Statistical Methods for Multivariate Group Differences
Code: 53837
ECTS: 3.0
Lecturers in charge: izv. prof. dr. sc. Toni Babarović
Take exam: Studomat
English level:

1,0,0

All teaching activities will be held in Croatian. However, foreign students in mixed groups will have the opportunity to attend additional office hours with the lecturer and teaching assistants in English to help master the course materials. Additionally, the lecturer will refer foreign students to the corresponding literature in English, as well as give them the possibility of taking the associated exams in English.
Load:

1. komponenta

Lecture typeTotal
Lectures 15
Seminar 15
* Load is given in academic hour (1 academic hour = 45 minutes)
Description:
Course description
Understanding of the theoretical basis of the MANOVA models, discriminant, canonical and cluster analysis; analysing and evaluating research papers in which these multivariate methods were applied; creating research designs where these methods should be used as an adequate methods of analysing the collected data; independent practical application of MANOVA models, discriminant, canonical and cluster analysis in psychological research using the SPSS software package

e-learning level 1
english level 1

Competency
Evaluate multivariate statistical procedures with regard to their limits and to satisfying the theoretical assumptions in concrete situations of application.
Design fundamental and or applied psychological research and construct and metrically evaluate psychological instruments.
Choose appropriate research methods and strategies for interventions in accordance with the characteristics of the members of different social groups and cultures as well as the specifics of their environmental and social context.

Learning Outcomes
1. Explain the statistical and mathematical logic of MANOVA, discriminant, canonical and cluster analysis
2. Evaluate and assess the statistical requirements for the implementation of these multivariate methods.
3. Design a research plan suitable for applying these multivariate methods
4. Assess and analyse the data suitable for MANOVA, discriminant, canonical and cluster analysis using the software package SPSS
5. Interpret the results obtained from one of these methods within the given research problem
6. Explain statistical parameters obtained from the analysis
7. Evaluate the quality of the data and results obtained by these multivariate analysis
8. Assess the range and limits of the statistical conclusions derived by factor MANOVA, discriminant, canonical and cluster analysis
9. Estimate the quality of the scientific interpretation of the results obtained these methods.


Week plan
1. Statistical logic of MANOVA - multivariate extension of ANOVA
2. Significance of multivariate F-test and interpretation of the MANOVA results
3. Using covariates -- MANCOVA
4. Logic of Discriminant analysis and analogy with MANOVA
5. Formation of discriminant functions; discriminative weights and discriminative loadings
6. Significance and interpretation of discriminant functions
7. Error estimates in group classification and validation of the results of discriminant analysis
8. Logic of Canonical analysis
9. Requirements for the canonical analysis
10. The formation of canonical functions; canonical weights, canonical loadings and cross-loading
11. Significance and interpretation of canonical functions
12. The relation between canonical analysis and other multivariate techniques
13. Cluster analysis - basic logic and methods of use
14. Methods of calculating the distance between the cases and the formation of clusters
15. Determining the number of clusters to retain


Grading
Activity in class (lectures and exercises) - 20%;
Seminal work - 30%;
Written exam - 50%.
Literature:
  1. Klecka, W. R. (1980) Discriminant Analysis, Sage 19, London.
    Harris, R. J. (1975) A Primer of Multivariate Analysis, Academic Press, New York.
    Overall J. E. & Klett C. J. (1972) Applied Multivariate Analysis, McGraw-Hill Book Inc. New York.
Optional literature:
  1. Hair, J.F, Anderson, R.E., Tatham, R.L. & Black, W.C., (1998) Multivariate Data Analysis, Prentice Hall, New Jersey.
3. semester
PSI-dipl. (5054): Izborni metodološki kolegiji - Mandatory studij - Psychology
Consultations schedule:
News

Informations

Hrvatski studiji Sveučilišta u Zagrebu pokrenuti su i ustrojeni 16. studenoga 1992., isprva samo kao dvosemestralni Sveučilišni komparativni studij hrvatske filozofije i društva. Taj je program potom preoblikovan u program redovitog četverogodišnjeg studija.

Address: Borongajska cesta 83d, Zagreb (map)
© 2013. - 2019. Sveučilište u Zagrebu Hrvatski studiji. Sva prava pridržana. Webmaster
powered by QuiltCMS