This longitudinal study examined the relation between continuity and change in the Big Five personality traits and life events. Approximately 2,000 German students were tracked from high school to university or to vocational training or work, with 3 assessments over 4 years. Life events were reported retrospectively at the 2nd and 3rd assessment. Latent curve analyses were used to assess change in personality traits, revealing 3 main findings. First, mean-level changes in the Big Five factors over the 4 years were in line with the maturity principle, indicating increasing psychological maturity from adolescence to young adulthood. Second, personality development was characterized by substantive individual differences relating to the life path followed; participants on a more vocationally oriented path showed higher increases in conscientiousness and lower increases in agreeableness than their peers at university. Third, initial level and change in the Big Five factors (especially Neuroticism and Extraversion) were linked to the occurrence of aggregated as well as single positive and negative life events. The analyses suggest that individual differences in personality development are associated with life transitions and individual life experiences.

In multilevel modeling, group-level variables (L2) for assessing contextual effects are frequently generated by aggregating variables from a lower level (L1). A major problem of contextual analyses in the social sciences is that there is no error-free measurement of constructs. In the present article, 2 types of error occurring in multilevel data when estimating contextual effects are distinguished: unreliability that is due to measurement error and unreliability that is due to sampling error. The fact that studies may or may not correct for these 2 types of error can be translated into a 2 × 2 taxonomy of multilevel latent contextual models comprising 4 approaches: an uncorrected approach, partial correction approaches correcting for either measurement or sampling error (but not both), and a full correction approach that adjusts for both sources of error. It is shown mathematically and with simulated data that the uncorrected and partial correction approaches can result in substantially biased estimates of contextual effects, depending on the number of L1 individuals per group, the number of groups, the intraclass correlation, the number of indicators, and the size of the factor loadings. However, the simulation study also shows that partial correction approaches can outperform full correction approaches when the data provide only limited information in terms of the L2 construct (i.e., small number of groups, low intraclass correlation). A real-data application from educational psychology is used to illustrate the different approaches.

The social relations model (SRM) is a conceptual, methodological, and analytical approach that is widely used to examine dyadic behaviors and interpersonal perception within groups. This article introduces a general and flexible approach to estimating the parameters of the SRM that is based on Bayesian methods using Markov chain Monte Carlo techniques. The Bayesian approach overcomes several statistical problems that have plagued SRM researchers. First, it provides a single unified approach to estimating SRM parameters that can be easily extended to more specialized models (e.g., measurement models, moderator variables, categorical outcome variables). Second, sampling-based Bayesian methods allow statistically reliable inferences to be made about variance components and correlations, even with small sample sizes. Third, the Bayesian approach is able to handle designs with missing data. In a simulation study, the statistical properties (bias, root-mean-square error, coverage rate) of the parameter estimates produced by the Bayesian approach are compared with those of the method of moment estimates that have been used in previous research. A data example is presented to illustrate how discrete person moderators can be included in SRM analyses using the Bayesian approach. Finally, further extensions of the SRM are discussed, and suggestions for applied research are made.

Zusammenfassung: Viele Fragestellungen der pädagogischen Psychologie beziehen sich auf Einflüsse von Variablen der Lernumwelt auf individuelle motivationale, affektive und kognitive Merkmale. Nicht selten werden dabei Schüler direkt nach Kontextmerkmalen befragt und anschließend werden die Angaben aller Schüler auf Klassen- oder Schulebene aggregiert um ein Maß für den Klassenkontext zu gewinnen. In weiteren statistischen Analysen - z.B. auf der Basis von Mehrebenenmodellen - werden die aggregierten Maße als potenzielle Prädiktoren auf Klassen- oder Schulebene behandelt. Im vorliegenden Beitrag wird formal und anhand einer Simulationsstudie gezeigt, dass die Regressionsgewichte solcher aggregierten Merkmale in Mehrebenenanalysen als bloßes statistisches Artefakt auftreten können. Die Stärke des Artefakts steigt mit der Unreliabilität der Messungen auf Individualebene und mit der Zunahme der Intraklassenkorrelation des betrachteten Merkmals. Ergänzend werden Daten aus einer Studie von Köller und Baumert (2001) verwendet, um post-hoc abzuschätzen, inwieweit die dort berichteten Kontexteffekte eventuell statistische Artefakte darstellen. Implikationen für zukünftige Studien zu Kontexteffekten werden diskutiert.