3. Results
3.1. Differences in the concurrent pursuit of the three social achievement goals
Descriptive statistics, bivariate correlations, and Cronbach alphas of the measured variables are presented in Table 1. We conducted a cluster analysis - a technique that classifies people who share common characteristics (i.e., social achievement goals, in our case) into the same group (Borgen & Barnett, 1987) - to investigate students' different profiles regarding their social achievement goals. Cluster analysis is sometimes called person-centered analysis because it tries to uncover howthe variables combine across individuals. After dropping 98 outliers (6.1% of the sample), according to the criteria proposed by Cohen, Cohen, West, and Aiken (2003) we followed Hair and Black's (2004) recommendations, and first ran a hierarchical cluster analysis (with Ward's method estimate) to determine the number of clusters that emerged. Inspection of the BIC index suggested a three-cluster solution. Using the centroids of the three-cluster solution, we then performed a kmeans clustering.