Session Information
09 SES 01 B, Findings from Large-scale Assessments
Paper Session
Contribution
With a growing number of longitudinal studies also new methods are used to analyze the development of a lot of different characteristics and their relationship to other characteristics. One group of these methods are Growth Curve Models (GCM). But like many other methods GCM identify only one central tendency for the whole sample to describe the mean development of the analyzed characteristic(s). Sex is one of the few attributes that are used quite regularly as a grouping variable to look for different developments by analyzing males and females separately to identify gender differences (even if there is no theoretical framework what kind of differences can be expected concerning this attribute). Often, gender differences found by an approach like this are not very substantial while their importance at the same time seems to be overestimated. And this may even strengthen gender stereotypes (Hyde, 2005).
In contrast, using Mixture Models gives us the possibility to switch our focus by first identifying groups that are defined by their different developmental trajectories of the interesting characteristic. Thus, also the different developments themselves can be identified. And probable differences between males and females for example, then will become noticeable by their different proportions in these groups what also may be a more realistic view to gender differences and similarities. A lot of other aspects can be analyzed as well particularly with respect to the identified groups.
In this study, the development of German students' physically violent behavior from fifth to ninth grade is analyzed. Earlier studies have shown that only a small minority of students behave like this regularly (once per month or more often; Eder, 2012). Thus, analyzing a mean tendency of this behavior over all students will be dominated by the majority of students that nearly never show this behavior. And even though male students are expected to show a higher tendency towards physically violent behavior than female students (Euler, 2009), both tendencies should be quite low and quite similar.
The actual aim of this study is to show that the use of Mixture Models is an appropriate way to identify quite different development trajectories that would otherwise be ignored and that seem to enable a realistic representation of the varying students' behavior. For this, the developmental trajectories that are identified by traditional GCM and by Growth Mixture Models (GMM, Kreuter & Muthen, 2008) will be compared. The differences of these results will be discussed also concerning expected gender differences (and similarities). Besides this, also challenges, restrictions, and uncertainties of Mixture Models will be discussed.
Method
Expected Outcomes
References
Eder, F. (2012). PISA-2009. Zusatzanalysen für Österreich. Präsentation zentraler Ergebnisse, 19.6.2012, Universität Salzburg. Euler, H. A. (2009). Geschlechterspezifische Unterschiede und die nicht erzählte Geschichte in der Gewaltforschung. In H. G. Holtappels, W. Heitmeyer, W. Melzer & K.-J. Tillmann (Eds.) Forschung über Gewalt an Schulen. Erscheinungsformen und Ursachen, Konzepte und Prävention (pp. 191-206). Weinheim: Juventa Verlag. Hyde, J. S. (2005). The Gender Similarities Hypothesis. American Psychologist, 60, 581–592. Kreuter, F. & Muthen, B. (2008). Longitudinal Modeling of Population Heterogeneity. Methodological Challenges to the Analysis of Empirically Derived Criminal Trajectory Profiles. In G. R. Hancock & K.M. Samuelsen (Eds.) Advances in Latent Variable Mixture Models (pp. 53-75). Charlotte, NC: Information Age Publishing, Inc.
Search the ECER Programme
- Search for keywords and phrases in "Text Search"
- Restrict in which part of the abstracts to search in "Where to search"
- Search for authors and in the respective field.
- For planning your conference attendance you may want to use the conference app, which will be issued some weeks before the conference
- If you are a session chair, best look up your chairing duties in the conference system (Conftool) or the app.