大學生網路資訊之批判思考運作暨相關教學的實驗研究

Abstract

網路上充斥著虛實不分的訊息，無論是為了學習與求知、還是為一般生活資訊的搜尋，大學生面對網路資訊時，究竟是不願意深入思考、漫不經心的全盤相信？還是以理性深思，根據邏輯與推理法則判斷何者應信、何者應為？本研究共設計兩個實驗，探討相關問題。實驗一旨在探討大學生閱讀不實卻具說服性的網路資訊(亦稱網路謠言)時，其態度的改變與訊息處理歷程是採用中央路徑亦或是邊陲路徑，以推敲可能性模式(Petty & Cacioppo, 1981, 1986)作為理論依據。採前後測－控制組的準實驗設計，49位大學生為受試，實驗任務要求受試者閱讀不實網路訊息(隱形眼鏡族長時間使用電腦會引發健康隱憂，取自東森網路謠言追追追網站)，並上網搜尋、查證訊息的可信性。受試者分為三組，第一組為無自由意願—有周邊線索組、第二組為無自由意願組、第三組為有自由意願組，本研究假設無自由意願、呈現周邊線索(利益)的情境下，個體傾向採取邊陲路徑的思考方式，較不會進行深思熟慮的訊息處理，亦不利發揮批判思考。依變項為「一般批判思考意向」，採用葉玉珠（1999）編製之「批判思考意向問卷」；以及「網路資訊之批判思考技巧」，採自編「網路資訊之批判思考內容分析表」，由評分者登錄。結果發現：第一，雖然進行實驗任務後，各組受試對於不實網路資訊的相信程度均顯著下降，但組別與時間對一般批判思考意向之交互作用未達顯著水準。第二，三組進行實驗任務時，展現的「網路資訊之批判思考技巧」整體效果與運用策略向度出現顯著差異，「有自由意願組」顯著優於「無自由意願組」，又更優於「無自由意願-周邊線索組」。顯示使用者有無訊息處理的自由意願、資訊是否呈現周邊線索會影響其尋求中央或邊陲路徑的思考。不過，三組在搜尋網路資訊時所使用的「關鍵字組數」、「訊息總量」未達顯著差異，有可能的原因如Ennis (1993)主張批判思考是理性的深思，因此，即便是有動機要執行批判思考的第三組，也無法自發的進行複雜的批判思考策略。因此實驗二以直接教學，試圖提升大學生的網路資訊批判思考。 實驗二旨在探討個體處理網路資訊時，批判思考運作是否因為網路資訊之批判思考技巧教學而提升。採前後測－控制組的準實驗設計，30位大學生為受試。「網路資訊之批判思考技巧教學」之設計融入Big Six方案（程良雄，2001）、加州教育廳批判思考教學模式(Costa, 1990)，也融入IMPACT批判思考宇宙方案(IMPACT, 1991)，教導學生面對未經評估的資訊，如何澄清問題、資訊搜尋策略、找到與取得資訊、利用資訊、綜合及評鑑，也教導分辨事實與意見、資訊的相關性與可靠性，再運用邏輯推理與判斷的原則從事批判性思考。實驗的第一因子為教學(獨立樣本：教學組與控制組)，第二因子為時間(相依樣本：前測與後測)。實驗組接受「網路資訊之批判思考技巧教學」，控制組則否，兩組均閱讀不實的網路訊息(網路謠言)，並上網搜尋、查證訊息的可信性。依變項為一般批判思考技巧，採用Facione(2002)編制的「加州批判思考思考測驗A式」；以及「一般批判思考意向」、「網路資訊之批判思考技巧」。研究結果發現：第一，組別與前後測在「一般批判思考技巧」及「一般批判思考意向」之交互作用均未達顯著水準。第二，兩組進行「網路資訊之批判思考實驗任務」，所展現的「網路資訊之批判思考技巧」達顯著差異，教學組優於控制組。第三，兩組在實驗任務中搜尋資訊的「關鍵字組數」及「訊息總量」達顯著差異，教學組均優於控制組。第四，「網路資訊之批判思考實驗任務」能引導受試者理性的評鑑網路資訊的正確性、並改變受試者對網路資訊是否可信的態度。第五，以半結構性問卷的資料詮釋，結果支持「網路資訊之批判思考技巧教學」實驗的成效。 本研究的結論有五：第一，實驗前大學生受試者對「網路資訊批判思考實驗任務」中不實的網路資訊多表相信，顯示大學生對網路訊息的可信程度不夠警覺。第二，本研究設計的「網路資訊之批判思考實驗任務」，能引導受試者以深思熟慮評鑑網路資訊的正確性、並改變受試者對網路資訊是否可信的態度。第三，訊息處理意願與資訊呈現的周邊線索(利益)會影響資訊接受者在訊息處理歷程中的批判思考技巧運作。第四，網路資訊之批判思考技巧為領域特定（domain specific）的認知技能，可經由教學予以提升；本研究提供一基礎的教案設計，未來可以在大學或高中資訊教育課程中，作為培育資訊素養—訓練學生檢驗網路不實資訊、運作批判思考的教材內容。最後根據研究結果，提出網路資訊批判思考教學的具體建議，並提供若干建議供未來研究與教學實務參考。

Internet has become one of the significant social contexts that university students of this age heavily engage in. When university students browse and read Internet messages, are they willing to think/analyze mindfully and critically? Are they easily persuaded by unevaluated messages posted in websites or do they spend time doing logic inference to determine what to believe and what should be done? Do they carefully check the credibility of unevaluated Internet messages before they get persuaded? This study designed an experiment to explore previous questions and then conducted an instruction to help students think critically in reading unevaluated Internet messages. The first experiment is to examine whether students’ motivation and message cues influence their ways in processing unevaluated Internet messages. Based on Elaboration Likelihood Model (ELM; Petty & Cacioppo, 1981, 1986), there are two ways people make decisions and hence get persuaded in reading Internet information. First, when an individual is motivated and able to pay attention, one takes a logical, conscious thinking to decision-making. This leads to permanent change in one’s attitude as he/she adopts and elaborate on the arguments presented in Internet messages. It is coined as the central route. In other cases, one takes the peripheral route in which an individual does not pay attention to persuasive arguments but is swayed instead by surface characteristics. A quasi-experiment with pre-posttest and control group design was adopted and 49 undergraduate students recruited from a Research University in Northern Taiwan served as subjects. Subjects were randomly assigned to three experimental settings, (1) low-choice/peripheral-cue, (2) low-choice, and (3) high-choice. All subjects were asked to read a message presented on an Internet browser, “contact lens – the cause of cataract”, that has been identified as a rumor with inaccurate health information spread on the Internet. Then they verified the credibility of this message and were told that the university will initiate a campaign in campus to promote students’ awareness of contact lens causing cataract. The first group was asked to write a paragraph, calling peers’ attention to the problem of contact lens, which will be printed in the poster of the campaign. They were also told that they have to write it simply because the random assignment (low-choice) but the writing of the paragraph can earn them partial credit in a related course (reward - peripheral cue). The second group was treated the same as the first group except no reward was given (low-choice). The third group (high-choice) was told to write a paragraph to express whatever they think and feel about the message about contact lens. We assume that the high-choice group has higher motivation to elaborate on the message than the low-choice group and the group given low-choice/peripheral-cue. The low-choice/peripheral-cue group is directed away from the argument or conscious thinking to the surface features of health information. The dependent variables were critical thinking disposition (measured by Critical Thinking Disposition Scale, Yeh, 1999) and critical thinking skills on Internet information, categorized by three experts according to the CTSII coding system developed by the authors. The results are as follows. First, every group’s credulous extent of unevaluated messages decreased after the experiment tasks, but the interaction of groups and time (pretest-posttest) on critical thinking disposition were not statistically significant. Second, three groups displayed significantly different critical thinking skills on Internet information, in terms of total score and strategy use. The high-choice group showed more critical thinking skills than the low-choice group and the low-choice group in turn showed more critical thinking skills than the low-choice/peripheral-cue group. Meaning that students’ motivation and peripheral cues impact whether they take central or peripheral routes while processing unevaluated Internet messages. Third, the number of keywords and messages that subjects employed to search information between groups are not statistically significant. It is likely that critical thinking is mindful and difficult; thus even the high-choice group with motivation to engage in elaborative thinking is not able to perform it. In this reason, the study then conducted an instruction to help students thinking critically in reading unevaluated Internet messages. The aim of the second experiment was to explore a little further into the effectiveness of the instruction designed to teach university students’ critical thinking of unevaluated Internet information. A quasi-experiment with pre-posttest and control group design was adopted and another 30 undergraduate students recruited from a Research University in Northern Taiwan served as subjects. The critical thinking instruction integrated Big Six, California critical thinking instructional model (Costa, 1990) and IMPACT (IMPACT, 1991). An experiment with 2 (between-subjects: experimental and control group) ×2 (within-subjects: pretest and posttest) factorial design was held to clarify this effect. Subjects were randomly assigned to two experimental settings. The experimental group received instruction, while the control group received no instruction. Then they were asked to read an unevaluated Internet messages that had been identified as a rumor and then to verify its credibility. The dependent variables were critical thinking disposition (measured by Critical Thinking Disposition Scale, Yeh, 1999), critical thinking skills (California Critical Thinking Skills Test Form-A, Facione, 2002) and critical thinking skills of Internet information, categorized by three experts according to the CTSII coding system developed by the authors. The results are as follows. First, the interaction of groups (experimental and control group) and time (pre-posttest) on critical thinking disposition and critical thinking skills were not statistically significant. Second, the critical thinking of Internet information between groups was statistically significant; the experimental group was statistically higher than the control group. Third, the number of keywords and messages that subjects employed to search information in experimental group were significantly greater than those in the control group. Fourth, subjects’ credulous extent of unevaluated messages decreased while perceived importance of rational evaluation about Internet information enhanced after the experimental task. Fifth, above results were corresponded to findings conducted by semi-structured interview. The effect on the critical thinking of information instruction was significant. In conclusion, (1) when university students browsed and read Internet messages, they were credulous of Internet messages; they failed to evaluate online messages critically. (2) The experimental task helped students think and evaluate critically in reading unevaluated Internet messages and that helped them change the credulous attitude of unevaluated messages. (3) Information receivers’ motivation and peripheral cues influenced their ways in taking central or peripheral route while they were processing unevaluated Internet messages. (4) Critical thinking of Internet information was likely domain specific and it could be enhanced by adequate instruction. In the last part, we further related above findings to previous research and provided suggestions for educators.