Constructing and confirming a cognitive model of mechanical kinematic representation by recording eye movement during the reading of texts and diagrams

• Main Authors: Yu-Cin Jian, 簡郁芩
• Other Authors: Chao-Jung Wu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/92728018646693341205

博士 === 國立臺灣師範大學 === 教育心理與輔導學系 === 101 === The process of forming kinematic representations contributes to high-order cognitive behavior in humans. The purpose of this study was to construct a cognitive model of kinematic representations within a machine. Two experiments were designed to confirm this model. This model examined the function of text and diagrams for forming internal representations of configurations and kinematic information. Moreover, this model also examined how configuration and kinematic information interact with each other during reading. Readers’ eye movements and comprehension were monitored as they read a piece of text. Experiment 1 consisted of a two-stage procedure: readers first read diagrams and then read a text-and-diagram article about a flushing cistern. We investigated whether readers could construct kinematic representations of diagrams with or without arrows (arrow group versus non-arrow group). Results showed that step-by-step question scores were higher for the arrow group than for the non-arrow group after reading the mechanical diagrams; however, this difference disappeared when both groups read the text-and-diagram article and then revised the step-by-step questions. In addition, scores on troubleshooting questions were higher for the arrow group than for the non-arrow group after reading the text-and-diagram article. In our analysis of eye movements, the arrow group had shorter mean saccade lengths on the diagrams and had longer gaze durations toward the first diagram than the non-arrow group. As for the non-arrow group, the strategy they seemed to use was comparing the status between the two diagrams. Therefore, there were more saccades between the two diagrams for the non-arrow group than for the arrow group. Experiment 1 confirmed that diagrams with arrows conveyed kinematic information, but the kinematic information conveyed was not the same as what could be conveyed by words. In Experiment 2, we investigated whether readers have the ability to form internal representations of mechanical configurations described via written text. The relationship between a mechanical configuration and kinematic information was also of interest. Readers first read the mechanical configurations, which consisted of text or a diagram, and then read text describing kinematic information of the same system. Participants were assigned to one of three groups (configuration group, non-configuration group, or diagram group). Results showed that the configuration group had higher configuration tests scores than the non-configuration group. However, both groups made progress on the revised configuration tests after reading the kinematic text. In addition, the configuration test scores regarding local connective relations were higher for the configuration group than for the diagram group. However, on the global dimension of the same test, the diagram group did better than the configuration group. In our analysis of eye movements, the configuration group displayed shorter total fixation durations and rereading times of the kinematic text than did the non-configuration and diagram groups. Experiment 2 confirmed that readers were able to form a mental representation of the mechanical configuration described by written words. However, both text and diagrams have their advantages for describing configurations; the former’s advantage is on the level of describing local connective relations, while the latter’s advantage is on the level of global analogic relations within the mechanical system. Readers can utilize mechanical configurations to form kinematic information. Conversely, kinematic information could also be used to revise and adjust readers’ mental representations of mechanical configurations. In sum, the model of the current study has been confirmed through the results of two experiments. We discuss relevant empirical research and theoretical underpinnings as they pertain to the current study.