以眼動型態探討數學圖文閱讀理解的歷程

This project completed three studies. ”The effects of prompt on comprehension and eye movement when reading geometry proof” included 26 subjects as prompt group and 27 subjects as non prompt group. Through comprehension test and eye movement patterns, this study exam whether prompt influence comprehension and reading process. Prior knowledge test had no group difference. Preliminary results showed prompt group had shorter reaction time of comprehension test in argument item than non prompt group. The accuracy of comprehension tests had no prompt effect. In other word, prompt increased the efficiency of reasoning but not the effectiveness. These result showed that adult readers However, longer reaction time and other reading process might cover the positive effect of prompt. Therefore, the result didn’t show the benefit of prompt. The results of this study were published in PME 37. ”The effect of text and formula symbols on comprehension and eye movement of reading mathematical text” dealt with whether different writing symbols influent the comprehensive process when reading popular mathematical text. The result of 69 subjects from reading comprehensive test and indexes of eye movement indicated that no group effect of accuracy in yes/no questions and calculation items. Text group had better quality of description than formula group. Although TFD had no group effect, formula group had longer TFD of figure in the critical segment than formula group which pointed out the possibility of information supply in figure for formula group. Formula which seems simplify didn’t increase the reading efficacy and had lower reading benefit. Furthermore, this study discussed polysemy of mathematical representation. The part results were published in 2012 ASET annual international conference. “The effect of illustrations in multistep arithmetic word problems: Evidance from problem-solving and eye movements” investigated illustration whether enhance children’s solution in mathematic word problems and affect the reading processing. The finds displayed that illustration significantly improved children’s performance in large group test, but marginal enhance performance in small sample. The effect of illustration was dependent on children’s solution type, especially in semantic consistent problems. The illustration improved successors’ solution, no effect to reversal error solver, and impair other error solvers’ performance.