探討光學成像關係性推理之大腦認知歷程與腦波動態變化

Abstract

本研究之目的為探討學生使用心智模型進行光學成像關係性推理時的大腦認知歷程與腦波動態變化。本研究包含兩個實驗。實驗一參與者依學科背景分為科學主修與非科學主修兩組，每組為41人。實驗二參與者為科學主修之學生60人，且皆接受過實驗一的光學成像關係性推理學習課程。 實驗一探討不同學科背景的學生在光學成像關係性推理學習課程時，其眼動情形、腦波變化與學習成效之間的關係。結果顯示兩組學生均有良好的學習成效，且科學主修之學生在不同測驗階段之回答正確率皆顯著高於非科學主修之學生。另外，眼動資料顯示科學主修之學生在［學習課程階段］中之圖片與文字關鍵區域中的平均凝視時間皆顯著高於非科學主修之學生。再者，腦波資料顯示科學主修之學生在各腦區的theta波活化較強，而非科學主修之學生在枕葉區域的alpha波抑制較強。根據學習成效、眼動情形及腦波變化三者之間的關係，推論科學主修之學生在關鍵區域內具有較持續的注意力及較深度的認知處理。 實驗二探討學生運用心智模型進行光學成像關係性推理及關係整合時的腦波動態變化情形。結果顯示學生在「單一平面鏡」的正確率最高且心智模型操作推理時間最短。在「透鏡組一（1平面鏡＋1凸透鏡）」及「焦距凸透鏡」之正確率較低且心智模型操作推理時間較長。腦波分析結果顯示，在［視覺訊息處理階段］時，枕葉區域在各類型的光學成像關係性推理問題中皆有明顯的alpha波抑制情形。在［心智模型操作推理階段］以及［成像結果作答階段］時，大腦前扣帶回區域皆發現有明顯的theta波活化情形，且在「單一平面鏡」時，大腦前扣帶回theta波活化較其他類型的光學成像關係性推理問題弱。此外，學生進行心智模型操作推理時，前半段時間會以圖像中的部分特徵進行心智模型的操作及翻轉，因此可發現有明顯的頂葉alpha波抑制情形，而後半段時間則是進行圖像重建、特徵整合以及重複驗證等認知活動，而這些認知活動可能反映不同光學成像關係性推理問題的複雜度，因而使得頂葉的alpha波活化強度隨著光學成像關係性推理問題的複雜度提高而有逐步增強的現象。

This study employed the electroencephalography (EEG) to explore the brain dynamics with underlying cognitive process during optic relational reasoning. This study consisted of two research experiments. There are forty-one science major and forty-one non-science major undergraduate students were recruited to involve in the experiment 1, and there are sixty science major undergraduate students who have received the optic relational reasoning learning content in the experiment 1 were recruited to involve in the experiment 2. The first experiment explored the EEG dynamics, eye movement patterns and learning performance of science and non-science major students during the optic relational reasoning learning content. Results showed science major students significantly outperformed than non-science major students on their learning performance of pre-test, learning practice and post-test. The eye-movement patterns indicated the mean fixation duration within area of interest (AOI) was greater in science major students than that in non-science major students, regardless of picture or word area. In addition, the EEG data revealed the theta augmentation was greater in science major students than that in non-science major, regardless frontal, occipital parietal and temporal lobes. In addition, there is greater alpha suppression in occipital was observed in non-science major students compare to science major students. In summary, the science major students allocated greater mean fixation duration at the AOI and greater theta augmentation in the different brain areas contributed to their better learning performance. The second experiment explored the EEG dynamics among six optic relational reasoning conditions which involved different level of relational reasoning ability. Results indicated students had the highest accuracy and the fastest mental model relational reasoning time in the single mirror condition, on contrary, students had the lowest accuracy and longest mental model relational reasoning time during the one mirror and one convex combination condition and convex with focal length. The greater alpha suppression was observed in the occipital during the visual information processing stage, and pronounced theta increase was found in anterior cingulate cortex (ACC) during the mental model relational reasoning stage and subjects’ response stage. In addition, the theta augmentation was lower on single mirror condition than others. During the mental model relation reasoning stage, the parietal alpha suppression was obvious earlier and we considered it is the time that students are performing mental model manipulation and mental rotation, and following parietal alpha increase might reflect the demands of other cognitive functions including image rebuilt, feature integration and confirmation that associated with the level of complexity at different optic relational reasoning conditions.