Growth Mixture Models Outperform Simpler Clustering Algorithms When Detecting Longitudinal Heterogeneity,Even With Small Sample Sizes |
| |
| Authors: | Daniel P Martin Timo von Oertzen |
| |
| Institution: | 1. University of Virginiadm4zz@virginia.edu;3. University of Virginia |
| |
| Abstract: | The expansion in the number of people entering higher education has resulted in a substantial increase in the proportion of students enrolling in nontraditional modes, such as part-time study. This study examined the question of whether part-time study curtails the development of the types of intellectual capabilities needed for a knowledge-based economy, because the students would have markedly less exposure to a stimulating campus environment than their full-time counterparts. Graduates from discrete full- and part-time programs from 1 university in Hong Kong completed a survey seeking perceptions of the nurturing of a range of capabilities, together with measures of teacher-student relationships and type of teaching experienced. Two hypotheses were tested by structural equation modeling: (a) the same mechanism for capability development operated for full- and part-time modes and (b) the principal element of the mechanism was the nature of teaching and the quality of teacher-student interaction. Hypothesis 1 was supported because configural invariance between hypothesized models for capability development between the 2 modes was found. Hypothesis 2 was also supported because the models showed that the principal influence on capability development came from teaching for understanding, through promoting active learning experiences and the degree and quality of teacher-student interaction. |
| |
| Keywords: | comparative simulation growth heterogeneity longitudinal clustering |
|
