個人性別態度檢測量表之建構效度研究

本研究旨在探討「個人性別態度檢測量表」能否測量出短期、小幅度之改變。藉由多元帶領者（包含實習諮商心理師、諮商心理師、社工師等）在大專院校帶領之各種性別相關活動或課程，蒐集前後測分數，並且分析測量量表分數之變化與相關因子，以驗證該量表反應性（responsiveness）。　　量表數據資料總數收回298份，剔除無效的16份後，共剩下282份數據資料。並透過SPSS統計軟體做重複測量變異數分析（Repeated Measure ANOVA），與Bonferroni作為Post Hoc事後檢定，以回答本研究問題：（一）個人性別態度檢測量表之整體前後測分數，是否達顯著差異？（二）不同性別的參與者，在性平推廣課程之前後測分數是否達顯著差異？（三）不同學院的參與者，在性平推廣課程之前後測分數是否達顯著差異？（四）不同地區的參與者，在性平推廣課程之前後測分數是否達顯著差異？（五）由不同身份之帶領者所帶領的性平推廣課程，其課程參與者的前後測分數是否達顯著差異？。而本研究假設在如此多元跟豐富的組合當中，若仍無一達顯著差異，則可能推論的是該量表未具有足夠的敏感度，倘若能夠達顯著，則得以證實，該量表能夠反應出個體的短期改變，意即該量表具有足夠的敏感度。　　分析結果發現分量表一女人刻板印象，在四個研究問題中皆呈現α≦.05達顯著差異，而男人刻板印象、多元性別友善兩分量表，則都沒有達顯著差異。可能的解釋之一是僅女人刻板印象該分量表有足夠的敏感度，得以測得參與者的短期改變，而男人刻板印象、多元性別友善兩者分量表則不具備足夠敏感度。　　最後，本研究依據研究結果進行討論並提出相關建議，同時回顧歷程以整理研究限制與自我檢討，以供未來性平相關單位以及實務研究工作之參考。

關鍵字

性別態度；反應性；重複測量變異數分析

並列摘要

The study aims to explore whether the "Personal Gender Attitude Test Scale" can measure short-term, small changes. Various gender-related activities or courses led by multiple leaders（including intern counseling psychologists, counseling psychologists, social workers, etc.）in colleges and universities, collect pre and post test scores, and analyze the changes in the scores of the measurement scale and related factors to verify the responsiveness of the scale. 　　A total of 298 copies of the scale data were recovered, after excluding 16 invalid copies, a total of 282 data materials remain. And do Repeated Measure ANOVA through SPSS statistical software, with Bonferroni as a Post Hoc test to answer the question of this research:（1）Is there a significant difference in the overall pre- and post-test scores of the Personal Gender Attitude Test Scale?（2）Participants of different genders, are there significant differences in the test scores before and after the sex education program?（3）Participants from different colleges, are there significant differences in the test scores before and after the sex education program?（4）Participants in different regions, are there significant differences in the test scores before and after the sex education program?（5）Sex promotion courses led by leaders of different identities, is there a significant difference between the pre-test scores of its course participants? And this research hypothesizes that in such a diverse and rich combination, if there is still no significant difference, it may be inferred that the scale is not sufficiently sensitive. If it can be significant, it can be confirmed that the scale can reflect Short-term changes in the individual mean that the scale has sufficient sensitivity. The result of the analysis found that the stereotype of a woman on the subscale one, in the four research questions, there are significant differences in α≦.05, and the two subscales of male stereotypes and multi-gender friendliness, there is no significant difference. One possible explanation is that only women’s stereotypes have sufficient sensitivity for this subscale, to measure short-term changes in participants, the two subscales of male stereotypes and multi-gender-friendliness are not sufficiently sensitive. 　　Finally, this research discusses and makes relevant recommendations based on the research results. At the same time, it reviews the history to sort out the research limitations and self-examination, so as to provide reference for future sex-related units and practical research work.