analizie danych marketingowych
Structural equation models (SEM and variables measured on the polychoric ordinal scale in the analysis of marketing data

Author(s): Piotr Tarka
Subject(s): Economy, Marketing / Advertising
Published by: Wydawnictwo Uniwersytetu Ekonomicznego we Wrocławiu
Keywords: SEM model; polychoric ordinal Likert scale; estimation methods

Summary/Abstract: In the hereby article, the author describes the methodological aspects in reference to the items/observed variables being measured on the basis of polychoric ordinal Likert scale, which can be used in the construction process of the structural equation models (SEM). With this objective in mind, there was designed and conducted a survey on the basis of data gathered from 200 sample units. Due to the applied measurement level (i.e., 5-point scale) for the observable variables, and the assessment of the parameter estimates, as well as the size of standard errors and descriptive statistics (indicating the goodness-of-fit of data to the SEM model), there were included, in the comparative analysis, the following estimation methods: ML,WLS, WLSk, DWLSkTM, DWLSKtmv.

  • Issue Year: 2016
  • Issue No: 459
  • Page Range: 265-276
  • Page Count: 12
  • Language: Polish
Toggle Accessibility Mode