Article
Portuguese
ID: <
10670/1.dnc6qk>
Abstract
This article is part of a broad suitability assessment study in the use of statistical techniques multivarised in theses and disertations of two higher education institutions in the field of marketing on consumer behaviour between 1997 and 2006. Eleven multivariate techniques (regression analysis, discriminatory analysis, logistic regression analysis, canonical correlation, multivariate variance analysis, joint analysis, structural equation modelling, factor analysis, conglomerate analysis, matching analysis, multidimensional scaling) are targeted in this article and have shown great potential for use in marketing studies. The aim of the reported work was to analyse the appropriateness of the employment of these techniques to the needs of the research problems presented in the thesis and dissertations and also to assess the level of adequacy in meeting their assumptions. In general, the results suggest that there is a need for an increase in the commitment of researchers to the verification of all theoretical precepts for the application of multivariate techniques.