Study level: Graduate study programme in psychology
Lecturers: Vesna Buško, Ph.D., Blaž Rebernjak, Ph.D.
ECTS: 5
Language: Croatian
Semester: 1st or 3rd (winter)
Status: elective
Form of instruction with class hours: 30 hours of lectures and 30 hours of exercises
Prerequisites: none
Student evaluation: writen (essays) - evaluation is done during the semester through data analysis during exercises and essays
Course description
Main content of the subject includes a study of issues regarding the application of selected group of methods for multivariate data analysis particularly aimed at the analysis of multivariate dependencies among and between the sets of metric and nonmetric variables. A selection of these analyses includes discriminant and multiple regression analysis, logistic regression, canonical correlational analysis, multivariate analysis of variance. The program is regularly limited to 3 methods that have been systematically studied and applied on empirical data. Each topic ends with written report where students, answering to general questions as well as to those related to actual data sets and the analyses performed, have to show that they have mastered the most important issues in application and quantitative interpretation.
Course objectives
The course aims at making students capable of ungoverned work in selection, evaluation of adequacy, and technical implementation of selected methods of multivariate data analysis, as well as in quantitative interpretation of the results obtained by use of these methods.
Required readings
Tacq, J. (1997). Multivariate Analysis Techniques in Social Science Research. London: Sage publications.
Klecka, W. R. (1980). Discriminant Analysis. Beverly Hills: Sage publications.
Norusis, M. J. (1993). SPSS for Windows - Professional Statistics. Chicago: SPSS Inc.
Recommended readings
Grimm, L.G., Yarnold, P.R. (Eds.) (1995). Reading and Understanding Multivariate Statistics. American Psyhological Association., Washington.
Jaccard, J. et.al. (2002). Interaction effects in multiple regression. London: Sage.