「心田寬恕量表」之中文化與信、效度分析

Reliability and Validity of the Chinese Version of the Heartland Forgiveness Scale

過去華人寬恕測量研究多聚焦在單一面向，實為研究缺口，尚待多向度測量來延伸開展寬恕各面向研究。故本研究旨在採用Thompson等人（2005）的理論發展適用於臺灣大學生的「心田寬恕量表」。該量表分成三個分量表，包括：寬恕自己、寬恕他人及寬恕情勢。預試以分層便利抽樣抽取229人，經項目分析及信度分析後，保留題目共18題。正式樣本涵蓋全臺灣北中南區之大學生共503人，依各區百分比進行取樣分配。結果發現全量表Cronbach’s α係數為 .87，各分量表 Cronbach’s α係數為 .78∼ .79，顯示本量表穩定性佳及內部一致性良好。建構效度方面，各項整體適配度指標均達標準，顯示本量表整體模式適配情形頗為理想。再以心理復原力量表（Connor-Davidson Resilience Scale, CD-RISC）為效標，具良好同時效度。此外，針對不同背景變項進行考驗，發現不同性別或宗教信仰並不會影響寬恕傾向。本量表深具價值及應用性，採嚴謹直譯及回譯方式，可以提供跨文化的比較，及後續實務應用，對於探索華人不同年齡及族群的寬恕狀態，初步診斷或介入，均能提供有效參考依據。

The field of positive psychology emerged in the 21st century. Positive traits are among the three pillars of this discipline and concern character strengths and virtues. A total of 6 virtues and 24 strengths exist. Park et al. (2004) defined strengths as“positive traits that reflect people’s thinking, feelings, and behaviors.”Forgiveness is one example. This strength can support mental health and is negatively correlated with multiple maladjustment indices, such as depression and anxiety (Mauger et al., 1992).
Early studies on forgiveness have emphasized the forgiveness of other people and examined the transformation of a victim’s negative attitude toward their transgressor to a positive one, as in the design of the Enright Forgiveness Inventory (Enright & Fitzgibbons, 2000) and the Transgression-Related Interpersonal Motivations Inventory (McCullough et al., 1998). Subsequently, attention has been focused on other facets of forgiveness, such as forgiveness of the self, which is defined as an individual’s recognition of their own errors, assumption of responsibility, and process of moving away from and back toward the self. On the basis of the multifaceted theory of forgiveness, studies have developed appropriate instruments to measure this construct, such as the Forgiveness of Self and Forgiveness of Others scales (Mauger et al., 1992), State Self-Forgiveness Scales (Wohl et al., 2008), Multidimensional Forgiveness Inventory (Tangney et al., 1999), and Heartland Forgiveness Scale (Thompson et al., 2005).
The progress of forgiveness research in Chinese-speaking society is approximately 10 years behind than that in the West. The first Chinese-speaking scholars to conduct studies in this field did not emerge until approximately 2000. Most research continue to treat forgiveness as a one-dimensional construct. Therefore, more studies are required to address this gap and examine forgiveness as a multifaceted perspective. Thompson et al. (2005) contributed critically to this promising and vital endeavor. They treated forgiveness as a multifaceted construct and examined the process of granting forgiveness instead of seeking forgiveness. Expanding previous theoretical frameworks, they included an additional facet of forgiveness in addition to that of others and the self: Forgiveness of situations. This approach has been influential in the field, and the corresponding scale has been translated into more than 20 languages.
The present research introduced the theoretical framework of the multifaceted forgiveness scale (i.e., the Heartland Forgiveness Scale) proposed by Thompson et al. (2005). We first translated the scale from English to Chinese by using robust procedures, including back translation, and then examined its reliability and validity. Using stratified convenience sampling, we recruited 229 college students for a pilot test. After item and reliability analyses, the original 18 scale items were retained. Subsequently, for a formal test, we recruited 503 college students from North, Central, and South Taiwan. The number of participants from each area was proportionate to that area’s population to ensure that our sample was representative. Results indicated that the scale exhibited high reliability (Cronbach’sα= .87). The reliability of each of the subscales was high: Forgiveness of the self (α= .79), forgiveness of others (α= .79), and forgiveness of situations (α= .78). In addition, we examined the composite reliability (CR) and the average variance extracted (AVE) of each of the subscales. The CR was .87 for forgiveness of the self, .86 for forgiveness of others, and .87 for forgiveness of situations; their AVE values were .53, .50, and .52, respectively.
After establishing the reliability of the scale, we conducted confirmatory factor analysis by using SPSS AMOS to assess the construct validity of the scale. We used the maximum likelihood to estimate parameters and adapted the criterion proposed by Hair et al. (2009) to evaluate the overall model fit in terms of absolute fit measures, relative fit measures, and parsimonious fit measures. With respect to absolute and relative fit, the fit indices indicated that our model fit the data: Goodness of fit index (GFI) = .94, standardized root mean squared residual = .05, root mean square error of approximation = .06, and adjusted GFI (AGFI) = .91. We obtained the same results for the parsimonious Fit: Parsimony normed fit index = .69, and parsimony GFI = .63.
We examined the concurrent validity of the scale by testing the correlation between forgiveness and resilience, as measured using the Connor-Davidson Resilience Scale (CD-RISC). The results indicated that the entire forgiveness scale and its subscales were all positively significantly correlated with the CD-RISC and its subscales (i.e., positive acceptance of change, strengthening effect of stress, and perceptions of control). The r values ranged from .31–.56 (p < .001). The results demonstrated the validity of the Heartland Forgiveness Scale and corroborated the relationship between forgiveness and resilience, a critical well-being index.
By grouping individuals on the basis of demographic variables, we compared individuals’trait level forgiveness. Using multivariate analysis of variance, we entered data on sex, religion, and their interaction into the model to predict individuals’scale scores. None of the predictors were significant (F < 1.5, p > .30).
We determined that the Heartland Forgiveness Scale exhibited high reliability and validity in our Taiwanese college student sample. The strengths of the study include an effectively planned pilot test, a translation procedure with proven reliability, a representative sample of the formal test, and robust statistical tests. However, this study has some limitations that should be addressed. First, future studies should collect data on different age groups, such as older adults. Changes occurring in people’s perception regarding forgiveness with age can affect the factor structure of the scale. Future studies should recruit participants with different occupations. Our study included only college students, and future studies validating the scale with other populations in Taiwan are warranted. Because of the positive correlation between forgiveness and well-being, future studies should develop interventions to help individuals exhibit greater multifaceted forgiveness. The facets of forgiveness are interconnected, and enhancing one facet might help individuals to develop another. However, with an organized intervention addressing the three facets together, individuals may benefit more fully from this character strength. The implications of this study are wide-ranging, and we encourage scholars and practitioners to use our empirically validated multifaceted Heartland Forgiveness Scale. We hope the introduction to this theoretical framework and instrument stimulate interest and spark intellectually diverse conversations among forgiveness researchers and practitioners in Taiwan, enabling the field to steadily develop and society to become more forgiving of others, the self, and situations.