Abstract:

Being a successful internet reader requires the use of new strategies to cope with the reading and navigation demands of hypertext (such as Wikipedia), that go beyond print reading strategies (Afflerbach & Cho, 2010). In the last decade there have been great efforts to understand how university students apply it. But, further studies are needed to analyze how middle school students behave in this new scenarios. To help fill in this gap we propose the use of an innovative methodology that could be used to explore reading comprehension and navigation in Wikipedia documents.
Many studies of learning in hypertext have focused on offline measures of comprehension, i.e. the final product of reading such as success in answering comprehension questions. A problem with these measures is that they don’t capture the underlying cognitive processes that ultimately lead to that final result. Both aspects are causally related, as the former leads to the latter. To cope with this problem it has been pointed out the importance of analyzing online data incombination to offline measures (Rouet & Passerault, 1999). Three main online measures have been used in the literature: students’ eye movements, navigation decisions, and think aloud protocols. Each of this measure provides a unique view of the students’ cognitive processes used during the learning task (e.g. time spent reading a relevant word, the decision whether to select a link, the regulation reason that leads to a specific behavior…).
However, in the literature of hypertext comprehension online measures have been traditionally used in isolation. We propose data triangulation by using the methodology of cued retrospective think aloud (van Gog, Paas, van Merrienboer and Witte, 2005), which would provide a richer picture of the strategies used by middle school students to read and navigate in Wikipedia. In cued retrospective think aloud students first perform the task at their own pace, while their eye movements are recorded. After completing the task, students are shown a video that reproduces their exact behavior (eye-movements and navigation) in the Wikipedia document. Students are requested to explain what they were thinking while they read and navigate each specific part of the hypertext, using the video as a cue. With this methodology we aim to capture how students combine information decoding, interpretation and use of comprehension and navigation strategies, which is not present in each of the online measures if used in isolation.For example, analyzing eye-movements we can observe if a student paid special attention to an irrelevant link, with the navigation information we can know if she finally clicked on it or didn’t, and the think – aloud protocol let us know why did she take that decision.
Following this rationale, we are currently applying that technique in a study in which 31 high-school students that read and navigate in a Wikipedia document to learn about the French Revolution. We are currently analyzing the data, and preliminary analysis will be reported at the conference. We expect that the use of the cued retrospective think aloud will allow us to identify complex reading and navigation strategies used by high-school students, and its relation with comprehension outcomes.

Keywords:

Eye-movements, hypertext, cued think aloud, wikipedia, navigation decisions.