Could not download file: This paper is available to authorised users only.


R. Stock, A. Hiemisch

Ernst-Moritz-Arndt Universität Greifswald (GERMANY)
Statistics as a research tool is taught in university courses across a variety of disciplines. Consequently, instructors are confronted with diverse aptitudes and attitudes of learners. This raises a demand for educational settings which customize learning processes so that students can learn at their own speed with materials tailored to fit their current knowledge.

Traditional courses where learning activities and materials are structured by the instructor often demand either too much or too little from students, thus impeding the development of motivation for and interest in the subject of statistics. One step towards a higher individualization are blended-learning formats in which class lectures are supplemented by e-learning material. It is, however, not enough to provide e-learning materials for different knowledge levels. Additionally, the classroom must provide opportunities for individual support and feedback.

In a randomized control group design (N = 22 – 39) we compared a traditional structured advanced-level statistics course in psychology with a semi-structured blended-learning course. Different from blended learning settings in which the e-learning component takes place outside the classroom, at least half the time of the class was dedicated to working with e-materials. The course provided a pool e-learning materials with varying levels of difficulty, from which students could choose. Furthermore, they decided between different learning activities, e.g. whether they worked alone or in small groups. Whenever necessary, the lecturer answered questions, provided assistance and gave feedback.

Indicators of subjective Learning experiences (cognitive overload, workload, intrinsic motivation, current self-perceived statistics abilities) were measured at three points of time during the semester. The students also worked on a final test for knowledge and statistical competence at the end of the course.

While there emerged a small to medium size effect for higher cognitive overload in the traditional course (d = 0.20 – 0.67), the perceived workload was at most times slightly greater in the blended-learning course (d = 0.34 – 0.43). This may be due to the additional effort in self-regulation necessary to structure one’s own learning in the blended course. Intrinsic motivation was higher in the blended learning course at all three measurement points (d = 0.60 – 1.07) as were the self-perceived abilities of the learners even though to a smaller degree (d = 0.25 – 0.48). In the final test, not only a difference of knowledge (d = 0.32) but also of competences (d = 0.66) emerged with both being higher in the blended-learning course.

University courses supported by the use of e-learning materials in the classroom seem to foster individualized learning processes, thereby resulting in higher intrinsic motivation and less cognitive overload (even though the workload is perceived high). This effect was not limited to subjective variables, but was also found for the objective performance measures, specifically statistical competence.