Summary: | 碩士 === 國立臺灣師範大學 === 科學教育研究所 === 106 === This study explored the impact of signaling and segmentation principles on college students' learning process and outcomes of reading science text with illustrations. Eighty participants read the article "life cycle of HIV". The manipulation of color-coding (as the signal, SI) and segmentation (SE) resulted in four groups in this study, including no-SI/no-SE, SI/no-SE, no-SI/ SE, SI/ SE. The use of color-coding as a signal between the text and illustrations was expected to increase transition saccades between them. The segmented design, which divided the fifteen steps in the text into four sections, was implemented to reduce cognitive load. Eye tracking records and post-test were obtained in this study to probe into the learners’ reading process and outcomes. The eye movement indicators include fixation durations on the illustration / text, fixation counts on the illustration / text, percentage of the first 10 fixations on the text, the number of transition saccades between illustration and text, search efficiency from text to the corresponding illustration, total fixation durations on target words, total fixation durations on target illustrations, first-pass forward fixation durations, first-pass reinspection time, and look-back time. The post-test results showed that text with color-coding and segmented design led to better learning performance than that without these two manipulations; however, there was no interaction between these two factors. There were three main findings from the eye movement analyses. (1) The participants did not change the text-directed reading pattern because of the signaling and the segmentation manipulation. (2) In the signaling groups, learners made more transitions between illustration and text, increased the accuracy in mapping from the text to the correct location of the corresponding illustration (higher search efficiency), and spent more time processing illustrations. (3) In the segmented groups, learners increased processing of textual representations, and understood the relationships and changes among steps better than the other groups.
|