Preliminary validation of a Chinese version of the comprehensive assessment of acceptance and commitment therapy processes

Acceptance and Commitment Therapy (ACT) improves mental health by augmenting psychological flexibility. Studies showed that psychological flexibility affects mental health; however, although studies have examined individual aspects of psychological flexibility, few have comprehensively examined an entire psychological flexibility model, especially in China. The Comprehensive Assessment of Acceptance and Commitment Therapy Processes (CompACT) is a comprehensive evaluation method that reflects psychological flexibility. The objectives of this study were to develop a Chinese version of CompACT that fits with Chinese culture and examine its reliability and validity. We studied 4999 non-clinical Chinese civil servants and explored the characteristics of the Chinese version of CompACT. Exploratory factor analysis was used to measure Openness to Experience, Behavioral Awareness, and Valued Action that explained 54.3% of the total variance. Confirmatory factor analysis provided support for the three-factor model. The Cronbach’s α-coefficient of the scale was 0.87, suggesting that the Chinese version of the CompACT is reliable. The Chinese CompACT total score represents the overall level of psychological flexibility; it is significantly negatively correlated with negative emotions and positively associated with well-being. Our findings suggest that the Chinese version of the CompACT accurately assesses psychological flexibility in Chinese adults.


Introduction
Traditional cognitive behavioral therapy methods focus on alleviating symptoms of psychological disorders and have led to the development of evidence-based treatments. Nevertheless, comorbidities and heterogeneity are common in syndromal groups; therefore, a single application of cognitive-behavioral therapy protocol to a single disorder may not be optimal. A new generation of process-based therapies (PBT) has emerged to study core mediators and moderators based on testable theories (Hofmann & Hayes, 2019). PBT emphasizes the process of change. They enhance health by influencing the core psychological processes that broadly affect well-being (Dindo et al., 2017;Hayes et al., 2012a).
Recent psychological intervention studies highlighted the significance of measuring the mechanisms underlying the changes in symptoms (Gu et al., 2015). Psychological flexibility refers to selecting behaviors that accord with one's values while entirely in contact with the present moment (Hayes et al., 2006); such flexibility is negatively correlated with many psychological problems. The objective of this study was to develop a questionnaire suitable for measuring the complete psychological flexibility of Chinese people. The psychological flexibility of the Chinese population can be studied effectively, and clients' progress can be monitored during therapeutic processes.
Psychological flexibility expands the interpretation of clinically relevant variables such as functioning and impairment, in addition to well-validated measures of symptoms such as depression, anxiety, stress, and anxiety sensitivity (Gloster et al., 2011). Psychological flexibility is at the core of the Acceptance and Commitment Therapy (ACT; Hayes et al. 2012b). Previous research has shown that it can support and maintain psychological well-being (Renshaw, 2018). Psychological flexibility can be conceptualized as a hexaflex model involving six processes: acceptance, defusion, self as context, present moment awareness, clear values, and committed actions (Hayes et al., 2006).
Psychological flexibility can be encouraged by effective treatment of an individual's negative emotions in the context of their overall mental health. It has been observed that higher psychological flexibility results in lower psychological stress and higher quality of life (Hayes et al., 2013). Psychological flexibility is positively correlated with recognizing and adapting to situational demands and the ability to maintain and balance life needs (Waugh et al., 2011;Gloster et al., 2017). In addition, psychological flexibility has been closely linked to effective treatment of trichotillomania (Lee et al., 2018), depression (Bai et al., 2019), and trauma-related problems (McLean & Follette, 2016). It has been shown to play a moderating role in the change of overall well-being and various psychological symptoms (Puolakanaho et al., 2020;Bardeen et al., 2013). Swiss researchers found that psychological flexibility can modulate a variety of stressors. Gloster et al. (2017) proposed that increasing psychological flexibility may be a desirable and achievable public health target.
Studies on psychological flexibility have been plagued by faulty measurements . Although the Acceptance and Action Questionnaire-II (AAQ-II) (Bond et al., 2011) is by far the most popular psychological flexibility measuring method in use today, it has notable drawbacks such as its emphasis on acceptance and defusion processes and its overlap with general distress measures (Wolgast, 2014). Measures such as the Brief Experiential Avoidance Questionnaire (Gámez et al., 2014), the Cognitive Fusion Questionnaire (CFQ) (Gillanders et al., 2014), and the Valuing Questionnaire (VQ) (Smout et al., 2013) can be used to examine a single process of psychological flexibility.
Researchers have begun to develop comprehensive tools to measure psychological flexibility. Kashdan et al. (2020) created the Personalized Psychological Flexibility Index (PPFI) that includes dimensions of avoidance, acceptance, and harnessing. Preliminary research has shown that PPFI has good reliability and validity and is not conflated with negative emotionality. However, PPFI is idiographic and temporary, and it measures individual coping strategies with depression while pursuing essential goals. Doorley et al. (2020) are working on further revisions to the PPFI to resolve this problem. Francis et al. (2016) compiled the Comprehensive Assessment of Acceptance and Commitment Therapy Processes (CompACT) as an overall measure of the ACT process. It consists of 23 items: openness to experience, behavioral awareness, and valued action. The three dyadic steps of psychological flexibility proposed by Hayes et al. in 2011, "open," "aware," and "engaged," are conceptually consistent with this three-factor structure (Callahan et al., 2020;Hayes et al., 2011). The CompACT has the potential to measure ACT processes (Francis et al., 2016). Recent research provided empirical support for the validity of the CompACT in measuring psychological flexibility. For example, Twiselton et al. (2020) reported moderate to significant correlations between psychological flexibility measured using CompACT, individual well-being, and relationship quality. Gallego et al. (2020) found that higher CompACT scores were negatively associated with public speaking anxiety symptom scores; openness to experience scores also significantly predicted public speaking anxiety symptom scores. These findings suggest that CompACT is a psychometrically sound instrument used in clinical assessment and efficacy studies.
Moreover, among participants with a self-harm history, subscales of the CompACT have demonstrated differential relationships with various non-suicidal self-injury functions (Callahan et al., 2020). A clinical study reported that caretakers of people with multiple sclerosis suffering from strain who attended an eight-week telephone-supported ACT self-help demonstrated significant improvements in the three-factor CompACT scores from the baseline to a six-month follow-up (Potter et al., 2020).
Improving psychological flexibility is a therapeutic goal of the ACT; therefore, most research on psychological flexibility in mainland China has been conducted in the context of ACT (e.g., Chen & Zhu, 2019;Cheung & Mak, 2016;Chong et al., 2017;Fang & Ding, 2020;Hu & Chen, 2020). However, few multidimensional measures of psychological flexibility have been developed to date. There is demand for developing and validating a version of the CompACT suited to the Chinese cultural context. Considering the substantial differences between Chinese and Western cultural backgrounds, it is inappropriate to translate the English version of CompACT into Chinese directly. Therefore, based on the original questionnaire, we sought to develop a Chinese version of CompactACT suitable for use in the Chinese context.

Aims and Objectives
Chinese adults were recruited to examine the psychometric characteristics and performance of the Chinese version of the CompACT. We aimed to test its psychometric properties and compare the instrument with depression, anxiety, stress level, and life satisfaction measurements. We hypothesized that this scale would be a comprehensive instrument to evaluate the psychological flexibility of Chinese adults.

Participants
The participants were civil servants from various parts of China. They were public officials responsible for delivering public services to local areas, exercising state power, and managing social and public affairs (Li, 2014). We sent the scales through an online platform. Before the study began, the participants were informed of the purpose of the questionnaire and that their anonymity was guaranteed. A total of 17,903 civil servants completed the questionnaire, 7842 of whom were excluded from subsequent analysis due to missing data. The effective response rate was 56.19%. The samples were randomly and evenly classified into Groups 1 and 2. We used Group 1 in this study and Group 2 in another study to determine subgroups based on the psychological flexibility, source of pressure, and life satisfaction in Chinese civil servants (in preparation).
Group 1 consisted of 4999 civil servants. Of the 4999 respondents, 73.5% were male, and 26.5% were female. Their average age was 36.6 ± 8.26 years. Most participants attended undergraduate programs (71.3%), and most were married (79.5%). Their characteristics are listed in Table 1. For factor analysis, we used SPSS 25.0 to randomly divide Group 1 (N = 4999) into the exploratory factor analysis (EFA) sample (n = 2509) and the confirmatory factor analysis (CFA) sample (n = 2490).

Translation of CompACT
The CompACT was translated into Chinese in a stepwise process. Two postgraduate students translated the English version into Chinese independently. Then, experts in psychology and the ACT revised it until they reached a consensus. Another postgraduate student then translated it back into English and compared it with the original version of the CompACT to ensure conceptual equivalence.

Development of Initial Chinese Version of CompACT
Considering the cultural differences involved in the translation, some items were deleted or modified to fit the Chinese context and to ensure that the participants fully understood the material. We first conducted a small-scale survey of some persons with bachelor's degrees (not in psychology) to assess their feelings regarding the items and their meanings. Based on their feedback and the ACT expert criteria, we removed items 1, 2, 7, 8, 19, and 23. These items may have been difficult for Chinese participants to understand because, in the context of the Chinese culture, the "feelings" and "thoughts" mentioned by items 1, 2, 7, and 8 and the concept of "importance" in items 19 and 23 are all for specific things or present states. The items would be more difficult for participants to understand and would be easily confused with the meaning of other items. Furthermore, after translation, the meaning of questions 19 and 23 in Chinese became identical to question 21 in the English scale. Then, we added some more items. In the English version of the CompACT, items from the "Openness to Experience" category refer to acceptance and defusion. Items from "Behavioral Awareness" relate to the awareness of the present moment. Items from "Valued Action" refer to personal values and committed active behaviors. After deleting the items mentioned above, to balance the number of questions in each stage of the CompACT, we selected items from AAQ-II, CFQ, and VQ that have been demonstrated to have good reliability and validity in the context of Chinese culture and added them to the scale. The initial items are shown in Table 2. We selected items 3 and 7 from the AAQ-II and item 6 from the CFQ to test acceptance, item 5 from the AAQ-II to test defusion, and item 13 from the VQ to describe individuals out of the present and therefore unable to take action.
The English CompACT version is scored on a 7-point Likert-like scale. The items in the Chinese version were rated on a 5-point Likert-like scale (1 = strongly disagree to 5 = strongly agree). Items 1 through 13 were also reversescored. A higher score indicates better condition.

Instruments
In addition to basic demographic information, all participants completed the Satisfaction With Life Scale (SWLS) (Xiong & Xu, 2009) and the Depression, Anxiety, and Stress Scale (DASS-21) (Lovibond & Lovibond, 1996) for the purpose of examining the relationship between CompACT results with negative emotions and well-being.
The DASS-21 is a self-report scale involving 21 items that is used to evaluate an individual's level of negative emotion within the past week. A higher score indicates more serious level of negative emotion. There are seven items per subscale: Depression, Anxiety, and Stress. The Chinese version of DASS-21 has been validated. The Cronbach's α values for the Depression, Anxiety, and Stress subscales, and that for the total scale, were 0.83, 0.80, 0.82, and 0.92, respectively . Unlike the original Chinese version of the DASS-21, in this study, participants were asked to score every item with the scale of 1 ("does not apply to me at all") to 5 ("applies to me very much"). The Cronbach's α in our case was .94 for the total score, 0.86 for Depression, 0.85 for Anxiety, and 0.83 for Stress.
The SWLS is a 5-item scale designed to assess individuals' life satisfaction. In this study, we used the Chinese version translated by Xiong and Xu (2009). The Cronbach's α was 0.78. The SWLS was rated using a Likert-7 scoring system from 1 ("absolutely incorrect") to 7 ("totally correct"). A higher score indicates that the individual is more satisfied with their life. The scale showed excellent internal reliability (Cronbach's α = 0.78). We used a standard 1-5 Likert rating criteria in the present study; the Cronbach's α was 0.85.

Statistical Analyses
EFA was conducted to explore the underlying factor structure of the scale. CFA was conducted to assess the fitting of model data. The goodness of CFA fit should be evaluated against the well-established criteria as follows: a Bentler-Bonett normed fit index, incremental fit index, comparative fit index, and Tucker-Lewis index greater than 0.90, a parsimony-adjusted comparative fit index above 0.50, and a root mean square error of approximation below 0.05 (Medsker et al., 1994). After removing items based on EFA, we calculated Cronbach's α to assess the internal consistency. We used Pearson's correlation to examine the correlations of all variables and items. SPSS 25.0 software was used to calculate reliability, relevance, and EFA; SPSS AMOS version 21 was used to calculate CFA.

EFA
The correlation coefficient between each question and the total score ranged from 0.33 (item 14) to 0.70 (item 3). After each item was deleted, the Cronbach's α values were between 0.81 and 0.87, suggesting that each item was equally necessary and important. The EFA method revealed the potential latent variables of the Chinese version of the CompACT. The items with factor loading less than.45 or communality less than.40 were deleted. Items 6 and 14 were deleted because their communality was less than.40. Finally, 18 items remained. The Kaiser-Meyer-Olkin Index (0.91) and Bartlett sphericity test (χ 2 = 16,092.237, df = 153, p < 0.001) indicated that the data were adequate to perform factor analysis. As shown in Table 3, EFA extracts factors according to the eigenvalue of each factor greater than 1, and three factors were extracted by principal component analysis with varimax rotation. No items loaded were less than 0.45, and there were no cross-loadings within.20 of a primary loading. Each factor included six items. The first factor accounted for 19.4% of the variance, the second factor accounted for 17.4%, and the third factor accounted for 17.5%. The three factors together explained 54.3% of the total variance. Table 3 shows good commonalities and loadings on the primary factor ranging from 0.527 (item 8) to 0.803 (item 11). Factors were named according to the original questionnaire: "Factor 1" is openness to experience, "factor 2" is behavioral awareness, and "factor 3" is valued action.

CFA
Having obtained the factor structure of EFA and after removing items based on EFA, we performed CFA on 2490 subjects. AMOS 21 was applied for maximumlikelihood CFA to examine the three-factor structure of CompACT. As shown in Table 4, the statistical results suggested that the models fit the actual data. This model (Model A) is shown in Fig. 1 . Because two of the three latent variables in model A are moderately correlated, we further verified that a second-order factor could explain all the first-order factors. After operation and modification, the second-order model (Model B) is shown in Fig. 2. The fitting indexes of Model B are entirely consistent with model A.

Consistency and Reliability
We then calculated the Cronbach's α coefficient for the total scale and each subscale of the Chinese version of the CompACT and found acceptable results. The Cronbach's α was 0.87 (> 0.80). The results of each subscale shown in Table 5 suggest that the scale has good reliability. Table 6 shows the correlation coefficient between the scores of each sub-scale in the Chinese version of the CompACT and those on the DASS and SWLS. We found a positive correlation between the factors of the Chinese CompACT version with SWLS and a negative correlation with DASS.

Discussion
The results confirmed our hypothesis that the Chinese CompACT could be a comprehensive instrument to evaluate the psychological flexibility of Chinese adults. Psychological flexibility has been associated with adaptive personality traits and lower negative emotionality  and plays a critical role in determining psychological health (Bai et al., 2019;Coto-Lesmes et al., 2020). Positive change is thought to result from processes associated with psychological flexibility in ACT interventions (Hayes et al., 2006;Hayes et al., 2012a). For these reasons, total measurement psychological flexibility is desirable. The Chinese CompACT we developed provides a useful instrument for studying psychological flexibility and ACT in China. We examined the psychometric properties of the developed CompACT in a Chinese adult sample. Though several items and scoring criteria in the Chinese version differ from the original, the overall findings show that the Chinese CompACT has sufficient scale reliability and validity.
Psychological flexibility is a complex and comprehensive conception. In China, prior measurements of psychological flexibility have been limited to experiential avoidance and acceptance. The CFA findings supported the three-factor structure of psychological flexibility in our Chinese civil servant sample. Using point of view of data fitting, the structure of the Chinese version of the CompACT questionnaire can be explained using either a first-order or second-order model. There was no significant difference in fitting index between the two models. However, the first-order model only shows a moderate correlation between three dimensions of psychological flexibility-openness to experience, behavioral awareness, and practical action. The second-order model further explains the source of this correlation-there is a total psychological flexibility factor above the three dimensions, which determines and influences the three dimensions below. This study developed a comprehensive measure of Chinese psychological flexibility using total scores for the first time. Unlike prior attempts, the Chinese CompACT also measures an individual's ability to pursue personal value with experiencing distress, which fits the definition of psychological flexibility. Three factors are involved in five 1 3 processes of psychological flexibility. The scores of the three subscales indicate the participants' levels of different processes of psychological flexibility. When the scale is used for psychological counseling, the subscale scores show the process of transforming the client's psychological flexibility during the counseling. As in the English version, the Chinese CompACT lacks items focused on the self as context. The reason for this is the lack of established measures within the literature from which we could extract items. Furthermore, developing the self as context is difficult to capture using verbal descriptors (Francis et al., 2016). To solve this problem, the definition of the "self as context" should be more operational and measurable.
The Chinese version of the CompACT was based on the English version, rather than directly translating from English to account for cultural differences. During the translation and selection of the items, we ensured that the meaning of each topic was clearly expressed in the Chinese context. According to the results of the small-scale interviews conducted before the formal questionnaire distribution, the topics we deleted were ambiguous and easily confused in the Chinese context. Then, to ensure the integrity of the measurement content of the scale, several items were added. In this way, the Chinese version of the questionnaire accorded with the content of the psychological flexibility and could be understood by Chinese individuals. Then, EFA was conducted, and the results showed that the Chinese version of CompACT was featured with a three-factor structure when the meaning of each factor was the same despite differences in some items between the two versions. Our findings suggest that this structure is consistent with ACT theory and represents different psychological flexibility processes. Unlike the original CompACT that does not overlap with the AAQ-II, the Chinese version of the CompACT has three items from the Chinese version of AAQ-II. In the initial revision, we removed some items that we felt were too difficult to understand in the Chinese cultural context. Therefore, it was necessary to add several items to expand the "Openness to Experience" subscale. The AAQ-II did not load on other domains of psychological flexibility other than acceptance and diffusion processes. The Chinese AAQ-II version was translated by Cao et al. (2013) with excellent reliability and validity. As the CompACT serves as a comprehensive measure of psychological flexibility, and because AAQ-II has been demonstrated to be an effective tool for measuring individual experience avoidance, on the advice of experts, we selected three items from the Chinese AAQ-II that reflect the individual's acceptance of their experience, then added them to the "Openness to Experience" subscale. These items were consistent with the subscale and the total score across the scale.
The Chinese CompACT will provide a useful instrument for studying psychological flexibility and ACT in China. According to the literature, psychological flexibility plays a  crucial role in determining psychological health (Bai et al., 2019;Coto-Lesmes et al., 2020). Positive change is thought to result from processes associated with psychological flexibility in ACT interventions (Hayes et al., 2006;Hayes et al., 2012a). Therefore, studying the complicated relationship between psychological flexibility and mental health is critical. It should be noted that there is a consistency between the results obtained in this study and previous studies-the three subscales and the total score positively correlated with the SWLS and negatively correlated with the DASS-21. Individuals with higher levels of psychological flexibility tend to be better aware of their present activities and more nonjudgmental of their emotions, feelings, and thoughts. They can take more effective actions and are less likely to experience negative emotions or complaints about their lives than less flexible individuals. The scores our sample provided on the Chinese version of the CompACT appear to accord with the extent to which psychological flexibility promotes emotional adjustment and well-being. As a special group, Chinese civil servants play a role of bridge between the general public and the government (Zhu et al. 2014); they also play a relatively important role in influencing the rise and fall of the country as a whole (Lu & Liang, 2016). As society evolves and reforms continue to deepen, the government's management of civil servants has grown increasingly standardized and stringent . Compared with people in other occupations, Chinese civil servants have been shown to be at a higher risk of suffering as a result of psychosocial stressors such as traditional bureaucratic culture, intense competition, and heavy workloads and responsibilities (Chen, 2005;Nian, 2004;Wang, 2008). Many early studies have illustrated that civil servants in various regions of China have different levels of anxiety, depression, burnout, and other psychological problems than the general population due to the specific aspects of their work and environment (Li et al., 2013;Xu et al. 2010;Hao et al., 2014;Hao et al., 2015). Ethnicity, resilience, and demographic characteristics are all considered to be factors affecting civil servants' mental health (Fu et al., 2017;Hao et al., 2015;Hong et al., 2015;Huang et al., 2018). There are a number of questionnaires on the measurement of mental health symptoms in Chinese civil servants, including the Hamilton Depression Rating Scale (HAMD) (Huang et al., 2018), the twelve-item Short Form Health Survey (SF-12) (Lu & Liang, 2016), the General Health Questionnaire (GHQ-12) , the Symptom Checklist 90 (SCL-90) (Zhu et al., 2014), and the Civil Servants Stress Scale (CSSS) . Poor psychological flexibility leads to many psychological and behavioral problems (Zhang et al., 2012), but no previous researchers have measured the psychological flexibility of Chinese civil servants. The revised Chinese version of the CompACT scale may provide an effective approach to this type of research.
This study has some limitations. First, the participants were all civil servants who lived with unique professional circumstances and work environments; therefore, the results may not generalize to other occupational groups or the general population. Further research is necessary to confirm the scale's psychometric properties and performance among more diverse populations. Many participants from across the country were selected to keep the results consistent with the Chinese context; however, the fitting index results may have room for improvement due to the large sample size and regional diversity. Future researchers can verify this questionnaire in populations with smaller sample sizes and more uniform characteristics. Finally, although the internal consistency of the questionnaire was tested using the Guttman split-half test and Cronbach's α, it remains necessary to test the Chinese CompACT's test-retest reliability.

Conclusions
The Chinese version of the CompACT was proven in this study to acceptably reveal psychometric characteristics within Chinese adults. It can be used to effectively evaluate psychological flexibility across a variety of reliable scales.