Review by: Brahm, Taiga (2009-06-08)

In their article, Krause et al. report on a study examining the effects of cooperative learning and a feedback intervention on the learning success of statistics students in an e-learning environment. The computer-based learning environment, named Koralle, is aimed at university students of the social sciences in an advanced statistics class. In terms of content, it deals with correlation analysis. Koralle is based on the principles of situated learning, i. e. the learners deal with complex problems relevant to them and content is embedded in a realistic context. Since the subject matter (statistics) is regarded as a well-structured field, the learning environment mainly uses worked examples which support the solution of statistical problems through step-by-step guides. In a first experimental study (Tyroller, 2005), it could be shown that students learning with Koralle had indeed received better grades, however, some of them also lacked deeper understanding. The authors identified the lack of cooperation in self-regulated e-learning such as Koralle as one possible reason. Another reason might be that especially students with little prior knowledge did not use the standardized feedback provided effectively. In consequence, the present study focused on the effects of cooperative learning and a feedback intervention. Cooperative learning is defined as "a setting where people learn together in a group that is small enough to allow active participation of each group member" (p 160). Since the students have to work cooperatively in front of a computer, dyads were used. Additionally, the authors decided for homogenous groups in order to assure that all students participated actively.Feedback can be implemented in different manners, i. e. it varies in form and degree of elaboration (feedback design). It also matters how the students deal with feedback (feedback reception). Among many other forms, feedback may be designed as knowledge of results, knowledge of the correct response and as elaborated feedback where elaborated feedback is seen as most effective. In e-learning, feedback can be provided automatically by the computer which is very economic and can be available for the student immediately. Automatic feedback can also take many forms, e. g. it can be standardized or adapted to the students' answers. It is also possible to provide feedback by a tutor, e. g. via e-mail or discussion boards). The study tested three hypotheses:

• Both interventions – cooperative learning and feedback – will promote objective learning outcomes.
• There will be an interaction effect of cooperative learning and feedback such that the feedback intervention is most successful for those students working cooperatively (group-feedback hypothesis).
• There will be an interaction effect of the feedback intervention and students prior knowledge such that feedback is especially beneficial for those students with little prior knowledge.
• Groups will be more successful in problem-solving than individuals.
• Both interventions will enhance students' perceived performance and perceived competence.

The study was conducted as a 2x2 factorial laboratory experiment with 137 students. It included a pre-/posttest design with the factors social context (individual vs. group) and feedback intervention (available or not). Students were randomly assigned to the four intervention groups. As mentioned above, for the cooperative setting, students were put together in homogeneous dyads based on their prior grades and prior knowledge reported by themselves. Data was gathered through questionnaire with established scales.

Results show that students with the feedback intervention scored higher than students without the feedback intervention, however, the main effect of social context was not significant. The group-feedback hypothesis could not be confirmed. Contrary to the second hypothesis stated above, the feedback intervention was more supported in the individual than the cooperative learning condition. In accordance with the hypothesis stated above, feedback was more beneficial for students with little prior knowledge. The feedback intervention also promoted objective learning outcomes, however, did not influence perceived learning outcomes and perceived competence.

The feedback intervention used in this experimental study was appropriate and induced learning gains among the students, especially with little prior knowledge. This effect was still in place after controlling for time-on-task, however, since the intervention included a multiple-choice test to generate the automated feedback it might be that the additional practice during the test might have improved the learning results as well. The question as to why the feedback intervention was useful remains. Further research should analyze feedback processes in more detail as to find out the mechanisms that lead to the beneficial effects of feedback as Krause et al. suggested. Cooperative learning did not improve the learning outcomes. The students were monitored during the study so that lack of participation and interaction can be excluded as a reason for this. Additional process analyses on the quality of the interaction could reveal further reasons. The study could not test the long-term effects of the interventions. Thus, it might also be that "elaborative and reflective processes in cooperation promote long-term retention and, thus, sustainability in learning" (p 168).

In sum, the study shows that adaptive elaborated feedback can support students when using e-learning environments. Process analyses and field studies would further support the effects found in the study.