|Abstract: ||本文說明TANET 2007研討會之摘要格式，本研究以社會認知為理論基礎(Bandura, 1973 & Bandura,1986)，進行兩次實驗研究，探討網路論戰文章、網路認知曲解對情緒與攻擊行為之影響。|
第一次的實驗目的：探討研究者所設計的三篇低、中、高的BBS論戰文章，是否能區分為三個層次。研究者邀請69位大學生，先閱讀研究者所設計之三種網路論戰文章之後，再區分出三篇文章的敵意與厭惡程度。研究結果顯示，三篇文章的敵意(F=4.11, P<.05)與厭惡(F=3.31, P<.05)程度均有差異。閱讀高論戰文章的受試者之敵意與厭惡的得分也顯著大於閱讀低論戰文章的受試者。所有受試者的敵意與厭惡得分也有相關(r=.49, P<.01 )。以上結果顯示，受試者閱讀論戰文章之後，會感受到敵意與厭惡，並且能區分這三篇論戰文章的論戰程度，亦也同時證明研究者所設計之文章具有敵意且有層次之分。研究者遂進一步對三篇文章進行內容分析，將文章命名為直接攻擊論戰、間接攻擊論戰、對立論戰文章。
第二次的實驗目的在探討研究者所設計的三篇網路論戰文章、網路認知曲解對於情緒與攻擊之影響。359位BBS使用者，經由系統隨機分配實驗一中的三篇BBS論戰文章的其中一篇來閱讀，並填寫網路問卷。研究結果顯示，雖然閱讀文章之後的受試者其攻擊行為與認知曲解量表得分並無交互作用(F=1.160, P>.05)，但是網路認知曲解高低分組之攻擊得分達顯著差異(F=76.222, p<.05)。此結果支持了Crick & Dodge(1994), Berkowitz(1993)認為認知曲解會增加攻擊行為的見解相同。雖然在兩次的情緒前後測得分與受試者之攻擊、認知曲解無交互作用(F=.064, P>.05)，但是認知曲解高低組別的情緒在前後測卻是有顯著差異的(F=4.550, p<.05)，而且全體受試者在實驗前後測的情緒得分有顯著差異。從實驗一、二中可知，雖然受試者可以區分網路論戰文章的敵意程度、閱讀文章後情緒也的確受到影響，但卻不會因為閱讀論戰文章而提高其攻擊行為。但是值得注意的是：高認知曲解的受試者無須閱讀論戰文章其情緒與攻擊行為就會被引發。本研究顯示認知曲解為造成網路攻擊行為的主要原因。但是網路論戰文章是否真無法引發攻擊行為，或許有待學者在未來做更深入研究來證實。
Based on the cognitive behavior theory (Bandura, 1973 & Bandura,1986), this study explored the effects of flaming messages and Internet cognitive distortion on emotion and aggression. There were two study of the research. The first study focused on whether the online flaming messages can be classified into several levels. 69 college students read one of BBS messages which were conduct into three levels of flaming. Then they are asked to judge the hostility and disgust levels from the messages (from 1 to 7). The results displayed that there are significant differences in hostility (F=4.11, P<.05) and disgust (F=3.31, P<.05) scores across the three levels of flaming messages. The disgust and hostility scores of high flaming messages were higher significantly than the low flaming messages. There was also a statistically significant relationship between hostility and disgust scores (r=0.49, p<.01).The results above explained that subjects could feel both disgust and hostility, and classify the flaming levels among the messages. The flaming messages were proven to classify into three levels: direct aggressive, indirect aggressive, and disagreement flaming.
The study 2 focused on whether the different level online flaming messages and Internet cognitive distortion had effects on the emotion and aggression. 359 BBS users read one of flaming message which had ever conducted into three types of flaming in study 1, and then finish the online questionnaires. The results indicated that there was not a significant difference in aggressive scores across interaction between the flaming messages and Internet cognitive distortion (F=1.160, P>.05). However, the aggressive grades of high and low level Internet cognitive distortions groups were significant different (F=76.222, p<.05). The result was the same with theories of Crick & Dodge(1994), Berkowitz(1993) that cognitive distortion would increase people’s aggressive behaviors. There was not a significant difference in twice emotion scores across interaction among the flaming messages, Internet cognitive distortion (F=.064, P>.05). However, the emotion scores of high and low level Internet cognitive distortions groups were significantly different (F=4.550, p<.05). The twice emotion scores were significantly different (F=38.349, p<.001). In Conclusion, flaming messages could cause aggressive behaviors and emotional change. Although the subjects could classify the flaming messages and their emotion change during the treatment, they did not behave aggressively after reading the flaming messages. However, subjects with high internet cognitive distortion behave aggressively and emotion change violently without reading flaming messages. The chief reason for online aggressive behaviors was possible the internet cognitive distortions. The results could be explored further in the future.