A multivariate method for analyzing and improving the use of student evaluation of teaching questionnaires: a case study

• Published in: Quality & quantity Vol. 45; no. 6; pp. 1415 - 1427
• Main Authors: Martínez-Gómez, Mónica, Sierra, Jose Miguel Carot, Jabaloyes, José, Zarzo, Manuel
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.10.2011; Springer; Springer Nature B.V
• Subjects: Case studies; Classification; Cluster analysis; College Students; Cronbach's alpha; Datasets; Education; Educational evaluation; Europe; Evaluation; Higher education; History, theory and methodology; Methodology; Methodology of the Social Sciences; Questionnaires; Reliability; Social Sciences; Sociology; Statistical methods; Structural equation modeling; Student Evaluation; Students; Teacher evaluations; Teaching; Teaching methods; Universities; Validity; Europe; Cluster analysis; Confirmatory factor analysis; Student questionnaire for teaching assessment; CHAID; Quality indicator; Case study; Multivariate Analysis; Evaluation; Questionnaire; Methodology; University; Interaction; Student
• ISSN: 0033-5177; 1573-7845
• DOI: 10.1007/s11135-010-9345-5

Student evaluation of teaching (SET) questionnaires are the most common methods of evaluation used by European universities to assess the quality of teaching delivered by their lecturers. A series of multivariate statistical methods were applied to analyze the underlying structure of the SET questionnaire used by the Universidad Politecnica de Valencia (UPV) in order to develop an appropriate methodology for extracting, analyzing, and interpreting the information contained in the questionnaire. In a first step, a confirmatory factorial analysis (CFA) was developed in order to evaluate the reliability, validity and dimensionality of it, by means of two relatively new parameters commonly used in structural equation modelling: the compound reliability and extracted variance for each latent construct. In a second step, cluster analysis (CA) was used to test the ability of the questionnaire for the identification of different categories of lecturers. In the last step, a tree classification method, the chi-squared automatic interaction detector (CHAID), was used in order to characterize the different lecturer’s categories obtained with CA according to all available information regarding the teaching staff and subjects.