Multilevel Analysis in SPSS Summer School


Summer School

Aims

Multilevel analysis, also known as hierarchical linear modelling or random coefficient modelling, is an increasingly common practice in organisational research, largely due to the nested nature of many data sets (e.g. individuals within teams or organisations). Whereas a few years ago this was only possible to perform in specialist software such as MLwiN or HLM, it is now included in many standard statistical packages, including SPSS.

Day 1 of the workshop will include a general introduction to multilevel analysis, and a guide to conducting basic multilevel analysis in SPSS. The introduction will present the rationale for multilevel modelling, including how to make decisions about when it is or is not appropriate, and suggest a commonly used strategy for employing the analysis to practical situations. The SPSS section will be an opportunity to learn how to develop, conduct and interpret multilevel analysis (including cross-level and moderated effects) using SPSS, including both worked examples and exercises for participants.

Day 2 will extend the learning from day 1 to cover longitudinal models and three-level models. Each section will again start with an introduction to the concepts, and will be followed by examples of how to perform such analysis in SPSS, with exercises for participants. The workshop will conclude with a discussion of other relevant statistical issues, including sample sizes, justification of aggregation, other software, and a general question and answer session.

Information

Topics covered during the workshop include

- Basic principles of multilevel analysis
- Running two-level models in SPSS, including cross-level analysis and interpretation of cross-level interactions
- Data manipulation necessary for multilevel analysis (including aggregation and data file merging)
- Multilevel analysis of longitudinal data
- An introduction to three-level models
- Inter-rater reliability and agreement
- Power and sample size

No prior knowledge of multilevel analysis will be necessary, although familiarity with regression analysis will be useful. Prior experience of SPSS will be assumed, and as the workshop will primarily use SPSS syntax for conducting analysis, some experience of this would be helpful too, although not essential.

Assessment

Exercises during the workshop.

Materials

Gavin, M. B., & Hofmann, D. A. (2002). Using hierarchical linear modeling to investigate the moderating influence of leadership climate. The Leadership Quarterly, 13(1), 15-33.

Johnson, R. E., Rosen, C. C., & Chang, C. H. (2011). To aggregate or not to aggregate: Steps for developing and validating higher-order multidimensional constructs. Journal of Business and Psychology, 26(3), 241-248.

Additional info

PhD candidates, Research Master students and faculty members of ERIM can register for this course via SIN Online Registration.

External (non-ERIM) participants are welcome to this course. To register, please fill in the registration form and e-mail it to miizuka@rsm.nl by 3 weeks prior to the start of the course. Please note that the number of places for this course is limited.

This course is free of charge for ERIM members (faculty members, PhD candidates and Research Master students). For external participants, the course fee is 250 euro per ECTS credit.