多面向道德情意量表之發展與信效度考驗

Multidimensional Affective Morality Scale:Development and Reliability and Validity Evaluation

本研究目的在發展「多面向道德情意量表」（multidimensional affective morality scale，簡稱MAM scale），並考驗其信效度。研究者汲取國內外道德學界既有經驗與成果，針對國中至研究所學生為對象，並奠基臺灣的在地脈絡，以作為情意層面道德範疇概況或成效展現的具體指標，希冀對於國內外道德教育理論與實踐有所助益。本研究方法與歷程包括先行探究理念架構以擬定量表初稿、繼而針對初稿評定其內容效度、再經修改後進行量表預試與正式施測。研究者區分四個階段進行統計分析，並採用IBM SPSS Statistics 23與Amos 25 Graphics為分析軟體：一是將預試樣本（N＝585）進行項目分析，計算其內部一致性信度；二是將預試樣本區分為二群，以樣本_1（N＝283）進行探索性因素分析；三是以樣本_2（N＝302）進行驗證性因素分析；四是以正式施測樣本（N＝708）進行複核效化檢驗。本研究發展完成的MAM量表包括詳版與簡版，詳版通過內容效度與內部一致性考驗，包括四則道德故事與四個構面，計有44個題項；簡版則是奠基詳版，並具有良好的模式適配度（二階單因子驗證性分析模式），以及理想的組合信度、建構效度與複核效化檢驗，其內容包括三則道德故事與三個構面，計有12個題項。詳版可顯示道德情意量表的脈絡性與豐富性，簡版則可展現結構性與精簡性。藉此量表的發展，本文綜整結論以彰顯MAM量表的特點：一是本量表凸顯情意層面道德範疇，豐富且具深度的多面向意涵；二是本量表與文獻所探討的量表，有其相似度均凸顯道德情意內涵，亦有其故事所關注道德議題的獨創性與在地關聯；三是本研究嘗試以多種類的信效度，對於道德情意量表進行堅實的考驗。最後，本研究基於道德教育的角度，提供在量表運用與後續研究的建議，以期增進各教育階段學生的多面向道德情意發展。

This study developed the Multidimensional affective morality (MAM) scale by testing its reliabilities and validity. On the basis of both national and global academic experiences and research achievements, this study selected students from junior high school to graduate school in Taiwan as subjects to construct a concrete indicator for affective morality and enhance the theory and practice of moral education. This study first explored the rationale underlying the scale's draft. The researcher then examined the content validity and conducted the pretest and formal survey. Statistical analysis was performed utilizing the IBM SPSS Statistics (version 23) and AMOS Graphics (version 25). In Stage 1, an item analysis was performed and the internal reliability was computed for the pretest group sample (N = 585). Stage 2 comprised an exploratory factor analysis (EFA) for pretest_1 group (N = 283). Stage 3 comprised a confirmatory factor analysis (CFA) for pretest_2 group (N = 302). In Stage 4, the cross-validation for the formal survey_ group was assessed (N = 708). The researcher developed two versions of the MAM scale. The long version, comprising 4 moral stories, 4 constructs, and 44 items, passed the test for content validity and internal consistency. The short version, comprising 3 moral stories, 3 constructs, and 12 items, demonstrated goodness of fit (a one-factor second-order CFA model) and good composite/component/construct reliability, construct validity, and cross-validity. The long version of the MAM scale demonstrated contextualization and richness, whereas the short one demonstrated structure and refinement. Thus, the MAM scale has the following characteristics: (1) The MAM scale has good richness and offered an in-depth understanding of the multiple dimensions of affective morality. (2) Similar to other related scales, the MAM scale demonstrated the concepts of affective morality. In addition, the scale encompassed localization and creativity pertaining to moral issues. (3) The MAM scale's reliability and validity could be tested using various evaluation methods. Finally, the researcher offered several suggestions for the scale's application and further research to improve the development of affective morality among students in different educational levels.