Supporting cognitive processing in multimedia learning: The use of implementation intentions

Abstract

This dissertation set out to answer the question whether implementation intentions can support cognitive processing during multimedia learning and improve learning. The use of multimedia (i.e., the combination of text and pictures) has been shown to positively affect learning outcomes (Mayer, 2003). This so called ‘multimedia effect’ depends, among other things, on learners’ knowledge and use of effective cognitive processes during multimedia learning. Yet, knowledge of effective cognitive processes does not necessarily lead learners to make use of them; accordingly, learners often fail to effectively use multimedia materials, resulting in suboptimal learning (e.g., Schmidt-Weigand et al., 2010b). The failure to use effective cognitive processes can be conceptualized as a problem of volition (Corno, 2001). One technique that has consistently proven itself helpful in overcoming volitional obstacles is the use of implementation intentions (Gollwitzer, 1999). Implementation intentions are specific if-then plans that strongly link a situational cue to an intended action, thus facilitating automatic action initiation during good opportunities to act.

Four experiments were conducted to investigate the effectiveness of implementation intentions in multimedia learning. All experiments shared a similar procedure: First, learners received either received either instructions to learn with implementation intentions about certain multimedia processes or some other instruction, followed by a learning phase about the topic of cell division. After the learning phase, learners had to answer a learning test. In Experiments 2 and 4, eye-tracking was used as a process measure of learners’ attention distribution.

Experiment 1 investigated whether implementation intentions can foster effective processing during multimedia learning, thereby improving learning. Furthermore, it tried to shed light on the question of how learners’ task-specific motivation interacts with the use of implementation intentions. The use of implementation intentions had only marginal impact on learners’ recall and no impact on their transfer performance. However, for recall performance, there was an interaction between the use of implementation intentions and learners’ interest: Those learners with little interest in the learning task significantly benefited from using implementation intentions, whereas those learners with more interest did not.

Experiment 2 focused again on the question whether implementation intentions can improve multimedia learning by supporting the underlying cognitive processes. Additionally, I was interested in what type of multimedia learning processes or combination thereof should best be supported by the use of implementation intentions. All groups learning with implementation intentions outperformed the control group that learned without implementation intentions. There was no difference in learning outcomes between implementation intentions evoking different types of multimedia process (i.e., text processing, picture processing, or integration). Contrary to my hypothesis, three implementation intentions about a singular process type resulted in worse learning outcomes than just one implementation intention, possibly due to interference between similar implementation intentions. The group that learned with three implementation intentions about different types of processes (i.e., one pertaining to text processing, one to picture processing, and one to integration) had the best learning outcomes.

Experiment 3 aimed at replicating the main finding of Experiment 2 against a more conservative control group learning with goal intentions. The implementation intention group outperformed the goal intention group. Thus, the findings of Experiment 2 could be replicated.

Finally, in Experiment 4, I tried to further delineate the differences between the use of implementation intentions and other effective ways to support the use of multimedia processes, more specifically the use of instructional prompts (e.g., Kombartzky et al., 2010; Thillmann et al., 2009). In order to do so, I studied the effect of both the use of implementation intentions and of prompts under different conditions of cognitive load. Under conditions of low cognitive load, both the instructional prompt group and the implementation intentions group showed better learning outcomes than the control group. Under conditions of high cognitive load, however, instructional prompts lost their effectiveness so that learners’ performance did not differ from the control group. Meanwhile, implementation intentions remained effective even under conditions of high cognitive load, so that the implementation intention group demonstrated better learning outcomes than both the control and the instructional prompt group.

The sum of all results from the four experiments conducted for this dissertation suggest that implementation intentions are an effective means to foster the use of effective multimedia processes and improve multimedia learning. It can be assumed that implementation intentions are especially effective if they cover a broad range of multimedia processes and are phrased in such a specific way that they do not interfere with each other. Furthermore, this dissertation demonstrated that implementation intentions compare favorably to instructional prompts because they are effective regardless of learners’ cognitive load.