以文本內容分析取向測量反芻反應之初探研究

Measuring Ruminative Responses by the Content Analysis of Essays: A Preliminary Study

本研究提出文本內容分析方法測量反芻反應，以補充自評問卷測量方法之不足。研究者根據 Nolen-Hoeksema （1987, 1991）反應風格理論與後續研究（Treynor, Gonzales, & Nolen-Hoeksema, 2003），將反芻區分為憂思和反思， 同時參考情緒書寫（Pennebaker, 1993, 1997; Pennebaker & Beall, 1986）之作法，建立文本分析評分系統；三名評分 者評定 98 名大學生受試者之個人壓力經驗文本中出現之憂思、反思、問題解決以及相關情緒，並採用反應風格量 表短版、憂鬱焦慮與壓力量表短版、正負向情緒量表、生活滿意度量表與之比較。本研究建立了一個具有評分者間 信度與效標關聯效度的文本內容分析方法，佐證了 （1） 反芻的兩個次類型「憂思」和「反思」的關係，（2） 兩者與負 面情緒的關聯，（3）「反思」與「問題解決」的關聯，初步解釋了反思的適應性功能，以及（4）憂思者有較多負面情緒， 且和自我責備有關，可解釋憂思具不適應性。本研究提出之文本內容分析方法能呈現反芻反應與探究其適應性，並 補足自評問卷測量會使受試者落入分心、脫離脈絡等的問題；此一初探對較少被觸及的反芻內容研究提出未來的精 進方向。

This study proposed a text content analysis method to measure ruminative responses as a complement to selfreport questionnaires of rumination. Researchers established a content analysis rating system that followed expressive writing (Pennebaker, 1993, 1997; Pennebaker & Beall, 1986). Based on the Response Style Theory (Nolen-Hoeksema, 1987, 1991) and additional theory refinements (Treynor, Gonzales, & Nolen-Hoeksema, 2003), rumination was categorized as either brooding or reflection. Essays about stressful experiences written by 98 university students were evaluated by three trained raters on the frequencies of brooding, reflection, and problem-solving, along with the levels of negative emotions. Self-rated questionnaires, including the Chinese Response Style Questionnaire-short form, the Depression Anxiety Stress Scales short form, the Positive and Negative Affect Scale, and the satisfaction with life scale, were also conducted for comparison with the ratings from the content analysis. This content analysis method attained good inter-rater reliability and criterion-related validity. The analysis results supported the hypotheses that: (1) there is a relationship between two rumination subtypes, brooding and reflection; (2) brooding and reflection are both related to depressive emotions; (3) reflection is related to problem-solving, which indicated the possibility of the adaptiveness of reflection; (4) brooders had more negative emotions related to self-blame, which explains the maladaptiveness of brooding. Conclusions: The content analysis method measured ruminative responses and the adaptiveness of rumination. The method kept participants' attention in the given context and prevented them from distraction when they presented their ruminative responses. These advantages might make the content analysis method surpass self-rated questionnaires when measuring rumination. This preliminary study on ruminative content, which had previously been insufficiently examined, provides future directions for improving the measure of rumination.