A Methodological Alternative to Media Comparison Studies: Linking Information Utilization Strategies and Instructional Approach in Hypermedia Learning

Literature reviews on hypermedia learning have yet failed to show consistent positive effects of learner-controlled nonlinear information access. We argue that a possible reason for this lack of evidence in favor of hypermedia learning results from the fact that not sufficient attention is paid to the strategies of information utilization learners deploy. The few studies that do analyze these strategies fail to link them to an instructional approach, which hampers a deeper interpretation of strategy patterns. Our study showed that different groups of learners can be distinguished according to their strategies used in an example-based hypermedia environment and that these groups differ with regard to learning outcomes, but not individual learner characteristics. Is Learning with Hypermedia Ineffective? Hypermedia environments are nonlinear information networks, which can be explored in multiple ways with the learner having control over the selection, sequencing, and pacing of information. Learner control is seen as beneficial for knowledge acquisition for different reasons: First, learners can adapt the presentation of information to their preferences and needs. Second, learner control requires a deeper information processing, because learners have to compare the options offered to come to an informed decision. Third, learner control may train meta-cognitive self-regulation abilities. Fourth, it may improve motivation to learn and learners’ attitude towards the topic. Despite these hypothesized advantages, literature reviews have yielded ambiguous results concerning the effectiveness of hypermedia learning over other forms of instruction (Dillon & Gabbard, 1998). In the remainder of the paper, two arguments will be brought forward and supplemented by some initial empirical support: First, it is proposed that analyzing the way learners make use of information offered in hypermedia environments might be a promising methodological alternative to global media comparisons. Most of the existing studies tend to ignore that once learner control is provided, learners’ information utilization behavior will show an increased variance. Thus, whether hypermedia is effective will depend on whether learners use the available information in a way that knowledge acquisition processes are facilitated. Second, the analysis of information utilization strategies should be linked to the instructional approach implemented in a hypermedia environment. Based on cognitive task analyses and empirical evidence, an instructional approach (e.g., learning from worked-out examples) can be characterized by a set of cognitive processes (e.g., example elaborations and comparisons) that enable knowledge acquisition (e.g., construction of a problem schema), where different cognitive processes require different information as input. Thus, an effective information utilization strategy can be defined a priori by the fact that information is selected, sequenced, and paced in a way that cognitive processes relevant to the instructional approach are facilitated. Until now, there are only a few studies that have considered strategies of information utilization when studying hypermedia learning. Barab, Bowdish, Young, and Owen (1996) used students’ navigational profiles based on different strategy indicators (e.g., number of pages retrieved, depth of search) to predict whether a user had been given a specific information search goal for browsing the system. Balcytiene (1999) identified different groups of learners (i.e., self-regulated learners, cue-dependent learners) by means of logfile and video analyses for her hypertext system on gothic architecture. Cue-dependent learners did not browse the environment in a systematic way and oriented themselves towards its local aspects, while self-regulated learners reflected on the accessed information and were oriented towards the global structure of the system. In accordance with this interpretation of the navigational profiles and video data, the posttest revealed superior retention for self-regulated learners. Lawless and Kulikowich (1996) used a cluster-analytical approach to identify different 2117 ha l-0 01 97 35 8, v er si on 1 14 D ec 2 00 7 Author manuscript, published in "28th Annual Conference of the Cognitive Science Society, 2003, Vancouver : Canada (2006)"