On the use of structural equation modeling in marketing image research

Purpose – The purpose of this paper is to assess the use of structural equation modeling in one specific field of marketing research, the image research. Design/methodology/approach – A meta-analysis of a sample of image marketing works using structural equation modeling (SEM). The period of investigation is limited to the last five years to test for possible positive return of previous assessments of SEM use on the current SEM application. Findings – Following this work, three major conclusions emerged: the study of homogenous samples of SEM models is required to get to accurate assessment of using the technique; SEM application is getting better probably due to learning from SEM reviews; and the reliance on a conjoint assessment of the various SEM issues is necessary to avoid parsimonious assessments. This study has provided a concise and refreshed view on the use of SEM in one marketing field, the image research. Research limitations/implications – 47 SEM papers and 99 models along five years were examined through this research. Although the authors reviewed four of the most consulted databases in marketing, the authors might miss several interesting works not available in these databases during the investigation. It is interesting to add on the works reviewed in this study and to re-conduct the analysis. The objective is not to doubt the consistency of SEM image research but to provide writers and readers with tools that enable them to produce better quality SEM research. Moreover, the quantitative analysis could be larger. Future research can consider computing other statistics. Finally, in the standards of most of marketing journals, this paper is a bit long. But as suggested by Babin et al. , journal editors should allow more space to SEM-based reviews as the nature of the discussion requires lengthening. Practical implications – Mastering the statistical tool in marketing research is as important as mastering the conceptual tool. Statistical learning and/or cooperation with statisticians is recommended. Originality/value – A multi-criteria review of works from one specific field in marketing research and across a recent period of time allowing for the test of possible positive return from previous reviews of SEM use on the quality of the current publications of SEM papers.