運用眼球追蹤法探索高中學生之論證架構理解與論證表現及科學認識信念之關係

Abstract

臺灣十二年國民基本教育課程綱要（以下簡稱108課綱）期望培養學生「帶著走的能力」，其中自然領綱中加強了對於「科學論證」的重要性，其目的是希望學生透過在科學教育中的學習，能夠獲得「靠證據說話」的技能，強調科學中證據及證據如何支持主張的重要性。 本研究利用眼球追蹤技術探討臺灣首都臺北市某公立高中三年級學生40位，於論證架構訓練後，是否能區辨科學文本文中Toulmin（1958）所提出的TAP（Toulmin Argumentation Pattern）論證元素，並進一步探索閱讀歷程、先備知識、科學辯證信念與辨認論證架構的關係。研究中先以問卷收集受試者的科學認識信念，接著以測驗卷檢驗受試者先備知識，然後讓學生閱讀論證架構說明文，再以改寫自《科學人》的閱讀材料檢測學生的論證元素區辨表現，並進行眼球追蹤實驗，以Tobbi-4C眼動儀紀錄受試者的閱讀歷程，最後以開放式問答題觀察學生是否能針對文本內容進行論證。此外，為了釐清論證架構說明文的介入效果，本研究操弄論證架構說明文內容，一組學生單純閱讀論證元素，另一組學生則閱讀論證元素加上論證分析範例。 本研究使用共變異數分析（ANCOVA）檢驗論證架構說明文中，「有無分析範例」的表現與學生先備知識是否產生交互作用，接著使用無母數U檢定（Mann-Whitney U Test）比較兩組之間的論證元素區辨表現差異；最後將受試者進行論證的答案進行編碼並評分，然後將編碼結果以及受試者論證區辨正確率與科學認識信念、閱讀歷程進行相關性分析（Spearman Correlation）。 研究結果顯示，「有無分析範例」與學生先備知識並無交互作用，並且兩組的受試者在區辨TAP要素的表現並無顯著差異。受試者的論證要素區辨能力與先備知識的高低有關，也與持有科學認識信念中的權威辯證信念及多元來源辯證信念有關，並且在眼動指標中的總注視時間、平均注視時間、首次閱讀時間也有關。同時受試者進行論證的表現與眼動指標中的首次閱讀時間有關。

The 12-year National Basic Education Curriculum Outline (referred to as the 108 Guidelines) of Taiwan aims to cultivate students’ “ability to apply what they have learned”. In the “Domain of Natural Science”, the importance of “scientific reasoning” is emphasized. The goal is for students to develop the skill of “speaking with evidence” through scientific education, highlighting the significance of evidence and how it supports claims in science. This research employs eye-tracking technology to explore whether 12th-grade students from a public high school in Taipei City, Taiwan, can identify the Toulmin Argumentation Pattern (TAP) elements proposed by Toulmin (1958) in scientific texts after receiving an online training about the TAP framework. Additionally, the study investigates the relationships among reading processes, prior knowledge, scientific epistemic beliefs, and the recognition of TAP structure and elements. The research began by collecting participants’ scientific epistemic beliefs through questionnaires, followed by assessing their prior knowledge using a multiple-choice test. Students then were given the online TAP training. Afterward, they were led to read an science article adopted from"Scientific American" and asked to identify the TAP elements in the text. During reading, the Tobii-4C eye tracker was used to record participants' reading processes. Finally, open-ended questions were used to evaluate students' ability to perform argumentation. To gauge the intervention effect of the TAP training, the study manipulated the content—students either read TAP structure alone or with an added example of the analysis for TAP elements. The research employed Analysis of Covariance (ANCOVA) to examine the interaction between performance of identifying TAP element under different conditions of TAP training and the extent of students’ prior knowledge. Furthermore, the Mann-Whitney U Test was utilized to compare the performance differences in identifying TAP elements between the two groups. Participants' argumentation responses were encoded, scored, and then correlated with their accuracy in recognizing TAP elements, scientific epistemic beliefs, and reading processes using Spearman Correlation. The findings revealed that the "presence of analysis examples" did not interact with students' prior knowledge, and there was no significant difference in the performance of TAP element discernment between the two groups. Participants' ability to distinguish TAP elements was related to their prior knowledge, epistemic beliefs about justification by authority and justification by multiple sources. Additionally, total fixation time, average fixation time, and first-pass time in eye-tracking metrics were related to the participants' performance. The ability to perform argumentation was found to be also associated with the first-pass time in eye-tracking metrics.