Solving problems on your own: how do exercises in tutorials interact with software learners' level of goal-orientedness?

Acquiring procedural knowledge cannot be done by simply reading about a new task domain; learners need to act within this domain and to interpret the results of their actions. One way to stimulate learners to perform fruitful actions is providing them with exercises. In this study, we contrasted two types of exercises for learners of a spreadsheet program: problem-solving exercises, providing learners with a well-defined goal, and (minimalist) on-your-own-exercises, inviting learners to explore options independently. On the basis of prior research, we expected exercise type to interact with the level of goal orientation. We assumed that learners with an ill-defined goal would profit more from problem-solving exercises, while learners with a better-defined goal would profit more from on-your-own sections. In order to test this hypothesis, we asked 60 people to become acquainted with the spreadsheet program Quattro Pro. Half of the participants worked with a (minimalist) tutorial that contained problem-solving exercises; the other half worked with a tutorial that contained on your own-sections. Goal orientation was operationalized by manipulating the task instruction the participants received before the training. Half of them received an open task instruction. This instruction asked them to imagine that they needed to get familiar with Quattro Pro in view of their activities as a committee member of a gymnastic club. The other half received a more specified instruction. These participants were asked to imagine that they needed to get familiar with Quattro Pro in order to be able to make the league tables for their gymnastic club. Furthermore, this task instruction specified the requirements these league tables should meet. After the training, the participants performed two tests: a skill test (consisting of some problem-solving tasks and an open task) and a knowledge test. About one week later, they performed these tests again. During the training and skill tests, participants were asked to think aloud. The results of the knowledge and skill tests confirmed our hypothesis. The two conditions that were expected to be beneficial (open task instruction + problem-solving exercises and specified task instruction + on-your-own sections) performed better than the conditions that were expected to be unhelpful (open task instruction + on-your-own sections and specified task instruction + problem-solving exercises). In order to achieve a better understanding of both exercise types, we also analyzed the participants' learning processes, especially the stages in which they were confronted with exercises. These analyses do not reveal clear explanations, but they do offer interesting ideas for further research.