EMOTIONALLY FOCUSED THERAPY FOR COUPLES IN TAIWAN By Chi-Fang Tseng A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Human Development and Family Studies — Doctor of Philosophy 2022 ABSTRACT EMOTIONALLY FOCUSED THERAPY FOR COUPLES IN TAIWAN By Chi-Fang Tseng This is the first study to begin to explore the effectiveness and predictors of change of emotionally focused therapy for relationship satisfaction and depressive symptoms among couples in Taiwan. This one-arm pragmatic trial assessed the clinical outcomes of 17 couples using paired-samples t-tests and multilevel modeling. Paired-sample t-tests revealed no statistical differences in relationship satisfaction and depressive symptoms before and after EFT. Additionally, multilevel modeling indicated no change in relationship satisfaction over time. However, there was a quadratic change in depressive symptoms over the course of EFT. While the study results were unexpected, it is important to note that most couples did not receive the recommended treatment “dose” in this pragmatic trial. In addition, the small sample size limited statistical power. In addition to assessing treatment outcomes, predictors of change were also examined. Findings showed that high traditionalism at intake predicted an increase in relationship satisfaction for women. Attachment was also a significant predictor of change; men with high attachment avoidance at intake demonstrated a significant decrease in depression, and men and women with high attachment anxiety at intake also experienced a significant decrease in depression. Lastly, emotional expressivity at intake was associated with an increase in relationship satisfaction and a decrease in depression for both men and women. Our study suggested that traditionalism, attachment, and emotional expressivity are important predictors of change among couples in Taiwan who receive EFT. While more research is needed, these findings offer preliminary support for the types of partners who may be more likely to experience change after receiving EFT. Copyright by CHI-FANG TSENG 2022 ACKNOWLEDGEMENTS I would like to thank my parents, Mi-Ying and Jung-Shih. Thank you so much for supporting my dreams and goals since I was a child. Being your daughter makes me feel supported, loved, and cared for. After studying music for 10 years, I decided to change my path in college to study education, and after college, I changed my path again to study counseling psychology. After graduate school, I decided to study abroad to pursue my dream career in couple and family therapy. These transitions in my life have paved the way to who I am today. You never questioned my decisions but always provided reassurance and support and made me realize I can achieve anything I want. Thank you for always being proud of me. I will continue to strive to serve humanity and contribute to society. To my primary advisor, Dr. Andrea Wittenborn: Thank you so much for bringing me all the way to Michigan State University for this incredible journey. You are an amazing advisor, mentor, and friend, and your continuous support in my professional and personal life has shaped who I am today. To my committee members: Dr. Adrian Blow, your mentorship and leadership in our department have shown me what a true leader is. Thank you for supporting me every step of the way through the program. Dr. Desiree Qin, I appreciate everything you have done to encourage me, and I am so grateful for you. I have really enjoyed our conversations about the meaning of life; you have no idea of the impact those talks have had on me. Dr. Ting Liu, thank you for connecting me with so many therapists in Taiwan and being willing to work with me and trust me in our important EFT research work. I also want to express appreciation for my college professor, Dr. Wei-Wen Lin. You constantly reminded me that I am awesome and motivated me to strive during my college years. Your support, encouragement, and willingness to show up for me when I needed it meant so v much as I navigated my career goals. To my former advisor, Dr. Wen-Tao Chao: If not for you, I would not be studying in the U.S. and pursuing my dream career in couple and family therapy. Thank you for making me believe in myself and encouraging me to study abroad to pursue my dream. Thank you for always saying that you are waiting in Taiwan for me to return so we can do amazing work and research together. I also want to thank Dr. Joe Wetchler and Dr. Lorna Hecker. Being an international student in a master’s program was hard, but you both saw my strengths when I could not. I would not be in a PhD program without your support and encouragement. To my family in Taiwan—my sister, aunt, and cousins: Thank you for always being there for me. Thank you for always waiting for me to come back home. I am also grateful for my Taiwanese friends in the U.S. You are all like my family and my strongest support system. To my cohort at Michigan State University—Morgan, Melissa, Miki, Meng, Ronald, and Deb: It is such an honor to work with you all. You all have inspired me in so many ways, and I cannot tell you how grateful I am to be with you on this PhD journey. Let’s continue to do our amazing work! To my child, Tzu-Chen Harper Hung: You make my life meaningful. You were born during my PhD and raised during the pandemic. We were together 24/7, and I was so privileged and honored to have you by my side while I was doing my PhD. You are adventurous, curious, brave, and beautiful, and you remind me to continue my work to build a better world. Most importantly, to my husband, best friend, and soulmate, Shih-Chang Hung: I love that we are an amazing team. I love how we supported each other throughout our PhD journeys. Knowing that you believe in me and my work means the world to me. Thank you for supporting me to dream big and continue to do my amazing work. We finish our PhDs with tears and joy! vi TABLE OF CONTENTS LIST OF TABLES .......................................................................................................................... x LIST OF FIGURES ...................................................................................................................... xii CHAPTER I: INTRODUCTION.................................................................................................... 1 Statement of the Problem ............................................................................................................ 1 The Purpose of the Study ........................................................................................................ 3 CHAPTER II: LITERATURE REVIEW ....................................................................................... 5 Theoretical Foundation of Emotionally Focused Therapy ......................................................... 5 Emotionally Focused Therapy Process of Change ................................................................. 6 A Review of Literature on Emotionally Focused Therapy Research ......................................... 7 EFT and Emotional Expressivity ............................................................................................ 8 EFT Among Diverse Populations ........................................................................................... 9 Application of EFT to Taiwanese Couples in Taiwan .......................................................... 10 Attachment ............................................................................................................................ 11 Emotions .............................................................................................................................. 13 Other Predictors of Success in EFT ...................................................................................... 14 CHAPTER III: METHODOLOGY .............................................................................................. 16 Procedure .................................................................................................................................. 16 Participants................................................................................................................................ 17 Measures ................................................................................................................................... 18 Demographics ....................................................................................................................... 18 Patient Health Questionnaire ................................................................................................ 18 Dyadic Adjustment Scale ...................................................................................................... 19 Experiences in Close Relationships ...................................................................................... 19 Femininity Ideology Scale–Short Form ................................................................................ 20 Male Role Norms Inventory–Short Form ............................................................................. 21 The Multidimensional Scale of Chinese Individual Traditionality–Short Form .................. 21 The Multidimensional Scale of Chinese Individual Modernity—Short form ...................... 22 Emotional Expressivity Scale ............................................................................................... 22 Materials Translation ............................................................................................................ 23 Statistical Analysis .................................................................................................................... 24 Plan of analysis ..................................................................................................................... 26 CHAPTER IV: STUDY ONE ...................................................................................................... 27 Abstract ..................................................................................................................................... 27 Introduction ............................................................................................................................... 28 An Evidence-Based Couple Intervention—Emotionally Focused Therapy ............................. 29 The Application of EFT in Taiwan ....................................................................................... 30 The Purpose of the Study ...................................................................................................... 31 Method ...................................................................................................................................... 31 vii Procedures ............................................................................................................................. 32 Participants ............................................................................................................................ 33 EFT Intervention ................................................................................................................... 33 Instruments ............................................................................................................................ 34 Demographic Information ................................................................................................. 34 Dyadic Adjustment Scale .................................................................................................. 34 Patient Health Questionnaire ............................................................................................ 35 Material Translation .......................................................................................................... 35 Data Analysis ........................................................................................................................ 36 Paired-Samples T-Tests .................................................................................................... 36 Multilevel Modeling ......................................................................................................... 37 Missing Data ..................................................................................................................... 38 Results ....................................................................................................................................... 39 Paired-Samples T-Tests ........................................................................................................ 40 Multilevel Modeling ............................................................................................................. 40 Model Building-DAS Unconditional Model .................................................................... 40 Model Building-PHQ-9 Unconditional Model ................................................................. 41 Indistinguishable Dyadic Multilevel Growth Models....................................................... 42 Discussion ................................................................................................................................. 42 Limitations ............................................................................................................................ 45 Conclusions ............................................................................................................................... 46 CHAPTER V: STUDY TWO ....................................................................................................... 47 Abstract ..................................................................................................................................... 47 Introduction ............................................................................................................................... 48 Attachment ............................................................................................................................ 49 Emotional Expressivity ......................................................................................................... 51 Traditional/Modern Values and Masculinity/Femininity Ideologies .................................... 52 Current Study ........................................................................................................................ 53 Methods .................................................................................................................................... 54 Procedures ......................................................................................................................... 54 Participants........................................................................................................................ 56 Measures ................................................................................................................................56 Attachment Avoidance and Anxiety ................................................................................. 56 Femininity Ideology .......................................................................................................... 57 Masculinity Ideology ........................................................................................................ 57 Traditional Values ............................................................................................................. 58 Modern Values .................................................................................................................. 58 Emotional Expressivity ..................................................................................................... 59 Relationship Satisfaction .................................................................................................. 59 Depressive Symptoms....................................................................................................... 60 Materials Translation ............................................................................................................ 60 Data Analysis ........................................................................................................................ 61 Missing Data ......................................................................................................................... 62 Results ....................................................................................................................................... 63 Indistinguishable Dyadic Multilevel Growth Models........................................................... 64 viii Relationship Satisfaction .................................................................................................. 64 Depressive Symptoms....................................................................................................... 65 Discussion ................................................................................................................................. 67 Attachment Avoidance and Anxiety ..................................................................................... 67 Traditional Values/Modern Values and Femininity/Masculinity ......................................... 69 Emotional Expressivity ......................................................................................................... 70 Limitations ................................................................................................................................ 71 Conclusion ................................................................................................................................ 71 APPENDICES .............................................................................................................................. 73 APPENDIX A Tables ............................................................................................................... 74 APPENDIX B Figures .............................................................................................................. 92 BIBLIOGRAPHY ....................................................................................................................... 100 ix LIST OF TABLES Table 1. Descriptive Information for Relationship Satisfaction and Depression Over Time ....... 75 Table 2. Dyadic Multilevel Growth Model of Relationship Satisfaction and Depression ........... 76 Table 3. Means and Standard Deviations of Predictor Variables ................................................. 77 Table 4. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Attachment Avoidance ................................................................................. 78 Table 5. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Attachment Anxiety ...................................................................................... 79 Table 6. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Femininity Ideology ..................................................................................... 80 Table 7. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Male Role Norms.......................................................................................... 81 Table 8. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Traditional Values ........................................................................................ 82 Table 9. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Modern Values ............................................................................................. 83 Table 10. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Emotional Expressivity................................................................................. 84 Table 11. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Attachment Avoidance ................................................................................. 85 Table 12. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Attachment Anxiety ...................................................................................... 86 Table 13. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Femininity Ideology ..................................................................................... 87 Table 14. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Male Role Norms.......................................................................................... 88 Table 15. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Traditional Values ........................................................................................ 89 x Table 16. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Modern Values ............................................................................................. 90 Table 17. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Emotional Expressivity................................................................................. 91 xi LIST OF FIGURES Figure 1. Relationship Satisfaction Trajectories Over the Course of EFT. .................................. 93 Figure 2. Depressive Symptoms Trajectories Over the Course of EFT. ...................................... 94 Figure 3. Femininity Ideology Association with Rate of Change in Relationship Satisfaction Trajectory ...................................................................................................................................... 95 Figure 4. Male Role Norms Association with Rate of Change in Relationship Satisfaction Trajectory ...................................................................................................................................... 96 Figure 5. Traditional Values Association with Rate of Change in Relationship Satisfaction Trajectory ...................................................................................................................................... 97 Figure 6. Attachment Avoidance Association with Rate of Change in Depressive Symptoms Trajectory ...................................................................................................................................... 98 Figure 7. Attachment Anxiety Association with Rate of Change in Depressive Symptoms Trajectory ...................................................................................................................................... 99 xii CHAPTER I: INTRODUCTION Statement of the Problem Approximately 5% of adults around the world suffer from depression (World Health Organization, 2021). Globally, the number of people living with depression increased by 18.4% from 2005–2015, and depression is now the world’s leading cause of medical disability (World Health Organization, 2017). Depression has the potential to cause severe symptoms that affect one’s cognitions, emotions, relationships, physical health, and daily functioning (National Institute of Mental Health, 2019). Depression is a debilitating condition, and it is therefore essential to examine its related risk factors. Numerous studies have found a reciprocal relationship between relationship satisfaction and depression (Atkins et al., 2009; Beach et al., 2003; Whisman et al., 2004), and distressed couples are 25 times more likely to suffer from depression than couples who are not distressed (Weissman, 1987). Relationship dissatisfaction or distress has also been found to predict an increase in depression over time (Beach & O’Leary, 1993) and to increase the likelihood of experiencing a major depressive episode in a given year (Whisman & Bruce, 1999). Relationship distress is common and can have a detrimental effect on physical and emotional health. Relationship dissatisfaction has been found to contribute to lowered productivity at work, which in turn increases the risk of economic hardship and chronic stress (Robles et al., 2014). It has also been found to increase the risk of mortality, cardiovascular disease, and negative health outcomes (Liu & Waite, 2014; Robles et al., 2014). Relationship distress is a modifiable mechanism of depression, and improving relationship problems may improve depression (e.g., Wittenborn et al., 2019). 1 Couple therapy plays an important role in treating couple distress and depression. Research findings have indicated that couple therapy is as effective as individual therapy in treating depression, and it is more effective than individual therapy at improving relationship satisfaction (Barbato et al., 2018). In Asia, mental health problems are seldom recognized by society due to a lack of access to services and social stigma (Kudva et al., 2020; Lauber & Rössler, 2007). There is an urgent need for evidence-based clinical interventions to increase relationship health and decrease depressive symptoms. Without such interventions, couples in Asia will continue to suffer the negative consequences of untreated depression and relationship problems. Leveraging couple therapy as an evidence-based intervention that can increase relationship health and decrease depressive symptoms presents a critical opportunity to improve mental health and relationship health in Asia. Emotionally Focused Therapy (EFT) is an empirically supported couple therapy approach developed in the early 1980s with the goal of treating relationship distress among couples (Johnson & Greenberg, 1985). Couples who participated in EFT have reported an increase in relationship satisfaction following treatment (e.g., Johnson & Greenberg, 1985). Dalgleish et al. (2015) recruited 32 couples to undergo EFT and reported an overall increase of 0.39 points in relationship satisfaction per weekly EFT session. Furthermore, a meta-analysis of 20 studies that included 332 couples found medium to large effects for pre-test to post-test, EFT versus other couple interventions, and pre-test to follow-up (Spengler, Lee, Wiebe, & Wittenborn, in press). EFT has also been found to improve depression (Denton et al., 2012; Dessaulles et al., 2003; Wittenborn et al., 2019). Studies have compared the efficacy of EFT in reducing partners’ depression to that of pharmacotherapy treatment (Dessaulles et al., 2003) and the efficacy of EFT plus pharmacotherapy to pharmacotherapy alone (Denton et al., 2012). Dessaulles et al. (2003) 2 found that EFT and pharmacotherapy both reduced depressive symptoms in women at post- treatment, but only those who received EFT showed greater improvement than pharmacotherapy condition at six-month follow-up. Separately, Denton et al. (2012) compared the use of EFT with medication to the use of medication alone and found that women in both conditions reported improved depressive symptoms over six months of treatment, but only women in the antidepressant plus EFT condition reported more improvements in relationship quality compared with women in the antidepressant condition. Finally, a randomized controlled trial compared the effectiveness of EFT with usual care (i.e., other couple therapy approaches) and found that men and women in both groups reported a significant decrease in depressive symptoms over the course of treatment, while couples in the EFT group reported a significantly greater increase in relationship satisfaction compared to usual care (Wittenborn et al., 2019). The majority of research on EFT has been conducted in Western cultural contexts. Based on the existing findings, EFT is expected to be an effective treatment method for couples in Taiwan because attachment and emotions—two fundamental concepts of EFT—are experienced across cultures (Ekman, 2007; Johnson, 2019). In Iran, EFT has shown to be an efficacious approach to improve depression and marital satisfaction in Iranian couples (Ahmadi et al., 2014; Soleimani et al., 2015; A. Soltani et al., 2013; M. Soltani et al., 2014). However, research on the effectiveness of EFT among Taiwanese couples in Taiwan is limited. The Purpose of the Study Most EFT research has been conducted in controlled settings (e.g., Dandeneau & Johnson, 1994; Denton et al., 2000; Denton et al., 2012; Dessaulles et al., 2003; Johnson & Greenberg, 1985; Johnson & Talitman, 1997). To date, few research studies have tested the effectiveness of EFT in common practice settings (e.g., Ahmadi et al., 2014; Wittenborn et al., 3 2019). Moreover, most EFT studies have been conducted in Canada and the United States among predominantly Western and non-Asian samples, though a small body of work has tested the efficacy of EFT among Iranian couples. Additional evidence is needed to examine whether findings on EFT with Western couples are similar among non-Western populations in non- Western contexts. The first research question this study will examine is: Does EFT improve clinical outcomes (i.e. relationship satisfaction and depression) among Taiwanese couples in Taiwan? Based on the assumption of the universality of attachment, EFT is thought to be applicable to couples across cultures. However, previous research studies have indicated that Asian couples might display avoidance and anxious attachment more than Western couples as these attachment styles are assumed to be more culturally acceptable (Wang & Mallinckrodt, 2006). Moreover, traditional values may persist in Chinese culture despite the rapid economic growth and modernization over the past few decades (Inglehart & Baker, 2000; Yang, 1998). The differences between traditional and modern values as well as ideologies of femininity and masculinity may also be important factors in the prediction of change in EFT. Therefore, the second research question aims to explore: Do attachment style, traditional/modern values, femininity/masculinity ideology, and emotional expressivity at baseline predict growth in relationship satisfaction and decrease in depressive symptoms among Taiwanese couples in Taiwan who receive EFT? 4 CHAPTER II: LITERATURE REVIEW Theoretical Foundation of Emotionally Focused Therapy EFT combines systemic and experiential approaches, viewing relationship distress from the lens of Bowlby’s (1969) attachment theory (Johnson, 2019). In attachment theory, seeking contact with attachment figures is considered an innate survival mechanism. Individuals’ attachment needs for comfort are activated during stressful situations and people seek comfort from important loved ones to soothe their stress and return to normal, autonomous activity (Bowlby, 1988; Johnson, 2019; McFarlane & van der Kolk, 1996). A two-dimensional framework has been used to understand and categorize patterns of attachment. In this framework, attachment behaviors and responses are organized across the two dimensions of anxiety and avoidance (Fraley et al., 2000). Securely-attached individuals, characterized by holding a sense of worthiness and an expectation that others are responsive, have low attachment anxiety (i.e., they view themselves positively) and low attachment avoidance (i.e., they view their partners positively) (Bartholomew & Horowitz, 1991). Insecure attachment develops when attachment figures are not emotionally accessible and responsive; in these situations, people develop high attachment anxiety or high attachment avoidance. High attachment anxiety includes hyperactivating strategies such as blaming or criticizing, and high attachment avoidance is characterized by withdrawing or distancing from one’s partner (Johnson, 2019). Previous research has explored attachment as a predictor of relationship satisfaction in EFT. In examining predictors of success in EFT, Johnson and Talitman (1997) found that the overall level of attachment security did not predict relationship satisfaction after couples received EFT. However, male partners with higher levels of proximity-seeking behaviors (a subscale of 5 the attachment security measure) at the beginning of EFT were more likely than female partners to report higher relationship satisfaction at the end of treatment (Johnson & Talitman, 1997). Dalgleish et al. (2015) examined predictors of success in EFT and found that individuals with higher initial levels of attachment anxiety experienced greater changes in marital satisfaction during the course of EFT. However, participants’ initial levels of attachment avoidance did not predict changes in EFT (Dalgleish et al., 2015). A similar study found that attachment avoidance decreased over the course of EFT (Burgess Moser et al., 2016). These findings suggest that EFT might be particularly beneficial for people with high attachment anxiety. Pre-therapy attachment avoidance levels may not determine the success of EFT, but attachment avoidance levels could decrease over time. Emotionally Focused Therapy Process of Change EFT consists of three stages: cycle de-escalation, changing interactional positions, and consolidation and integration of therapeutic gains (Johnson, 2019). In the first stage, cycle de- escalation, therapists conceptualize couples’ issues using an attachment framework and assist couples to develop an understanding of their negative interactional cycles. Therapists continue to track and identify the negative interactional cycles and access unrecognized emotions underlying these cycles. During stage one, therapists help couples reframe their problems in terms of unmet attachment needs, unacknowledged underlying emotions, and negative interaction cycles. By the end of stage one, a hostile partner might become less reactive, and the withdrawn partner may be more emotionally engaged (Johnson, 2019). In the second stage, changing interactional positions, partners start to identify and express disowned attachment needs and also support their partners’ expression of attachment needs (Johnson, 2019). Therapists’ tasks in this stage are to guide couples to respond and support 6 each other in an emotionally attuned way. For example, after a partner shares attachment vulnerabilities and needs in a session, a therapist might support the other partner in processing their experience of hearing their partner’s vulnerable emotions and supporting them in responding to their partner. This process leads to a shift in interactional positions, which creates new emotional engagement and bonding events (Johnson, 2019). The third stage of EFT, consolidation and integration, involves helping couples find new solutions to long-standing issues. The goal of this stage is to have couples leave therapy not only non-distressed, but able to maintain emotional engagement that will allow them to continue to strengthen their bond after therapy concludes (Johnson, 2019). At the termination phase, the relationship becomes a secure base, and relationship issues no longer trigger attachment insecurity and evoke negative interactional cycles (Johnson, 2019). A Review of Literature on Emotionally Focused Therapy Research Numerous studies have demonstrated the effectiveness of EFT among couples. In a meta-analysis of four EFT outcome studies conducted prior to 1999, Johnson et al. (1999) found a large effect size of 1.31, suggesting that EFT is an effective approach to couple therapy. Another meta-analysis, including 23 couple intervention studies published between 1963 and 2002, focused on which couple treatment was most effective for different marital distress levels: mild, moderate, and severe (Wood et al., 2005). The meta-analysis showed that EFT demonstrated better effectiveness compared to Behavioral Couple Therapy in treating couples with moderate levels of marital distress (Wood et al., 2005). In an extension of Johnson et al.’s (1999) meta-analysis of EFT outcome studies, Beasley and Ager (2019) conducted a systematic review of EFT outcome research that was published between 1999 and 2017. A total of nine studies were included in this review, including the application of EFT among Iranian couples 7 (Ahmadi et al., 2014), infertile Iranian couples (Soleimani et al., 2015; Soltani et al., 2014), couples with chronically ill children (Cloutier et al., 2002), couples facing medical conditions such as breast cancer (Naaman, 2009) and end-stage cancer (McLean et al., 2013), couples with psychological difficulties such as depression (Denton et al., 2012; Dessaulles et al., 2003), childhood trauma (Dalton et al., 2013), and somatoform disorders (Walsh, 2002). The systematic review concluded that EFT helped couples increase their marital satisfaction after receiving treatment, with a large effect size of 2.09, and the improvement in marital satisfaction was maintained at follow up (Beasley & Ager, 2019). The most recent meta-analysis included 20 studies totaling 332 couples and found medium to large effects for EFT pre- to post-test, pre-test to follow-up, and in comparison to other couple interventions (Spengler et al., in press). EFT and Emotional Expressivity The primary function of emotions is to communicate one’s needs and motives (Frijda, 1986). In EFT, emotional experiences and expressions are critical to treatment because they facilitate the process of unfolding key emotions among couples to identify and express attachment needs (Johnson, 2019). In a study of predictors of success in EFT, Johnson and Talitman (1997) found that male partners described by their partners as emotionally inexpressive at intake made the most improvement in relationship satisfaction at treatment termination. A hallmark of EFT is the establishment of a safe environment in which couples can feel safe to reveal their vulnerable emotions and needs, and it is possible that this experience facilitated greater growth for these partners who were less apt to express their emotions and attachment needs prior to initiating therapy. In another study, Motaharinasab et al. (2016) examined the effect of EFT on emotional expressivity among married women in Iran. They found that women showed a significant increase in emotional expressivity after completing ten EFT sessions. 8 However, as previously discussed, restricted expression of emotions is considered culturally appropriate in Asian cultures (Lim, 2016; Ryder et al., 2008). In Chinese culture, holding back emotional expression is encouraged and preferred in interpersonal relationships (Tsai & Levenson, 1997). Li (2006) found that tolerance (i.e. suppressing one’s emotions) may be seen as a passive and indirect way to handle conflicts in the Western culture, but it is considered an appropriate emotion regulation strategy in Chinese marriage. In Li and Hsiao’s (2016) recent study to investigate the effect of tolerance on marital satisfaction among 1,173 individuals in Taiwan, they found that tolerance in marriage is related to increased marital satisfaction among men. They concluded that tolerance is an implicit method of emotional expressivity that plays an important role in the maintenance of marital satisfaction among Chinese couples (Li & Hsiao, 2016). In EFT, it is assumed that couples will increase their emotional expressivity toward each other throughout the course of treatment to help them access unrecognized emotions and to restructure interaction patterns. EFT Among Diverse Populations In a review of EFT research studies from 1985 to 2016, Wiebe and Johnson (2016) found a lack of empirical research of EFT among couples across different cultures. EFT empirical studies on non-Western populations are scarce. In a small body of outcome studies among Iranian couples in Iran, EFT has been found to reduce depression and anxiety (Soltani et al., 2014), increase marital satisfaction (Soleimani et al., 2015b), increase intimacy (Soltani et al., 2013), and decrease marital conflicts (Ahmadi et al., 2014). These studies confirmed the effectiveness of EFT in treating couples in Iran. However, EFT research in Taiwan is still in its infancy. Liu and Hung (2019) conducted a case study of a Taiwanese couple in Taiwan who completed 14 sessions of EFT. In this case 9 study, the clinical process of the application of EFT skills and case conceptualization were analyzed (Liu & Hung, 2019). In another study to explore the experiences of Taiwanese couples who received EFT in Taiwan, Sun (2019) interviewed six unmarried couples who completed six EFT sessions with certified EFT therapists in Taiwan to understand their experiences of EFT. Couples shared that EFT helped them to understand and reframe partners’ behaviors from the perspective of attachment needs, recognize negative interactional cycles, increase acceptance when handling conflicts, and increase an understanding of partners’ attachment needs through an understanding of their family of origin (Sun, 2019). Thus, EFT research in Taiwan has been limited to case study (Liu & Hung, 2019), qualitative interviews of couples’ experiences of receiving EFT (Sun, 2019), clinical application (Wang & Wang, 2015), and the effectiveness of EFT in the form of the psychoeducational “Hold me Tight” group (Sun & Liu, 2015). While there is limited research on the efficacy of EFT in Taiwan, therapists are increasingly seeking training in EFT and couples are receiving EFT treatment (Asia EFT, n.d.). Therefore, it is important to carefully examine the effectiveness of EFT for couples in Taiwan. Application of EFT to Taiwanese Couples in Taiwan The divorce rate in Taiwan has continued to rise in the last four decades; in 2020, the divorce rate was 10%, the second highest in Asia (Ministry of the Interior Taiwan Department of Statistics, 2021). The prevalence of depression is also high. In 2002, approximately 8.9% of Taiwanese individuals aged 15 and over had mild depression and 5.2% had major depressive disorder (Ministry of Health and Welfare, 2004). More recent data shows that an estimate of 5.9% of Taiwanese individuals were prescribed antidepressants in 2019 (Ministry of Health and Welfare, 2020). However, only 20% of adults with depression reported seeking any kind of treatment for their depression in the previous year (Liao et al., 2012). Given the reciprocal 10 relationship between relationship distress and depression, it is imperative to provide an evidence- based treatment to increase relationship health and decrease depressive symptoms. EFT was developed by Sue Johnson in Canada, and it is important to examine whether EFT is appropriate for couples in Taiwan. Clinical interventions developed in a Euro-American cultural context are grounded in Euro-American cultural norms. It is potentially problematic to apply such interventions to populations outside of North America without thorough examination of the theoretical foundation and rationale of the intervention (Bernal et al., 2009). Attachment and emotions, the two central concepts of EFT, are universally experienced across cultures (Ekman, 2007; Johnson, 2019). Given this assumed universality, the following sections will discuss considerations for applying EFT to Taiwanese couples in Taiwan. Attachment In 1967, Mary Ainsworth first proposed the existence of different attachment patterns, and she supported this theory with the universality of infant-mother attachment relationships observed in her study in Uganda. In a 2008 review of 21 studies conducted in Africa, China, Israel, Japan, and Indonesia, van IJzendoorn and Sagi-Schwartz found that the attachment phenomenon between infants and mothers and different types of attachment existed in every culture. In addition to the universality of infant-mother attachment, numerous studies also supported that attachment in adulthood is universal. In a meta-analysis of 33 studies—including samples of adolescents, fathers, and mothers from different countries who completed more than 2,000 Adult Attachment Interviews—van IJzendoorn and Bakermans-Kranenburg (1996) found that the classifications of autonomous, dismissing, and preoccupied attachment were applicable across cultures. 11 Attachment theory is considered applicable to individuals around the world; however, several researchers have argued against its universality because the assumptions of attachment constructs are derived from Western cultures (Rothbaum et al., 2000; Wang & Mallinckrodt, 2006). In a large-scale study of more than 17,000 participants from 62 cultures to examine how adult romantic attachment varied across cultures, Schmitt et al. (2004) reported that secure adult attachment was normative in 79% of the 62 cultures examined, and preoccupied attachment was more prevalent in East Asian cultures than in other regions. Similarly, Ditommaso et al. (2005) found that Chinese college students had significantly lower attachment security than Canadian college students. In another study, 280 Taiwanese college students and 268 U.S. college students were asked to complete the Experiences in Close Relationship Scale, a self-report measure of adult attachment that asks participants to respond to each item according to their belief of “how an ideally emotionally and psychologically healthy person of your own gender in your culture would answer” (Wang & Mallinckrodt, 2006, p. 195). The results indicated that Taiwanese men and women idealized more avoidant behaviors than U.S. men and women, and Taiwanese men idealized more anxiety behaviors than U.S. men, though not U.S. women (Wang & Mallinckrodt, 2006). In countries with high cultural acceptance of avoidant and anxious attachment, couples may cultivate these patterns of attachment because they are part of the social norm. Attachment is considered a universal experience; however, expression of preferred attachment patterns may differ across cultures. It is expected that EFT will be an effective intervention with Asian couples given that its development is based on attachment theory. However, EFT therapists need to be sensitive to how Asian couples display attachment needs and express love and connection. 12 Emotions The word emotion is derived from the Latin emovere, which means to “move out, remove, or agitate” (Online Etymology Dictionary, n.d.). Emotion is viewed as adaptive, and its functions are to communicate one’s needs, and to process and organize experiences and perceptions about self and others (Bowlby, 1991; Johnson, 2019). Emotion occurs in the form of automatic appraisal, meaning that it happens automatically and an individual cannot choose whether to feel an emotion (Ekman, 2007). In addition to the universal function of emotions, Ekman (2007) proposed that there are seven universal emotions that cross language, cultural, and racial barriers: anger, contempt, disgust, enjoyment, fear, sadness, and surprise. EFT was developed in the context of Western culture, but its application to Asian couples may be appropriate given that core emotions are experienced across cultures. Using emotions to shape the meaning of narratives and communicate with one another is also considered a universal experience (Johnson, 2019). Emotional expression occurs across cultures, yet individual displays of emotion vary among different cultures. Ekman (2018) proposed the concept of display rules regarding how and to whom one can show their emotions. These display rules are socially and culturally constructed and develop over time. Display rules can “specify that an emotion be suppressed, de-amplified, exaggerated, or even masked altogether” (Ekman, 2018, para 1). For example, moderating and controlling one’s emotional expressivity is seemingly the cultural norm in Asia (Ryder et al., 2008). Expression of high arousal emotions (e.g., excitement or anger) is discouraged, and it is considered more culturally appropriate to express emotions moderately (Lim, 2016). Due to this cultural norm, it is common for individuals from Asian cultures to suppress or indirectly express their high arousal emotions (Matsumoto et al., 2008). 13 To ensure EFT therapists maintain sensitivity to cultural differences of attachment needs and emotions, Liu and Wittenborn (2011) proposed guidelines for therapists in working with diverse populations. First, it is the therapist’s responsibility to recognize the meanings and functions related to clients’ emotional expressions and attachment needs. Therapists must familiarize themselves with the display rules their clients use to show and process emotions (Liu & Wittenborn, 2011). Second, therapists need to strive to use words and metaphors that relate to clients’ emotional experiences (Liu & Wittenborn, 2011). The application of EFT to couples in Asia is appropriate given that attachment and emotions are experienced across cultures, and it is critical that therapists are mindful that attachment and emotions can be expressed differently in diverse cultural contexts. Other Predictors of Success in EFT Levels of masculinity/femininity and traditionality/modernity affect partners’ perceptions of intimate relationships and openness to emotion expressivity. In EFT, emotional expressivity is considered to be the fundamental process of successful EFT outcomes. Johnson and Talitman (1997) found that partners with higher traditionality might have extreme pursue/withdraw interactional patterns that make emotional vulnerability more difficult to access during therapy. A small body of research has examined the association between relationship satisfaction and partners’ levels of masculinity and femininity. Snyder et al. (1993) explored predictors of couple therapy outcomes and found that lower femininity scores at intake predicted higher marital distress at four-year follow up. However, some studies found femininity and masculinity did not predict relationship satisfaction after couple therapy (Jacobson et al., 1986). Partners’ levels of traditionality and modernity are also examined in several couple therapy intervention studies. Snyder et al. (1993) found the higher the level of traditionality at 14 couple therapy intake, the lower the relationship satisfaction post-treatment. However, Johnson and Talitman (1997) found that traditionality was not associated with relationship satisfaction in their EFT study. In a study examining the relationship between traditionality/modernity and relationship satisfaction among 150 married couples in Taiwan, Kao and Lu (2006) found that when husbands had higher traditionality than their wives, both husbands and wives reported poorer marital satisfaction. In addition, when wives had higher modernity than their husbands, both partners reported poorer marital satisfaction (Kao & Lu, 2006). 15 CHAPTER III: METHODOLOGY In the current study, I conducted a single-arm pragmatic clinical trial (Thorpe et al., 2009) of EFT for couples in Taiwan. Rather than conducting this study in a laboratory setting, I collaborated with community-based therapists who implemented EFT with couples in Taiwan. To do this, assessments of relationship distress and depression were administered to couples after each couple therapy session. This study also examined predictors of change to help researchers and clinicians understand factors that are associated with success or lack of response to EFT among couples in Taiwan. Procedure Certified EFT couple therapists in Taiwan were located through the Asia EFT website (www.asia-eft.com). To facilitate the recruitment process, I also collaborated with the coordinator and director of the Asia EFT Center. I contacted EFT couple therapists through email to invite them to collaborate on the study. Each therapist then invited their couple clients who met the inclusion criteria to participate in the study, and couple clients who showed interest in participating in the study reviewed the research participant information and consent form. Once couples signed the consent form, they were officially enrolled in the study. Therapists were contacted to collaborate in the study if they met the following criteria: (1) completed the requirements for certification as an EFT therapist by the International Centre for Excellence in Emotionally Focused Therapy (ICCEFT), (2) practiced EFT in Taiwan, (3) received supervision from EFT certified supervisors biweekly during the time of this research phase, and (4) worked with Taiwanese couples. A total of 14 therapists were contacted, and they all expressed interest in assisting with recruitment for the current study. However, four therapists did not recruit couples for this study due to confidentiality concerns among clients. One therapist 16 stated that her couple clients had concerns regarding whether confidentiality could be maintained. The other three therapists’ clinics had concerns over therapists recruiting participants for research in the clinic, worrying that it might affect clients’ perceptions regarding confidentiality and willingness to receive therapy in their clinics. Therefore, a total of 10 therapists assisted with recruitment in the current study. All couple participants were asked to complete EFT therapy sessions in their therapists’ practice settings. In addition to the baseline surveys they completed after the first session, couples also had to complete questionnaires after each session. They could complete assessments through online surveys or paper-pencil surveys in their therapists’ offices immediately following the completion of each therapy session. Couples were able to access online surveys through their smartphones; if smartphones were not accessible, therapists provided paper-pencil surveys. Couple participants each received NT$100 gift voucher (approximately $3 U.S. dollars) after every 3 sessions they completed (i.e., each partner received NT$100 after each 3 sessions upon completion of their third assessment) until the termination of therapy. Therapists received a NT$700 (approximately $23 U.S. dollar) gift voucher for each couple for their assistance. All the gift vouchers were sent electronically to couples and therapists by emails or texts. Participants Couple participants were eligible if (1) they enrolled in couple therapy with a participating EFT therapist in Taiwan, (2) each partner was age 18 or older, and (3) both partners were willing to participate in the study. Couples who met the following criteria were not included in this research: (1) clinically significant past or ongoing intimate partner violence, (2) reporting infidelity in the past three months, or (3) reporting severe suicidal ideation in the past three months. Exclusion criteria were assessed by the EFT therapists. Expressions of 17 vulnerability is a key to change in EFT; intimate partner violence, infidelity, and suicidal ideation can impede or deteriorate the therapeutic process given that sharing vulnerability is likely to be unsafe, dysfunctional, or ineffective (Johnson, 2019). Individuals with severe suicidal ideation also require immediate crisis intervention and management, which was not offered in this study. A total of 19 couples expressed interest in participating in this study after therapists explained the research information and process. Of those, 17 couples signed the consent forms and completed assessments after each therapy session. Thus, 17 couples were included in the analysis of this study. Measures Demographics Participants responded to demographic questions at the intake session only, including (1) age, (2) gender, (3) education level, (4) combined family income, (5) marital status, (6) years in current relationship, (7) years married, (8) number of children, (9) whether they have received couple therapy before, and (10) whether they were receiving any additional treatment at intake, including medication, individual therapy, or group therapy. Patient Health Questionnaire Participants completed the Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001) at intake and after each therapy session to assess depressive symptoms. This measure uses a 4-point Likert-type scale (0 = not at all, 3 = nearly every day) to ask participants to rate if they have been bothered by any of the following problems in the last 2 weeks, such as little interest or pleasure in doing things. The PHQ-9 score was obtained by adding the scores from the nine items, with higher scores indicating higher levels of depressive symptoms. For this study I used a version of 18 the PHQ-9 that was translated into Chinese and was tested with Chinese populations with a reliability of .86 and validity of .92 (Wang et al., 2014). Cronbach’s alpha in the current study ranged from .88 to .98 for men and .56 to .96 for women. Dyadic Adjustment Scale Participants completed the Dyadic Adjustment Scale (DAS; Spanier, 1976) at intake and after each therapy session to assess their intimate relationship adjustment. The DAS has four subscales: consensus (13 items), satisfaction (10 items), cohesion (5 items), and affectional expression (4 items). The total score (range of 0 to 151) is obtained by summing item responses, with higher scores indicating better adjustment to one’s relationship (Spanier, 1976). The DAS demonstrates high reliability, with a Cronbach’s alpha coefficient of .96 (Spanier, 1976). The DAS version used in this study was previously translated into Chinese and has been validated in various studies (Shek, 1995; Shek et al., 1993; Shek & Cheung, 2008). The Chinese version DAS showed high reliability with a Cronbach’s alpha coefficient of .91 (Shek et al., 1993). Cronbach’s alpha in the current study ranged from .92 to .98 for men and .87 to .98 for women. Experiences in Close Relationships Participants completed the Experiences in Close Relationships Relationship Structures Questionnaire (ECR-RS; Fraley et al., 2011) at their intake sessions. The ECR-RS is derived from the Experiences in Close Relationships—Revised Inventory (ECR-R; Fraley et al., 2000), and aims to assess individuals’ attachment styles on four relational domains: relationships with mother, father, romantic partner, and best friend. Each domain consists of the same 9 items assessing attachment-related anxiety and avoidance (Fraley et al., 2011). Only the romantic partner domain was used in this study. Participants were asked to rate on a 7-point scale (1 = strongly disagree; 7 = strongly agree) in regards to the extent to which they worry their partner 19 may reject them (anxiety dimension, i.e., I often worry that this person doesn’t really care for me) and how uncomfortable they are to open up and depend on their partner (avoidance dimension, i.e., It helps to turn to this person in times of need) (Fraley et al., 2000). The avoidance and anxiety scores were computed separately; the avoidance score was computed by averaging items 1 to 6 (after reverse coding items 1 to 4), and the anxiety score was obtained by averaging items 7 to 9, with higher scores indicating more avoidance or anxiety. Cronbach’s alphas in previous research were .87 (avoidance domain) and .91 (anxiety domain) (Fraley et al., 2011). Cronbach’s alpha in the current study was .90 for men and .45 for women regarding attachment avoidance, and for attachment anxiety, it was .70 for men and .91 for women. Femininity Ideology Scale–Short Form Participants completed the Femininity Ideology Scale Short Form (FIS–SF; Levant et al., 2017) at their intake sessions. This is a 12-item measure that uses a 5 point Likert-type scale (1 = strongly disagree; 5 = strongly agree) to examine traditional femininity ideology. Three subscales include statements to measure how women should or should not behave, think, and feel: dependence/deference (e.g., Women should not succeed in the business world because men will not want to marry them), purity (e.g., Woman should not read pornographic magazines), and emotionality and traditional roles (e.g., It is expected that women will have a hard time handling stress without getting emotional). Total scores range between 12 to 60, with higher scores indicating higher endorsement of traditional femininity ideology. Cronbach’s alphas for subscales ranged from .82 to .88 and the total scale was .85 in prior research (Levent et al., 2017). Cronbach’s alpha in the current study was .84 for men and .77 for women. 20 Male Role Norms Inventory–Short Form Participants completed the Male Role Norms Inventory Short Form (MRNI–SF; Levant et al., 2013) at intake sessions. The MRNI–SF is a 21-item measure that uses a 7-point Likert- type scale (1 = strongly disagree, 7 = strongly agree) with total sum scores ranging between 21 to 147; higher scores indicate greater levels of endorsement of traditional masculinity ideology. MRNI-SF includes 7 subscales: avoidance of femininity (e.g., Boys should prefer to play with trucks rather than dolls), negativity toward sexual minorities (e.g., All homosexual bars should be closed down), self-reliance through mechanical skills (e.g., Men should have home improvement skills), toughness (e.g., I think a young man should try to be physically tough, even if he is not big), dominance (e.g., A man should always be the boss), importance of sex (e.g., Men should always like to have sex), and restrictive emotionality (e.g. A man should never admit when others hurt his feelings). MRNI-SF was developed by selecting three of the highest loadings for each factor from the Male Role Norms Inventory-Revised scale (Levant et al., 2007). A Chinese version of MRNI-SF has been developed and tested (Liu et al., 2018), and was used in this study Cronbach alphas were .92 for men and .94 for women for the total scale; for subscales, alphas ranged from .79 to .90 for men and .75 to .88 for women (Levent et al., 2013). Cronbach’s alpha for the Chinese version was .86 (Liu et al., 2018). Cronbach’s alpha in the current study was .86 for men and .91 for women. The Multidimensional Scale of Chinese Individual Traditionality–Short Form Participants completed the Multidimensional Scale of Chinese Individuals Traditionality Short Form (MS-CIT–SF; Kao & Lu, 2006) at intake sessions. This scale was originally developed by a Taiwanese researcher (Yang, 1996) and included 50 items. MS-CIT–SF is a 15- item Chinese-language measure that uses a 6-point Likert-type scale (1 = strongly disagree, 6 = 21 strongly agree) with total sum scores ranging from 15 to 90; higher scores indicate greater levels of endorsement of traditionality. This measure includes five subscales: submission to authority, filial piety and ancestral worship, conservatism and endurance, fatalism and defensiveness, and male dominance. The Cronbach’s alphas with a Taiwanese sample was .81 (Kao & Lu, 2006). Cronbach’s alpha in the current study was .89 for men and .94 for women. The Multidimensional Scale of Chinese Individual Modernity—Short form Couple participants completed the Multidimensional Scale of Chinese Individual Modernity Short Form (MS-CIM–SF; Kao & Lu, 2006) at intake sessions. This measure is a 15- item Chinese-language measure that uses a 6-point Likert-type scale (1 = strongly disagree, 6 = strongly agree) with total sum scores ranging between 15 to 90; higher scores indicate greater levels of endorsement of modernity. Same as the MS-CIT–SF, this scale was originally developed with 50 items (Yang, 1996). This measure includes five subscales: Egalitarianism and open-mindedness, independence and self-reliance, optimism and assertiveness, affective hedonism, and sex equality. The Cronbach’s alpha with a Taiwanese sample was .71 (Kao & Lu, 2006). Cronbach’s alpha in the current study was .55 for men and .83 for women. Emotional Expressivity Scale Participants completed the 17-item Emotional Expressivity Scale (EES; Kring et al., 1994) at intake to assess participants’ emotional expressivity. Participants were required to respond to a 6-point Likert-type scale (1 = never true, 6 = always true), with higher sum scores indicating a higher level of emotional expression. Sample items include, “I keep my feelings to myself”, “My partner is not easily able to observe what I am feeling”, and “I don’t like to let my partner see how I am feeling”. This scale was translated and tested with Chinese populations in 22 China with a high reliability of .82 (Chan et al., 2010). The Chinese version of the EES was used in this study. Cronbach’s alpha in the current study was .75 for men and .73 for women. Materials Translation The target population of this study was Chinese-speaking Taiwanese couples in Taiwan. Therefore, the study materials, including recruitment flyers, informed consents, and some English measures, were all translated into Chinese. When translating supporting materials from English to Chinese, it is important to ensure the translation is “reliable, complete, accurate, and culturally appropriate” (Pan & Puente, 2005, p.5). In addition, translated measurements should also have semantic equivalence (i.e. words in English and Chinese represent the same meaning), conceptual equivalence (i.e. the commonality of the concept between English and Chinese), and normative equivalence (i.e. the translated measure should address the difference in social norms between two cultures) (Behling & Law, 2000; Pan & Puente, 2005). Geisinger (1994) proposed steps for measurement translation and addressed the importance of the inclusion of qualified translators who are fluent in both languages, extremely familiar with both cultures, and knowledgeable about the instrument. To meet this rigorous requirement, I completed all English to Chinese translations myself; I am a doctoral student studying couple and family therapy in the U.S. who is bilingual in both the Chinese and English language and is familiar with both cultures. The translation started by translating the documents and instruments from English into Chinese with detailed records of the specific translation challenges encountered and decisions made to deal with those challenges. After I completed the translation, I had two translator reviewers who are also native Chinese speakers with high English proficiency review the documents to ensure the text and semantics are accurate, to identify problematic areas, and to provide suggestions for improvements. After translator 23 reviewers provided suggestions, I modified the translations based on their suggestions and developed a final version of the translated documents. In this study, only FIS-SF, MRNI-SF, and EE were translated into Chinese as the rest of the assessments either had translated Chinese versions or were developed in the Chinese language. In addition to translation, the adaptation of an instrument on a question-by-question basis may be needed to ensure the question is culturally meaningful to the target population (Geisinger, 1994; Tran, 2009). In this study, some questions in the MRNI-SF had to be adapted. For example, the question “The President of the U.S. should always be a man” was revised to “The President of Taiwan should always be a man”. Another question, “Men should watch football games instead of soap operas”, was adapted and changed to “Men should watch baseball games instead of soap operas”. Watching football games is a popular activity in the U.S., while in Taiwan, football is not often played and is overshadowed by baseball in terms of popularity. Statistical Analysis To answer the research questions, I employed dyadic growth curve modeling to examine change over time and the nature of the change (linear or curvilinear); analyses were also used to understand baseline predictors of change over time. Since this study included a sample of both partners, a dyadic growth curve model was used to examine “how change over time is coordinated across two partners” (Kashy et al., 2008, p. 317). Dyadic growth modeling approaches are useful in examining dyadic longitudinal data as it can evaluate whether the rate of change for one partner is associated with the rate of change for the other partner (Kashy & Donnellan, 2008). Another advantage of using dyadic growth modeling is that it accounts for the nonindependence of the data (Kenny et al., 2006). That is, the data from each member of the dyad might be correlated. For example, in our study examining relationship satisfaction after 24 each therapy session, two members of a dyad reported on the same relationship. By using a more sophisticated analysis, such as dyadic growth modeling, the interdependence nested within a dyad is accounted for (Kenny et al., 2006). Researchers have proposed two ways to analyze dyadic data: structural equation modeling and multilevel modeling (Kashy & Donnellan, 2008; Ledermann & Kenny, 2017; Planalp et al., 2017). Multilevel modeling (MLM; Raudenbush & Bryk, 2002) was used to examine the dyadic growth models in this study. There are several strengths of using MLM in the current study. First, MLM deals with varying time points and time intervals across and within dyads without increasing the number of missing values (Atkins, 2005; Ledermann & Kenny, 2017). In our study, couples participated in a range of 1 to 17 sessions of EFT. In addition, couples attended therapy with varying time intervals; some couples had sessions weekly and some couples had varying session intervals. This variability of time points and intervals are common challenges in real-life effectiveness studies compared to efficacy studies that require a specific number and frequency of sessions. Second, MLM handles small sample sizes well by using restricted maximum likelihood (REML) (Ledermann & Kenny, 2017). Third, MLM uses maximum likelihood estimation with the expectation-maximization algorithm to deal with missing data, which is a common issue in clinical data (Bartle-Haring et al., 2020). In a clinical trial, clients may not complete the assessment for one session, but they might complete one at the next session, or clients may drop out of therapy before they receive the number of sessions anticipated to create an effect. MLM uses all the data from each participant and can provide unbiased estimates, as long as the data are missing at random (Atkins, 2005; Ledermann & Kenny, 2017; Schafer & Graham, 2002). 25 Plan of analysis The first set of analyses examined research question one: Does EFT improve clinical outcomes (i.e. relationship satisfaction, depression) among Taiwanese couples in Taiwan? Two separate multilevel models were used to test the two outcomes (i.e., relationship satisfaction and depression). The second set of analyses answered research question two: Do attachment style, traditional/modern values, femininity/masculinity ideology, and emotional expressivity at baseline predict an increase in relationship satisfaction and a decrease in depressive symptoms among Taiwanese couples in Taiwan who receive EFT? This question was answered by adding each predictor variable to the two separate multilevel models, yielding a total of 14 models. Each predictor was added to the model to test the relationship between the predictor and the initial level and rate of change for each outcome (i.e., relationship satisfaction and depressive symptoms). 26 CHAPTER IV: STUDY ONE Abstract This is the first study to begin to explore the effectiveness of emotionally focused therapy for relationship satisfaction and depressive symptoms among couples in Taiwan. This one-arm pragmatic trial assessed the clinical outcomes of 17 couples using paired-samples t-tests and multilevel modeling. Paired-sample t-tests revealed no statistical differences in relationship satisfaction and depressive symptoms before and after EFT. Additionally, multilevel modeling indicated no change in relationship satisfaction over time. However, there was a quadratic change in depressive symptoms over the course of EFT. The study results were unexpected, though it is important to note that most couples did not receive the recommended treatment “dose” in this pragmatic trial. In addition, the small sample size limited statistical power. Future research is needed to continue to examine the effectiveness of EFT for couples in Taiwan. 27 Introduction Depression is a prevalent and burdensome mental disorder. Globally, an estimated 5% of all adults suffer from depression (World Health Organization, 2021). In Taiwan, an estimated 8.9% of individuals aged 15 and older had subsyndromal depression and 5.2% had major depressive disorder according to data from 2004 (Ministry of Health and Welfare, 2020). More recent data indicate that about 1.19 million people in Taiwan sought treatment from psychiatrists for depression from 2016–2018 (Taiwan Association Against Depression, 2020), and about 5.9% were prescribed antidepressants in 2019 (Ministry of Health and Welfare, 2020). Depression is the leading global cause of medical disability (WHO, 2021). It affects physical health, daily function, and cognition, and it is a strong predictor of chronic disease and suicidal thoughts and behaviors (National Institute of Mental Health, 2019; Simon et al., 2016). Depression is associated with an increased risk of poor physical and mental health, and it has also been found to have a reciprocal association with relationship satisfaction (Atkins et al., 2009; Beach et al., 2003; Foren et al., 2015; Whisman et al., 2004). Further, the burden of relationship problems, including low relationship satisfaction, high conflict, separation, or divorce, is detrimental to physical and mental health (Slatcher & Selcuk, 2017; Doss et al., 2021). Greater relationship quality is related to better health and lower cardiovascular reactivity during relationship conflict (Robles, et al., 2014). Further, relationship distress has been found to predict an increase in depressive symptoms over time (Beach & O’Leary, 1993). Couple therapy is an effective approach to treat depression and improve relationship satisfaction (Barbato et al., 2018). Emotionally Focused Therapy (EFT), an evidence-based couple therapy, is an efficacious treatment for improving depression (Denton et al., 2012; Dessaulles et al., 2003; Wittenborn et al., 2019), and relationship satisfaction (Denton, 2000; 28 Dalgleish et al., 2015; Johnson & Talitman, 1997). However, most EFT research has been conducted in North America. Few researchers have examined the effectiveness of EFT to treat depression and improve relationship satisfaction in Asia (Ahmadi et al., 2014; Soleimani et al., 2015; Soltani et al., 2013; Soltani et al., 2014). To date, there have been no known studies of EFT among couples in Taiwan. The purpose of this study is to empirically evaluate the effectiveness of EFT in Taiwan. Specifically, the study hypothesis is that EFT will improve clinical outcomes (i.e., increased relationship satisfaction and decreased depression) among Taiwanese couples in Taiwan, based on the assumption in EFT that attachment is universal and thus the approach is applicable to couples across cultures (Johnson, 2019). An Evidence-Based Couple Intervention—Emotionally Focused Therapy This study examines EFT, a widely researched, implemented, and empirically supported couple therapy that is efficacious at decreasing depressive symptoms and relationship distress (e.g., Denton et al., 2012; Dessaulles et al., 2003; Wittenborn et al., 2019). The most recent meta- analysis included 20 studies totaling 332 couples and found medium to large effects for EFT pre- to post-test, pre-test to follow-up, and in comparison to other couple interventions (Spengler et al., in press). Another recent meta-analysis found that EFT had moderate effects on relationship satisfaction changes from the first to last couple therapy session (Rathgeber et al., 2019). In addition, couples who received EFT reported significant growth in relationship satisfaction across a 2-year follow-up (Wiebe et al., 2017). EFT consists of three stages: cycle de-escalation, restructuring of interactional positions, and consolidation and integration of therapeutic gains (Johnson, 2019). In the first stage, therapists track and identify negative interactional cycles and help couples access unrecognized emotions underlying their negative interactions. In the second stage, therapists guide couples to 29 respond to and support each other in an emotionally attuned way, which leads to changes in interactional patterns and new emotional engagement. In the third stage, therapists help couples explore and identify new and more secure solutions to their problems. Change occurs through EFT as a result of new corrective emotional experiences in therapy. When therapists can address and access each partner’s emotions and identify the attachment needs underlying their negative interaction cycles, couples can start to reconnect and create positive cycles of bonding and connection (Johnson, 2019). These shifts, in turn, lead to increased relationship satisfaction. The Application of EFT in Taiwan There is a lack of empirical research regarding EFT among couples outside of North America (Wiebe & Johnson, 2016). A small body of research has tested EFT outcomes among Iranian couples in Iran (Ahmadi et al., 2014; Soltani et al., 2013; Soltani et al., 2014; Soleimani et al., 2015). However, the existing EFT research in Taiwan has been limited to a case study (Liu & Hung, 2019), a qualitative study of couples’ experiences of EFT (Sun, 2019), a clinical study of EFT among couples who had affairs (Wang & Wang, 2015), and a study of the effectiveness of EFT in a “Hold Me Tight” psychoeducational group for unmarried couples (Sun & Liu, 2015). EFT was developed by Sue Johnson in a North American context grounded in Euro- American cultural values and beliefs. It would be problematic to claim that this intervention is effective in all populations, including those outside of the Euro-American cultural context without careful clinical research of the intervention in these cultures. Researchers have argued that EFT is applicable in other cultural contexts because the two core concepts of EFT— attachment and emotions—are universally experienced (Ekman, 2007; Johnson, 2019). Based on this argument, we assume that EFT can be applied to Taiwanese couples and lead to desirable outcomes. 30 The Purpose of the Study Most EFT studies have been conducted in a laboratory or other controlled setting (e.g., Johnson & Greenberg, 1985; Dandeneau & Johnson, 1994; Johnson & Talitman, 1997; Denton et al., 2000; Dessaulles et al., 2003; Denton et al., 2012). Few research studies have tested the effectiveness of EFT in real-life settings (e.g., Ahmadi et al., 2014; Wittenborn et al., 2019), and no studies have examined clinical outcomes among Taiwanese couples in Taiwan. There is a critical need to test the effectiveness of EFT among Taiwanese couples to determine its effectiveness in a different cultural context. In the absence of such information, therapists and couples cannot feel confident that EFT can be successfully used in Taiwan. The purpose of this study was to begin to explore whether EFT improves relationship problems and depression among Taiwanese couples in Taiwan. We hypothesized that couples who receive EFT will demonstrate significant increases in relationship satisfaction and significant decreases in depression over the course of EFT. Method This pragmatic clinical trial (Thorpe et al., 2009) examined the effectiveness of EFT in real-life settings in Taiwan. Rather than conducting this clinical research in a controlled setting, we collaborated with therapists trained in EFT to implement the intervention with couples in their practices in Taiwan. A pragmatic clinical trial enables therapists to apply an intervention flexibly as they would in typical clinical practice. This design imposes minimum constraints, includes few selection criteria, and allows therapists to determine the length of treatment for couples (Sox & Lewis, 2016). 31 Procedures Therapists who were certified in EFT were identified through the Asia EFT website (www.asia-eft.com) in collaboration with the coordinator and director of the Asia EFT Center. Therapists were contacted to collaborate on this study if they (1) had completed the EFT therapist certification requirements established by the International Centre for Excellence in Emotionally Focused Therapy (ICCEFT), (2) used EFT in their practice in Taiwan, (3) received supervision from EFT-certified supervisors biweekly, and (4) worked with Taiwanese couples. Each EFT couple therapist was contacted by email, and the study information was shared. Therapists who expressed interest in collaborating on the study were invited to a Skype meeting to explain the study information in detail and describe their role in the study. Therapists invited their couple clients who met the research inclusion criteria to participate in the study. Couple participants were included in this study if they (1) agreed to receive couple therapy from a participating EFT-certified therapist in Taiwan, (2) were aged 18 or older, and (3) agreed to participate in the study. Couples were excluded from participating in this study if they (1) reported ongoing or past intimate partner violence, (2) reported suicidal thoughts or ideation in the past 3 months, or (3) reported ongoing infidelity. These exclusion criteria were set based on the rationale that clients needing treatment for intimate partner violence, suicidal ideation, and infidelity are not often best served by couple therapy alone (Johnson, 2019). Once couples had reviewed the study information and signed the consent forms, they were officially enrolled in the study, and emails were sent to couples to confirm their enrollment. All participants received EFT in their therapist’s practice setting. Participants completed surveys at intake and after each therapy session. They were encouraged to complete surveys through their 32 smartphones after each therapy session; if smartphones were not accessible, therapists provided paper-and-pencil surveys, which were then sealed in envelopes to protect the participants’ confidentiality. Participants each received an NT$100 gift voucher (equal to approximately USD$3) after each three sessions for which they completed surveys until the termination of EFT. Therapists received an NT$700 gift voucher (equal to approximately USD$23) to compensate them for their time. All gift vouchers were sent to couples and therapists through emails or text messages. Participants The average age of participants in the sample was 33 years for male partners (SD = 7.90) and 32 years for female partners (SD = 7.47). Participant couples’ relationship duration ranged from less than one year to 21 years, with an average of 7 years (SD = 6.70). Among the couples, 65% did not have children, 24% had one child, and 12% had two children. Regarding educational attainment, 18% of male participants and 18% of female participants had completed high school, 65% of males and 59% of females had bachelor’s degrees, and 18% of males and 24% of females had graduate degrees. The most commonly reported income level among participants was above NT75,000 (equal to approximately USD$2,500) monthly. EFT Intervention In all, 17 couples were seen by 10 therapists, including 1 male therapist and 9 female therapists. All therapists had master’s degrees and held professional mental health licenses in Taiwan. Five had counseling psychologist licenses, four had clinical psychologist licenses, and one was a licensed social worker. Therapists had practiced for an average of 9 years (SD = 5.34; range: 4–22 years). Therapists provided EFT in their practices and were instructed to determine 33 the number of sessions needed to achieve positive outcomes based on the needs of each couple. All therapists were instructed to provide EFT according to the EFT manual (Johnson, 2019). Couple participants completed 1–17 sessions (M = 7), and only 41% of couples (N = 7) completed more than 8 sessions; the recommended treatment length for EFT is between 8–20 sessions (Alan & Zeidlhack, 2021). In all, 29% of couples (N = 5) dropped out of the study. Three couples left before the third session because they stopped attending therapy without explanation, one couple left before the second session because they ended their relationship, and one couple left before their seventh session because the therapist had observed an increase in violent behaviors and depressive symptoms and asked the partners to be seen individually. Instruments Demographic Information Demographic information collected from participants included age, gender, education level, combined family income, marital status, years in the current relationship, number of children, and whether they had received or were currently receiving any mental health treatment (e.g., medication, individual treatment, group therapy). Dyadic Adjustment Scale The Dyadic Adjustment Scale (DAS; Spanier, 1976) is a measure of relationship satisfaction that includes four subscales: consensus (13 items), satisfaction (10 items), cohesion (5 items), and affectional expression (4 items). The total score (range: 0–151) is obtained by summing all 32 scale items, and a higher score indicates higher relationship satisfaction (Spanier, 1976). The DAS has demonstrated high reliability with a Cronbach’s alpha coefficient of .96 (Spanier, 1976). The DAS has also been translated into Chinese and validated in various studies (Shek, 1995; Shek et al., 1993; Shek & Cheung, 2008). The Chinese-language version of the 34 DAS was used for the current study (C-DAS; Shek et al., 1993). The C-DAS has demonstrated high reliability with a Cronbach’s alpha coefficient of .91 (Shek et al., 1993). For this study, the internal reliability of DAS was high across treatment and ranged from .92 to .98 for men and .87 to .98 for women. Patient Health Questionnaire The Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001) is a tool used to assess depressive symptoms on a 4-point Likert-type scale (0 = not at all, 3 = nearly every day). Participants rated the extent to which they had experienced each symptom in the last two weeks. Examples of symptoms include “feeling down, depressed, or hopeless,” and “poor appetite or overeating”. The sum of the PHQ-9 score indicates participants’ depression level such that a higher score indicates a higher level of depressive symptoms. The PHQ-9 has been translated to Chinese and tested among Chinese populations with a high reliability of .96 and a high validity of .92 (Wang et al., 2014). The Chinese-language version of the PHQ-9 was used for this study. The internal reliability of PHQ-9 was high across treatment and ranged from .88 to .98 for men. For women, the reliability was poor to acceptable, ranging from .56 to .96 for women (see Table 1). Material Translation This study was conducted in Taiwan and all participants spoke Mandarin. Therefore, all study materials, including the research information flyer and informed consent materials, were written in Mandarin by the first author, who is fluent in both Mandarin and English. All therapy sessions were conducted in Mandarin. 35 Data Analysis In this study, we aimed to explore the effectiveness of EFT in Taiwan through a one-arm pragmatic trial. We used two types of analyses to begin to explore the research question, including paired-samples t-tests and multilevel modeling. Given our pragmatic approach and the exploratory nature of this study, we also conducted a paired-samples t-test in order to assess change between participants’ own intake and termination scores. This analysis was conducted because the number of sessions couples completed varied. In addition, we used multilevel modeling (MLM) to examine change over time using participants weekly scores throughout treatment. In MLM, we planned to use intent-to-treat (ITT) analyses (Gupta, 2011; Kruse et al., 2002). ITT is generally recommended for clinical trials research because it ensures that all participants are included in the analysis regardless of whether or not they completed treatment, thus reducing bias in the estimation of the treatment effects (Gupta, 2011; Kruse et al., 2002). However, we did not use ITT analyses for paired-samples t-tests because in order for pair- samples t-tests to accurately capture the differences between intake and termination scores, we could only include participants who had completed both intake and termination assessments. To ensure this approach was appropriate for our data, we also empirically tested whether missing data effected our model before proceeding. Paired-Samples T-Tests We calculated paired-samples t-tests to compare participants’ relationship satisfaction and depressive symptoms scores from intake to termination (Tabachnick & Fidell, 2013). Only couples who completed EFT (N = 9) were included in the t-tests to ensure that the results accurately reflected the differences in scores from intake to termination. Given that participants 36 in our study received a wide range of sessions, each participant’s score at their termination session was used. Multilevel Modeling This study includes data from both partners and aims to describe change over time in one treatment condition; thus, dyadic growth curve modeling was used to answer the research question. A dyadic growth curve model can examine “how change over time is coordinated across two partners” (Kashy et al., 2008, p. 317), and this approach also accounts for interdependence nested within a dyad (Kenny et al., 2006). Multilevel modeling (MLM) and structural equation modeling (SEM) are both appropriate methods for the analysis of dyadic data (Kashy & Donnellan, 2008; Ledermann & Kenny, 2017; Planalp et al., 2017). This study employs MLM (Raudenbush & Bryk, 2002) for data analysis for several reasons. First, couples in the study completed 1-17 sessions (varying time points) and attended therapy with varying frequency (varying time intervals). An MLM approach can account for varying time points and intervals across and within dyads without increasing the number of missing values (Atkins et al., 2005; Ledermann & Kenny, 2017). Second, the study recruited a small sample, which is common in clinical research (McNeish & Stapleton, 2016). The MLM approach accounts for a small sample size by using restricted maximum likelihood (REML; Ledermann & Kenny, 2017). In MLM, partners’ scores across the weeks of treatment (level 1) were nested within couples (level 2). This type of growth model is created by an intercept (initial score) and slope (rate of change), where the slope is grand mean centered. To develop the MLM growth model for relationship satisfaction, we first determined the number of random effects (i.e., intercept and/or random slope) in our model. Next, we evaluated if nonlinear or quadratic slopes better explained 37 the change across treatment. Then we used omnibus tests of distinguishability to determine if the model was distinguishable by gender. Chi-square difference (or deviance) tests guided our analysis in each step to determine better-fitting models where lower loglikelihood values indicated a more parsimonious model. After completing each of these steps for the DAS, the same analytic steps were completed to develop a model for the PHQ-9. Missing Data Since our study was a pragmatic trial in a real-world setting, it was expected that couples might not attend a session each week or couples might drop out of therapy before we would expect them to have improved. In this case, MLM can use maximum likelihood estimation to handle missing data, which is a common issue in clinical data (Bartle-Haring et al., 2020). MLM also uses all the data from each partner and provides unbiased estimates, as long as the data are missing at random (Atkins et al., 2005; Ledermann & Kenny, 2017; Schafer & Graham, 2002). Restricted maximum-likelihood methods (REML) and Kenward-Roger adjustment (Kenward & Roger, 1997) are the recommended methods for analyses with small sample sizes (McNeish & Stapleton, 2016). REML estimates unbiased variance components well, especially for continuous outcome variables (McNeish & Stapleton, 2016). Kenward-Roger adjustment avoids an inflated Type I error rate resulting from underestimated fixed-effect standard errors (Kenward & Roger, 1997). Thus, researchers recommend using the Kenward-Roger adjustment to guard against biased estimates (Bell et al., 2013). Since this is a study of naturally occurring therapy, the amount of sessions in which couples completed therapy varied. In our study, one couple continue treatment until week 17, while almost half of the couples (47%) completed treatment before the 10th week. However, our MLM model showed that missing at week 17 did not affect our model; one possible explanation 38 for that was that so many couples actually completed treatment before week 17 instead of dropping out from treatment. Because missing data in a study of this design is inevitable, we employed pattern-mixture models separately for the DAS and PHQ-9 to examine if missing data at Week 17 (observed = -1, missing =1) affected our model intercepts and slopes (Son et al., 2012). For the DAS model, there was no difference in model fit when missing at week 17 was added to the model, (χ2(2) = -2.61, p = .27). Specifically, missing at week 17 was not associated with the initial (p = .45) and rates of change in relationship satisfaction across treatment (p = .17). Similarly, for the PHQ-9 model, there was no difference in model fit when missing at week 17 was added to the model, ((χ2(2) = 2, p = .38). Particularly, missing at week 17 was not associated with the initial (p = .40), linear rates of change (p = .10), and the quadratic rates of change in depressive symptoms across treatment (p = .05). Together, the lack of improved model fit when missing was included in the model as well as the lack of associations of missing with the initial and rates of change provide evidence for missing at random in both the PHQ-9 and DAS models. Therefore, we used REML and Kenward Roger standard error corrections in our models. Results Before completing the paired-samples t-tests and MLM, we examined descriptive scores of the outcome variables. The majority of participants in the study (65%) were moderately distressed at intake, with an average DAS score of 85.13 (range: 47–140, SD = 21.54). In addition, the majority of participants scored at or above moderate to severely depressed, with a mean PHQ-9 score of 10.55 (range: 0–24, SD = 6.7). Specifically, 12% of male participants and 12% of female participants in the current study had moderate depression (PHQ-9 scores: 10–14), 24% of males and 24% of females had moderately-severe depression (PHQ-9 scores: 15–19), 39 and 6% of males and 12% of females had severe depression (PHQ-9 scores: 20–27). The mean, standard deviation, and internal reliability of relationship satisfaction and depressive symptoms are presented in Table 1. Paired-Samples T-Tests For couples (N = 9) who completed treatment and completed intake and termination assessments, we also ran paired-samples t-tests to evaluate the impact of EFT on couples’ scores on the DAS and PHQ-9. We compared DAS and PHQ-9 scores from intake to termination. For the DAS, there was no statistical difference in the relationship satisfaction scores before and after EFT for men (p = .49) and women (p = .33). For the PHQ-9, there was no statistical difference in depressive symptoms before and after EFT for men (p = .60) and women (p = .20). Multilevel Modeling Model Building-DAS Unconditional Model In MLM models, we included all couples (N = 17) in the analysis. For the DAS model, we began with an undistinguishable dyad model that modeled an intercept and slope as fixed effects. We then evaluated the number of random effects. We first included random intercepts into the model, both variance (p = .01) and covariance (p =.04) of intercept were significant. However, when we added random slopes into the model, model fit improved (χ (2) = -19.02, p < 01), but the random effect was not significant (p =.87). For parsimony, we only accounted for intercept random effect in the DAS model. Next, we evaluated if nonlinear or quadratic slopes better explained the growth of the DAS model. There was no difference in the model fit when a quadratic slope was included in the model (χ2 (1) = .32, p = .57); therefore, we included a linear slope in the model. Next, we evaluated if the DAS model could be distinguished by gender (female = 1, male = -1). The 40 omnibus test of distinguishability revealed that the model was not distinguishable by gender (χ2 (4) = -3.53, p = .47). Specifically, the intercept (p = .65) and slope (p = .77) did not differ by gender. After completing the above steps, an indistinguishable dyadic multilevel model for the DAS model, including a relationship satisfaction intercept, linear slope, and treatment weeks as fixed effects was the best fitting model (Kenny et al., 2006). Model Building-PHQ-9 Unconditional Model For the PHQ-9 model, we also used an undistinguishable dyad model that modeled an intercept and slope as fixed effects. We also evaluated the number of random effects. When we tested random intercepts, both variance (p = .03) and covariance (p =.04) of intercept were significant. However, when we accounted for slope random effect, model fit improved (χ (2) =12.90, p < .05), but the random effect was not significant (p =.21). For parsimony, we only accounted for the intercept random effect in the PHQ-9 model. Next, we determined if nonlinear or quadratic slopes explained the rate of change of the PHQ-9 model better. The model fit improved when adding the quadratic slope (χ2(1) = 4.78, p <.05). Specifically, there was small, but notable quadratic slope (b = .03, p <.05), while the linear slope was not significant (b = -.14, p =.08), for depressive symptoms across treatment. Hence, we included the quadratic slope in the PHQ-9 model. Last, we tested if the PHQ-9 model could be distinguished by gender (female = 1, male = -1). Similar to the DAS model, for the PHQ-9 model, the omnibus test of distinguishability revealed that the model was not distinguishable by gender (χ2 (5) = 8.64, p = .12). Particularly, the intercept (p = .85), linear slope (p = .05), and quadratic slope (p =.66) did not differ by gender. After completing these steps, an indistinguishable dyadic multilevel model for depressive symptoms that included an intercept, 41 quadratic slope, and treatment weeks as fixed effects was the best fitting model (Kenny et al., 2006). Indistinguishable Dyadic Multilevel Growth Models The variances, covariances, and residuals of the growth models were highly correlated with each other over time. For example, relationship satisfaction for each week was highly correlated (ρ = .75, p <.01) while depressive symptoms for each week was also highly correlated (ρ = .75, p <.01). Given these high correlations over time, we used an autoregressive (1) matrix structure for the variance, covariance, and residuals because it assumes covariances are correlated, and it factors in declining correlations over time (Leyland, 2004). For the DAS model, the average relationship satisfaction intercept was significant (b = 88.26, p <.01) and varied across couples (σ2 = 430.15, SE = 138.61, p < .05) (See Figure 1). However, there was no significant linear change in relationship satisfaction over time (b = -.09, p = .76). For the PHQ-9 model, the average depressive symptoms intercept was significant (b = 8.57, p < .01). The model also revealed that the average depressive symptoms intercept varied across couples (σ2 = 31.21, SE = 10.14, p < .05) (See Figure 2). There was no significant linear change in depressive symptoms (b = -.15, p =.09), but there was a significant quadratic change in depressive symptoms over time (b = .03, p < .05), indicating curvilinear growth. Discussion This is the first study to examine whether EFT was effective for couples in Taiwan. The primary purpose of this study was to examine the effectiveness of EFT for depression and relationship satisfaction among Taiwanese couples in Taiwan. The multilevel growth model findings indicate that couples receiving EFT did not report change in relationship satisfaction 42 over time. The paired samples t-tests (i.e., pre- and post-tests) also did not reveal a significant increase in relationship satisfaction from intake to study termination. This finding is unexpected because prior studies have found that EFT was more effective than a range of control groups (e.g., behavioral approaches, wait-list, other therapy approaches) at increasing relationship satisfaction (e.g., Johnson & Greenberg, 1985; Denton et al., 2000; Wittenborn et al., 2019). One prior study found that couples had an overall 0.39-point increase in relationship satisfaction each week during EFT (Dalgleish et al., 2015). Our study findings also provide preliminary information regarding the rate of change of depressive symptoms over the course of EFT. There was no significant linear rate of change in depressive symptoms, but there was a significant quadratic change (i.e., a curvilinear change in depression) over EFT. There are several possible reasons for the lack of increase in relationship satisfaction over time in this study. First, a highly plausible explanation is that couples did not receive enough sessions for EFT to improve partners’ relationship satisfaction. It is recommended that couples receive 8 to 20 sessions as they move through the three stages of EFT, with more sessions needed for more severe symptoms (Alan & Zeidlhack, 2021; Johnson, 2019). In our study, 17 couples completed a range of 1 to 17 sessions, with only 41% (N = 7) of couples completing more than 8 sessions. Therefore, couples did not likely receive the correct “dose” of treatment, especially given the severity of the symptoms experienced by partners. A second possible explanation for this finding is that the current study included a small sample size, and there was insufficient power to detect the treatment effect (McNeish & Stapleton, 2016). The lack of a robust sample may have contributed to the findings (Evans & Ildstad, 2001). Another possible explanation for the unexpected study findings relates to the public health system in which the study occurred. Taiwan’s national health insurance provides universal 43 coverage; however, counseling and therapy services in private practice are not covered by the national health insurance (Guo et al., 2013; Ministry of Health and Welfare, 2021). In Taiwan, the cost of therapy is high, ranging from $40 to $160 per hour, depending on the geographic location of the private practice and the therapist’s years of experiences (Taiwan Counseling Psychologist Union, 2011). The high cost of couple therapy could have contributed to the low number of completed sessions. It may be the case that couples discontinued therapy when they started to feel better. The first major change process in EFT is de-escalation of the negative cycle, which occurs at the end of the first stage of EFT (Johnson, 2019). At that point in therapy, couples often experience new ways of interacting and begin to feel newly hopeful about their relationship (Johnson, 2019). It is possible that couples began to feel some relief and, given the high cost, decided to terminate treatment. Doing so would prevent couples from receiving the full “dose” of treatment (e.g., three stages of EFT), which could alter the immediate and sustained impact of EFT (Burgess Moser et al., 2016). Regarding depression, the majority of participants scored above the cutoff point for moderate depression at the initial assessment. In Taiwan, only 20 percent of people with depression seek professional help (Taiwan Association Against Depression, 2020). For those who seek professional help, they often receive treatment from psychiatrists rather than counselors or therapists; Between 2016 to 2018, around 1.19 million people in Taiwan sought treatment from psychiatrists for depression (Taiwan Association Against Depression, 2020). In Asia, including Taiwan, mental health problems are seldom diagnosed due to the lack of accessibility and availability of services and stigma associated with seeking services (World Health Organization, 2017). For individuals who have intentions to seek help, most would prefer to use sources and support from their social networks, such as family and friends, rather than 44 from healthcare providers (Oliver et al., 2005; Arnault, 2009). This phenomenon might explain the severity of depression in our study sample since partners might suffer from depression, but delay seeking treatment until it becomes unbearable. For people experiencing severe depression, the most effective treatment is a combination of antidepressant medication and therapy (National Institute for Health and Care Excellence, 2009). It is possible that participants with severe depression may have experienced more improvement with a combination of EFT and antidepressant medication, though future research is needed to test this. It is also possible that EFT may need to be culturally adapted to fit the needs of couples in Taiwan in order to facilitate change. Bernal and colleagues (2009) have recommended the consideration of language, cultural patterns, meaning, and values when systematically modifying an evidence-based treatment for cultural adaptation. Some specific cultural characteristics in Chinese culture may affect how couples interact with one another and how EFT can be used to facilitate change. For example, tolerance in relationships is encouraged in the Chinese cultural context (Chang et al., 2020). Chang and colleagues (2020) found that tolerance is often used to repress negative emotions to maintain harmony among Taiwanese couples. Chen and Li (2007) also found that tolerance can predict an individual’s marital satisfaction and their partner’s supportive behaviors. The main mechanism of change in EFT is the expression of emotions to create new connections and experiences. This mechanism may conflict with the cultural concept of tolerance in that couples in Taiwan and other Chinese cultural contexts may use tolerance as a coping strategy. Limitations This study had several limitations. First, only 41% of couples completed more than 8 EFT sessions; the remaining couples completed 8 or fewer sessions (29%) or discontinued 45 treatment (29%). Couples are recommended to receive at 8 to 20 or more sessions, based on the needs of the couple, to improve (Alan & Zeidlhack, 2021). Partners who report more severe symptoms often need more sessions or even other treatment plans (e.g., adjunctive antidepressant medication). The pattern mixture model indicated that missing data at week 17 did not affect the model; however, the fact that the majority of couples in the study did not have the recommended treatment “dose” could explain the lack of clinical improvement. Additionally, another limitation of the study is the small sample size. A larger sample size might give the model more power to detect a treatment effect. Finally, while most couples completed their measures immediately after their session, others were delayed for various reasons and didn’t complete them until the following day. The validity of the assessment might be affected if all couples did not complete the assessment immediately after each treatment session. Conclusions This is the first study to begin to explore the effectiveness of EFT among couples in Taiwan. The results of the one-arm pragmatic trial did not indicate significant change in clinical outcomes; however, several limitations, including the small sample size and small treatment “dose” may have accounted for the limited change observed. Future research that studies a longer course of treatment with a larger sample of couples is needed to continue to explore whether EFT is effective for couples in Taiwan. 46 CHAPTER V: STUDY TWO Abstract This study examined predictors of change in relationship satisfaction and depression over the course of EFT in a pragmatic clinical trial of 17 couples in Taiwan. Results of multilevel modeling revealed that high traditionalism at intake predicted an increase in relationship satisfaction for women. Attachment was also a significant predictor of change; men with high attachment avoidance at intake demonstrated a significant decrease in depression, and men and women with high attachment anxiety at intake also experienced a significant decrease in depression. Lastly, emotional expressivity at intake was associated with an increase in relationship satisfaction and a decrease in depression for both men and women. Our study suggested that traditionalism, attachment, and emotional expressivity are important predictors of change among couples in Taiwan who receive EFT. While more research is needed, these findings offer preliminary support for the types of partners who may be more likely to experience change after receiving EFT. 47 Introduction Attachment theory, the theoretical foundation of Emotionally Focused Therapy (EFT), considers attachment a universal experience across cultures (Bowlby, 1988; Johnson, 2019). Based on this assumption, EFT is assumed to be applicable across a variety of populations. Emotion, another important change mechanism in EFT, regulates and organizes attachment responses (Johnson, 2019). Emotions occur automatically, and seven universal emotions have been identified that cross language, cultural, and racial barriers: anger, contempt, disgust, enjoyment, fear, sadness, and surprise (Ekman, 2007, 2018). EFT was developed in a North American context and has demonstrated efficacy in prior studies (e.g., Dalgleish et al., 2015; Wittenborn et al., 2019), as well as studies that have examined baseline client characteristics associated with change (Johnson & Talitman, 1997). However, it is important to examine predictors of change among couples receiving EFT in non-Western cultural contexts. Understanding the baseline characteristics of partners that are associated with improvement after receiving EFT is an important step in determining which couples might benefit from such treatment. Not all couples benefit from couple therapy; some couples fail to show improvement in relationship satisfaction over the course of treatment. Understanding predictors of change enables clinicians to ensure that their clients are receiving the treatment best suited to their needs (Atkins et al., 2005; Dalgleish et al., 2015). Few studies have examined predictors of success in EFT (e.g., Johnson & Talitman, 1997; Dalgleish et al., 2015), and none have focused on Taiwanese couples in Taiwan. The aim of this paper is to identify which baseline characteristics predict improvement among couples receiving EFT in Taiwan. Specifically, the paper will explore associations between partners’ attachment, emotional expressivity, and specific Taiwanese cultural characteristics (e.g., traditional values and 48 masculinity ideology) at baseline with changes in relationship satisfaction and depressive symptoms over the course of EFT. Attachment Attachment between infants and mothers exist in every culture (van Ijzendoorn & Sagi- Schwartz, 2008), and adult attachment styles have also been identified across cultures (van Ijzendoorn & Bakermans-Kranenburg, 1996). Regarding patterns of attachment, a securely attached adult is characterized by low attachment avoidance and low attachment anxiety, whereas an insecurely attached adult is characterized by high levels of attachment avoidance and/or anxiety (Bartholomew & Horowitz, 1991). Individuals with high attachment avoidance develop deactivating strategies (e.g., withdrawal, focusing on other tasks instead of relationship problems) to self-soothe whereas individuals with high attachment anxiety tend to develop hyperactivating strategies (e.g., consistent attempts to reach proximity) to ensure the support and love they are seeking is still attained (Johnson, 2019). While studies have examined attachment across cultures, some scholars argue that the assumptions of attachment constructs are rooted in Western culture (Rothbaum et al., 2000; Wang & Mallinckrodt, 2006). As a result, some researchers have argued against the universality of these constructs because cultural values and beliefs may affect how these attachment styles are perceived in different cultural settings. Schmitt and colleagues (2004) examined differences in adult attachment across 62 cultures and found that preoccupied attachment was more prevalent in Eastern cultures (e.g., Taiwan, Hong Kong, and China) than in other cultures. Ditommaso and colleagues (2005) also found that Chinese individuals had significantly lower attachment security than Canadian individuals. In addition, higher attachment avoidance has been reported among Asian Americans than among Caucasian and Hispanic Americans (Wei et al., 2004). 49 Wang and Mallinckrodt (2006) found that Taiwanese individuals idealized avoidance behaviors more than Americans and that Taiwanese men idealized anxiety-attachment behaviors more than American men. Baseline attachment styles could be an important predictor of success in EFT given that EFT emphasizes changes in attachment-related bonds, thoughts, and behaviors (Johnson, 2019; Dalgleish et al., 2015). Some studies have examined the influence of attachment style on relationship satisfaction during EFT treatment among couples. Couples who demonstrated insecure attachment prior to EFT treatment have been found to have a significant decrease in relationship satisfaction over the course of EFT treatment (Dalgleish et al., 2015; Burgess Moser et al., 2016); Specifically, compared to couples with low attachment anxiety, those with high attachment anxiety at intake experienced a decrease in relationship satisfaction across EFT sessions (Dalgleish et al., 2015). Similarly, Johnson and Talitman (1997) found that male partners who showed reluctance to turn to their partner for support at intake (e.g., a characteristic of attachment avoidance) had a higher level of relationship satisfaction at termination of EFT. These findings suggest that EFT could be beneficial for couples with insecure attachment, as the main goal of EFT is to help couples identify and express unmet attachment needs, as well as reconsider their partner’s criticism, anger, and frustration as attachment hurts and fear (Johnson, 2019). Though attachment behaviors are universal across cultures, some cultures may be more accepting of insecure attachment than others. Specifically, couples in Taiwan may display insecure attachment more often than couples in the U.S. because these attachment styles are more culturally acceptable (Wang & Mallinckrodt, 2006). EFT is assumed to be an effective intervention to increase relationship satisfaction and reduce depression for couples in Taiwan, 50 but it is important to examine how the attachment styles of partners receiving EFT affect the trajectory of change in relationship satisfaction and depression. Emotional Expressivity Emotional expression, which is thought to be one of the most important mechanism of change in EFT, enables couples to organize the ways they respond to each other in a way that helps their partner understand their needs (Johnson, 2019). Johnson and Talitman (1997) found that male partners who were described as emotionally inexpressive by their partners at intake had the greatest gains in relationship satisfaction at EFT termination. This finding highlights the importance of emotional expressivity in EFT, which could help male partners access and share their vulnerable emotions with their partners. Another study found that Iranian women demonstrated increased emotional expressivity after EFT (Motaharinasab et al., 2016). When examining the role of emotional expressivity in EFT among couples in Taiwan, it is imperative to consider how Chinese culture shapes and informs the expression of emotion. In Chinese culture, consideration of the reactions of others and emphasis on the maintenance of harmonious relationships affects individual’s comfort levels and willingness to express emotions (Hwang & Han, 2010). Tolerance and suppression of one’s emotions are qualities that are valued in Chinese culture (Chang et al., 2020). Tolerance facilitates harmony among Taiwanese couples, which can entail suppressing one’s emotions and sacrificing one’s needs (Chang et al., 2020; Chen & Li, 2007; Huang, 2009). Li (2006, 2012) also found that suppressing emotion is considered an indirect conflict management approach in Western culture, but is considered a helpful coping strategy in Chinese culture. Indeed, suppressing emotions has been associated with increased relationship satisfaction among couples in Taiwan (Li & Hsiao, 2016). 51 Emotional expressivity may look different in Taiwanese couples due to the Chinese cultural influence, and it is important to examine whether individual differences in emotional expressivity at baseline predict change over the course of EFT. This finding has important implications for EFT therapists regarding how emotion in the Taiwanese cultural context may affect clinical outcomes. Traditional/Modern Values and Masculinity/Femininity Ideologies Taiwan has undergone rapid economic and societal changes since 1960 (United Nations, 2014). It has become increasingly common for women in Taiwan to delay marriage and childbearing, and pursuit of higher education and career goals are increasingly seen as priorities for women (Wang & Yang, 2019). Moreover, individuals in Taiwan have increasingly embraced liberal values, such as the belief that one can enjoy a satisfactory life without a relationship or marriage and the adoption of singlehood as a life choice (Cheng & Yang, 2021; Tsai et al., 2021). Despite the influence of modern and progressive values, traditional values persist in Taiwan and continue to shape individual’s perspectives on relationships and family (Li & Hsiao, 2016). The Confucian emphasis on patriarchal and traditional values is still present in Taiwanese culture (Li & Hsiao, 2016; Yu, 2006). Several studies have found that Taiwanese women who adhere to traditional gender roles and values experience higher relationship satisfaction than women who endorse modern values (e.g., Yu, 2006). Taiwanese women who endorse egalitarian values also have lower relationship satisfaction than women in other East Asian countries (e.g., China, Korea; Fan & Qian, 2022). Due to the continued influence of traditional values on Taiwanese society, women with egalitarian values may need to consistently negotiate their values because they conflict with societal expectations (Fan & Qian, 2022; Qian & Sayer, 52 2016). Women in Taiwan may also experience low relationship satisfaction because they are sensitive to and aware of power inequalities, such as the unequal division of housework in relationships (Davis & Greenstein, 2009). Another approach to examining traditionalism and modernism is to look at masculinity and femininity ideologies. Masculinity and femininity ideologies refer to an individual’s internalization of cultural beliefs about male and female gender roles (Levant et al., 2007). For example, Levant and Richmond (2007) found that individuals in China, Russia, Japan, and Pakistan were more likely than Americans to endorse a traditional masculinity ideology. It is important for therapists to assess a couple’s masculinity and femininity ideologies to determine whether the couple’s conflicts may be related to their implicit assumptions regarding men’s and women’s roles in relationships (Levant et al., 2013). To date, only one study has examined the influence of traditionalism on EFT treatment outcomes (Johnson & Talitman, 1997), and it was conducted in North America. The study found that partners’ levels of traditionalism at intake did not predict EFT outcomes. The present study was conducted in Taiwan, where Confucianism and a mix of traditional and modern values shape individuals’ worldviews (Tsuya & Bumpass, 2004); thus, it is particularly important to examine how partners’ traditionalism, modernism, and masculinity and femininity ideologies affect EFT outcomes. Current Study Examining predictors of change over the course of EFT is important in distinguishing between couples who are responsive to EFT and those who are not (Johnson & Talitman, 1997; Roth et al., 2005). Considering cultural values and characteristics and exploring individuals’ and couples’ levels of traditionalism, modernism, masculinity, and femininity may help to explain 53 why some couples in Taiwan experience greater changes in relationship satisfaction and depression during EFT. The aim of this study is to examine how baseline attachment style, traditional and modern values, femininity and masculinity ideologies, and emotional expressivity predict changes in relationship satisfaction and depressive symptoms among Taiwanese couples in Taiwan who received EFT. Responses to couple therapy vary, and it is important to identify certain characteristics of individuals that might affect therapy outcomes. By developing an understanding of how couples in Taiwan responded to EFT based on their individual and couple characteristics, therapists can better predict which couples will be most responsive to EFT. Methods This study is a single-arm pragmatic clinical trial of EFT for couples in Taiwan. Participants received EFT in therapists’ practice settings and were asked to complete an assessment after each therapy session throughout the course of treatment. At the intake session, couples were asked to complete a demographic survey and an assessment of predictors that were hypothesized to be associated with change in EFT for couples in Taiwan. Procedures In this study, we collaborated with the coordinator of the Asia EFT Center to assist us with the recruitment of therapists in Taiwan. The coordinator assisted us in locating Certified EFT Therapists in Taiwan. Researchers then contacted each certified EFT therapist through email to explain the study purpose and compensation and to ask about their interest in participating in our research study. Once therapists expressed interest in the study, researchers then set up individual meetings with them through Skype to review the research study in detail. Therapists then helped researchers to recruit couples who met eligibility criteria at intake in their practices. 54 Therapists collaborated in our study if they: (1) received certification as an EFT therapist by the International Centre for Excellence in Emotionally Focused Therapy (ICCEFT), (2) used EFT in Taiwan to treat couples in Taiwan, and (3) received supervision from EFT certified supervisors biweekly. Therapists reached out to couples they planned to treat in their practice to inquire about their interest in participating in the study. A total of 10 therapists (9 women and 1 man) enrolled in this study. All therapists hold masters’ degrees in a mental health field, and all therapists were licensed mental health professionals in Taiwan (1 social worker, 4 clinical psychologists, and 5 counseling psychologists). Therapists had practiced for an average of 9 years (SD = 5.34; range: 4–22 years). The inclusion criteria for couple participants were: (1) more than 18 years old, (2) received couple therapy from a participating EFT therapist in Taiwan, and (3) both partners agreed to participate. Couples were excluded from the study if they: (1) experienced past or ongoing intimate partner violence, (2) reported infidelity in the past 3 months, (3) and experienced suicidal ideation in the past three months. EFT encourages couples to show vulnerability and express their inner needs and emotions. Expressing vulnerability could be problematic and put partners at risk if there is intimate partner violence or infidelity in the relationship (Johnson, 2019). In this study, participating EFT therapists assisted researchers in determining participants’ eligibility. Participants completed assessments in their therapist’s office independently after each session. Partners completed assessments independently through their smartphones; those unable to use smartphones received paper-pencil surveys from their therapist. At intake, partners completed an assessment of their demographic characteristics, attachment style, femininity and masculinity ideology, traditional and modern values, emotional expressivity, relationship 55 satisfaction, and depressive symptoms. After the second session, partners completed assessments of relationship satisfaction and depressive symptoms after each couple therapy session until treatment was terminated. Therapists received a NT$700 (approximately $23 U.S. dollar) gift voucher for the assistance in the study. Each couple received a NT$100 gift voucher (approximately $3 U.S. dollars) after they completed every three sessions until termination of therapy. Researchers sent gift vouchers to therapists and couples through SMS texts or emails. Participants All couples reported being in a different-sex relationship. Couple participants had been together for an average of 7 years (SD = 6.70), ranging from less than one year to 21 years. The average age of male partners was 33 years (Median = 35; SD = 7.90; range: 19–47 years) and female partners was 32 years (Median =31; SD = 7.47; range: 20–45 years). Regarding children, 65% of couples reported having no children, 24% had one child, and 12% had two children. The majority of participants had high educational attainment, with 18% of males and 24% of females holding graduate degrees and 65% of males and 59% of females holding bachelor’s degrees. The majority of the participants (41%) reported a monthly income level above NT75,000 (equal to approximately USD$2,500). Measures Attachment Avoidance and Anxiety Partners completed the Experiences in Close Relationship-Relationship Structures Questionnaire (ECR-RS; Fraley et al., 2011) at the intake session. ECR-RS assessed attachment styles of four domains: relationships with mother, father, romantic partner, and best friend. In our study, we specifically assessed participants’ attachment with their romantic partner. ECR-RS 56 consisted of 9 items on a 7-point Likert scale (1 = strongly disagree; 7 = strongly agree). Item 1 to 6 assessed how comfortable participants were to open up and depend on their partners (avoidance dimension, i.e., It helps to turn to this person in times of need), and items 7 to 9 assessed the extent to which participants worried their partners may reject them (anxiety dimension, i.e., I often worry that this person doesn’t really care for me) (Fraley et al., 2000). Avoidance scores were obtained by averaging items 1 to 6, with item 1 to 4 reverse coded. Anxiety scores were obtained by averaging items 7 to 9. In previous research, the internal reliability was .87 (avoidance domain) and .91 (anxiety domain) (Fraley et al., 2011). ECR-RS was translated into Chinese and tested with Chinese populations (Zhang et al., 2021). In our study, the internal reliability of attachment avoidance was .90 for men and .45 for women, and attachment anxiety was .70 for men and .91 for women. Femininity Ideology Partners completed the Femininity Ideology Scale–Short Form (FIS-SF; Levant et al., 2017) at the intake session. FIS-SF examined traditional femininity ideology through 12 items with a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree). FIS-SF includes three subscales: dependence/deference, purity, and emotionality and traditional roles. Scores were obtained by summing 12 items, ranging from 12 to 60, with higher scores indicating more endorsement of traditional femininity ideology. In prior research, the internal reliability was .85 (Levant et al., 2017). In our study, we translated this scale into Chinese, and the internal reliability was .84 for men and .77 for women. Masculinity Ideology Partners completed the Male Role Norms Inventory–Short Form (MRNI-SF; Levant et al., 2013) at the intake session. MRNI-SF examined traditional femininity ideology through 21 57 items using 7-point Likert-type scales (1 = strongly disagree, 7 = strongly agree). MRNI-SF includes seven subscales: avoidance of femininity, negativity toward sexual minorities, self- reliance through mechanical skills, toughness, dominance, importance of sex, and restrictive emotionality. The total sum scores ranged from 21 to 147, with higher scores indicating higher endorsement of traditional masculinity ideology. In prior research, internal reliability was .92 for men and .94 for women for the total scale with a North-American sample (Levant et al., 2013). In our study, we translated the MRNI-SF, and the internal reliability was .86 for men and .91 for women. Traditional Values Partners completed the Multidimensional Scale of Chinese Individual Traditionality– Short form (MS-CIT–SF; Kao & Lu, 2006) at the intake session. MS-CIT–SF examined traditional values by using 15 Chinese-language items (6-point Likert scale, 1 = strongly disagree, 6 = strongly agree). The total sum ranges from 15 to 90, with higher scores indicating higher endorsement of traditionality. MS-CIT–SF has five subscales: submission to authority, filial piety and ancestral worship, conservatism and endurance, fatalism and defensiveness, and male dominance. This scale was developed in Taiwan, and the internal reliability with a Taiwanese sample was .81 (Kao & Lu, 2006). In our study, the internal reliability was .89 for men and .94 for women. Modern Values Partners completed the Multidimensional Scale of Chinese Individual Modernity—Short form (MS-CIM–SF; Kao & Lu, 2006) at the intake session. MS-CIM–SF assessed modern values with 15 Chinese-language items on a 6-point Likert-type scale (1 = strongly disagree, 6 = strongly agree). The total sum score has a possible range of 15 to 90, with higher scores 58 indicating higher endorsement of modernity. This measure includes five subscales: Egalitarianism and open-mindedness, independence and self-reliance, optimism and assertiveness, affective hedonism, and sex equality. This scale was developed in Taiwan, and the internal reliability with a Taiwanese sample was .71 (Kao & Lu, 2006). In our study, the internal reliability was .55 for men and .83 for women. Emotional Expressivity Partners completed the Emotional Expressivity Scale (EES; Kring et al., 1994) at the intake session. EES assessed participants’ emotional expressivity by using 17 questions with a 6- point Likert scale (1 = never true, 6 = always true). The total sum could range from 17 to 102, with higher scores indicating higher levels of emotional expressivity. Sample items include, “I think of myself as emotionally expressive” and “I don’t like to let my partner see how I am feeling.” EES was translated in to Chinese and tested with a Chinese population with a high reliability of .82 (Chan et al., 2010). In this study, we used the translated Chinese version of EES, and the internal reliability was .75 for men and .73 for women. Relationship Satisfaction Partners completed the Dyadic Adjustment Scale (DAS; Spanier, 1976) at intake and after each therapy session until the termination of treatment. The DAS has four subscales: consensus, satisfaction, cohesion, and affectional expression. The possible score ranges from 0 to 151, with higher scores indicating higher relationship satisfaction (Spanier, 1976). The DAS demonstrates a high reliability, with internal reliability of .96 (Spanier, 1976). The DAS was translated into Chinese and has been validated in various studies with Chinese populations (Shek, 1995; Shek et al., 1993; Shek & Cheung, 2008), with high reliability of .91 (Shek et al., 1993). 59 We used the Chinese version of DAS (Shek et al., 1993) for this study, and the internal reliability was high across treatment, ranging from .92 to .98 for men and .87 to .98 for women. Depressive Symptoms Partners completed the Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001) at intake and after each therapy session until the termination of treatment. PHQ-9 assessed depressive symptoms through nine questions with a 4-point Likert-type scale (0 = not at all, 3 = nearly every day). Scores were obtained by summing up responses to the 9 questions, with higher scores indicating higher depressive symptoms. The PHQ-9 was translated into Chinese and was tested with Chinese populations with a reliability of .86 and validity of .92 (Wang et al., 2014). We used the Chinese version of the PHQ-9 in this study, and the internal reliability was high across treatment, ranging from .88 to .98 for men. The reliability was acceptable for women, ranging from .56 to .96. Materials Translation This study included Chinese-speaking couples in Taiwan. To ensure cultural sensitivity throughout the research study, all research materials (e.g., flyers, recruitment emails, consent forms, measures) were translated into Chinese. For measures, only FIS-SF, MRNI-SF, and EES were translated into Chinese; ECR-RS, DAS, and PHQ-9 had existing Chinese versions, and MS-CIT–SF, MS-CIM–SF were developed in the Chinese language. Translation of the FIS-SF, MRNI-SF, and EES were completed using the following procedures. Translation from English to Chinese should ensure semantic equivalence (i.e. words and phrases in English and Chinese represent the same meaning), conceptual equivalence (i.e. the commonality of the concept between English and Chinese), and normative equivalence (i.e. the translated measure should address the difference in social norms between two cultures) (Behling 60 & Law, 2000; Pan & Puente, 2005). It is important to include translators who are fluent in both English and Chinese, very familiar with both cultures, and knowledgeable about the measures (Geisinger, 1994). To meet this requirement, the first author completed all English to Chinese translation. The first author was fluent in both English and Chinese and studied couple and family therapy in master’s and doctoral programs in the U.S. She was also very familiar with both cultures. The translation process was completed as follows: (1) the first author translated materials and measures from English to Chinese and documented specific translation challenges encountered and decisions made to handle those challenges, (2) the first author asked two translator reviewers who were also fluent and bilingual in English and Chinese to review the translation. The goal of the review was to ensure the text and semantics were accurate, identify problematic translation, and provide feedback. (3) The first author then revised the translation based on reviewers’ feedback and developed a finalized version of translated materials and measures. It is also important to adapt measures on a question-by-question basis when translating the measure to ensure items are culturally meaningful to the target population after translation (Geisinger, 1994; Tran, 2009). When translating MRNI-SF, we determined there were a few items that needed to be adapted. For example, the question “Men should watch football games instead of soap operas” was adapted and changed to “Men should watch baseball games instead of soap operas”. Football is a popular sport in the U.S. but not popular in Taiwan, and the equivalent popular sport in Taiwan is baseball. Data Analysis We used multilevel modeling (MLM; Raudenbush & Bryk, 2002) to answer our research question: Do attachment style, femininity/masculinity ideology, traditional/modern values, and 61 emotional expressivity at baseline predict change in relationship satisfaction and depression among Taiwanese couples in Taiwan who receive EFT? This research question builds off the work from manuscript 1. Specifically, in manuscript 1, we developed multilevel growth models for relationship satisfaction and depressive symptoms. In developing both models, we determined the number of random effects in our model and whether nonlinear or quadratic slopes could explain the rate of change in relationship satisfaction across 17 weeks of EFT better than linear rates of change. During each step, we used chi-square difference tests (or deviance tests) to determine if model fit improved, where a more parsimonious model had a lower loglikelihood value. The current study extends upon the multilevel growth models of relationship satisfaction and depressive symptoms in manuscript one by adding seven separate predictors (i.e., attachment anxiety, attachment avoidance, femininity ideology, male role norms, traditional values, modern values, and emotional expressivity) to both relationship satisfaction and depressive symptoms growth models respectively, yielding a total of 14 models. Specifically, each predictor was added to the model to test the initial level (intercept) and rate of change (slope) for relationship satisfaction and depressive symptoms. Since theories and past research studies indicated that there were gender differences in the levels of these predictors (e.g., Levant, 2013), we then added gender (female = 1, male = 0) as a predictor to examine gender differences in the association between the predictor and initial levels as well as rates of change. Missing Data It is common for clinical studies to have missing data (Bartle-Haring et al., 2020). To handle missing data, MLM can use maximum likelihood estimation to provide unbiased estimates (Atkins et al., 2005; Ledermann & Kenny, 2017; Schafer & Graham, 2002). It is also 62 common to have small sample sizes in clinical research (Miller, 2020). To handle small sample sizes, it is recommended to use Restricted maximum-likelihood methods (REML) and Kenward- Roger adjustment (Kenward & Roger, 1997). We used two pattern-mixture models separately for relationship satisfaction and depressive symptoms to determine if missing data at Week 17 (observed = -1, missing =1) affected our model (Son et al., 2012). For both models, the model fit did not change when missing was added to the model (relationship satisfaction: χ2(2) = -2.613, p = .27; depressive symptoms: χ2(2) = 2, p = .377). Missing at week 17 was not associated with initial (p = .453) and rate of change (p = .166) in relationship satisfaction or initial (p = .401) and rate of change (p = .10) in depressive symptoms. Together, it provided evidence that missing data in both models were missing at random. We then used REML and Kenward Roger standard errors when running our models. Results Before running MLM, we examined the data to ensure that all the scales were normally distributed. We also examined descriptive statistics. Couples (N = 17) completed a range of 1 to 17 sessions (M = 7). Using the clinical cutoff of 10 points, the majority of the participants started out therapy moderately depressed to severely depressed (44%), as measured by the Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001). Using the cutoff of 97 points, the majority of couples (65%) experienced relationship distress when they started EFT, as measured by the Dyadic Adjustment Scale (DAS; Spanier, 1976). Means and deviations for all predictors are shown in Table 3. 63 Indistinguishable Dyadic Multilevel Growth Models Relationship Satisfaction For the relationship satisfaction model, we only accounted for the random intercept because the random slope was not significant (p =. 87). We also determined that a quadratic slope did not improve model fit, compared to only including the linear slope (χ2 (1) = .32, p = .57). For all models, we used an autoregressive matrix structure to estimate the variances, covariances, and residuals. For more details regarding the procedures and findings, please refer to the results section of manuscript 1. To examine predictors of change in relationship satisfaction, we ran seven separate models to test each predictor and gender differences. Three predictors were not associated with changes in relationship satisfaction. Specifically, attachment avoidance at intake was not associated with the rate of change in relationship satisfaction (b = -.31, p = .29) and this association did not differ between women and men (b = .21, p = .49; see Table 4). Attachment anxiety at intake, also, was not associated with the rate of change in relationship satisfaction (b = .12, p = .44) and this association did not differ between women and men (b = .03, p = .89; see Table 5). Similarly, modern values at intake was not associated with the rate of change in relationship satisfaction (b = .01, p = .81) and this association did not differ between women and men (b = -.05, p = .22; see Table 9). On the other hand, four predictors were associated with changes in relationship satisfaction and some differed by gender. Particularly, femininity ideology at intake was associated with a small increase in the rate of change in relationship satisfaction (b = .08, p = .01) and this association differed between women and men (b = .12, p = .00; see Table 6). We then used a two-intercept approach to evaluate gender differences further by using binary codes 64 for men (men = 1, women = 0) and women (women = 1, men = 0) to test both trajectories simultaneously. This revealed that higher femininity ideology at intake was associated with an increase in rate of change in relationship satisfaction for women (b = .19, p < .01), but not for men (See Figure 3). Similarly, male role norms at intake were associated with a small increase in the rate of change in relationship satisfaction (b = .05, p < .01) and this association did differ between women and men (b = .04, p = .01; see Table 7). A two-intercept model revealed that higher masculinity ideology at intake was associated with a small increase in the rate of change in relationship satisfaction for women (b = .09, p < .01), but not for men (See Figure 4). Traditional values at intake were not associated with the rate of change in relationship satisfaction (b = .12, p = .44), but this association differed between women and men (b = .06, p = .01; see Table 8). A two-intercept model revealed that traditional values at intake were associated with a small increase in the rate of change in relationship satisfaction for women (b = .11, p < .01), but not for men (See Figure 5). Finally, emotional expressivity at intake was associated with a small increase in the rate of change in relationship satisfaction (b = .08, p < .01), however, this association did not differ between women and men (b = -.02, p = .39; see Table 10). Depressive Symptoms For the depressive symptoms model, we only accounted for the random intercept because the random slope was not significant (p =. 21). We also determined that a quadratic slope (b = .03, p <.05) explained the rate of change of the model better than a linear slope (b = -.14, p =.08). Then, we determined that the model could not be distinguished by gender (χ2 (5) = 8.64, p = .12), indicating that depressive symptom trajectories were similar for both partners. For all models, we 65 used an autoregressive matrix structure to estimate the variances, covariances, and residuals. Refer to manuscript 1 for more extensive details regarding the procedures and results. To examine predictors of change in depressive symptoms, we ran seven separate models that tested the predictor and gender differences. Four predictors were not associated with changes in depressive symptoms. First, femininity ideology at intake was not associated with the rate of change in depressive symptoms (b = -.00, p =.78) and the association did not differ between men and women (b = -.02, p =.06; see Table 13). Male role norms at intake was also not associated with the rate of change in depressive symptoms (b = -.00, p =.87) and the association did not differ between men and women (b = -00, p = .88; see Table 14). Similarly, traditional values at intake was not associated with the rate of change in depressive symptoms (b = -.00, p =.55) and this association did not differ between men and women (b = -.01, p =.31; see Table 15). Lastly, modern values at intake was not associated with the rate of change in depressives symptoms (b = -.01, p =.56) and this association did not differ between men and women (b = .01, p =.31; see Table 16). Three predictors were associated with changes in depressive symptoms and some differed by gender. Specifically, attachment avoidance at intake was associated with a small decrease in the rate of change in depression (b = -22, p < .05). There was also a significant difference in the rate of change of depressive symptoms between men and women (b = .21, p < .05). Specifically, the two-intercept model showed that higher attachment avoidance at intake was associated with a large decrease in the rate of change in depression for men (b = -.43, p < .01; see Table 11 and Figure 6). Similarly, attachment anxiety at intake was associated with a decrease in the rate of change in depression symptoms throughout treatment (b = -.21, p < .01). The results also showed that the association between attachment anxiety and rate of change in depressive 66 symptoms was not significantly different for men and women; however, it was likely because attachment anxiety for both partners were decreasing at a similar rate of change. Upon further examination by running a two-intercept model, it showed that higher attachment anxiety at intake was associated with a decrease in the rates of change in depression for both men (b = -.20, p < .05) and women (b = -.22 p < .001; see Table 12 and Figure 7). Lastly, emotional expressivity at intake was associated with a small decrease in the rate of change in depression (b = -.02 p < .05). However, the association between emotional expressivity and the rate of change in depressive symptoms was not significantly different for men and women (b = -.01 p = .17; See Table 17). Discussion The aim of this study was to examine predictors of the trajectories of relationship satisfaction and depressive symptoms among couples in Taiwan over the course of EFT. Overall, the results provide meaningful evidence regarding sample characteristics that predict change in EFT. Identifying predictors of change in EFT is particularly important because this study is the first EFT effectiveness study among couples in Taiwan. The results provide insight regarding which couples in Taiwan might be most responsive to EFT. Attachment Avoidance and Anxiety The study findings indicated that attachment avoidance at intake did not predict change in relationship satisfaction over the course of EFT. However, men with high attachment avoidance at intake were found to experience a significant decrease (by 0.43 points on average) in depression each week during EFT. The nonsignificant association between attachment avoidance and relationship satisfaction is consistent with prior study findings (Dalgleish et al., 2015). High avoidance may indicate withdrawing patterns in a relationship, and such patterns make it difficult 67 for couples to engage in therapy because withdrawing partners use deactivating and over- regulating strategies as coping mechanisms (Horowitz et al., 1993; Myung et al., 2022). Withdrawer re-engagement, one of the key change events associated with EFT, is often characterized by withdrawers’ experiences of shame, grief, and hurt, all of which are emotions associated with deep sadness (Lee et al., 2017; Myung et al., 2022). The experience of sadness is a key indicator of a shift from a focus on others to a focus on self, and this transition provides opportunities for change (Myung et al., 2022). This pattern may explain the finding in the present study that men who had high attachment avoidance experienced a significant decrease in depression each week. Based on this theory, men’s depression would have decreased as they focused more on their own sadness than on their partner’s criticizing and blaming behavior. Additionally, findings indicated that attachment anxiety at intake did not predict changes in relationship satisfaction. However, attachment anxiety at intake did predict changes in depressive symptoms over the course of EFT, with men experiencing a decrease of 0.20 points and women experiencing a decrease of 0.22 points each week. These findings are inconsistent with prior study findings that attachment anxiety predicted growth in relationship satisfaction at the end of EFT (Johnson & Talitman, 1997) and during treatment (Dalgleish et al., 2015). Another important key change event in EFT is blamer softening, which requires successful withdrawer re-engagement, and blamer softening was previously found to predict change in relationship satisfaction (Burgess‐Moser et al., 2017; Dalgleish et al., 2015). However, the current study findings suggest that both men and women with high attachment anxiety at intake experienced a reduction in depressive symptoms over time. The present study findings highlight the importance of EFT as a tool to treat depression, particularly for men who are insecurely attached. Attachment anxiety is strongly related to 68 depression (Zheng et al., 2020), and EFT can serve as an effective treatment. Men with depression tend to use maladaptive coping strategies (e.g., withdrawal from relationship, abuse substance, etc.) to manage their depression instead of seeking help (Chuick et al., 2009). Men sometimes experience shame related to their depression and worry about being perceived as a failure by their partners (Wittenborn & Liu, 2012). Research has shown that men can be less emotionally expressive than women (Levant et al., 2013). EFT can serve as a helpful guide to encourage men to recognize and express emotions that arise in their relationships (Greenman et al., 2012). Traditional Values/Modern Values and Femininity/Masculinity In the present study, women who endorsed higher levels of traditional femininity ideology, masculinity ideology, and traditional values at intake experienced greater rates of change in relationship satisfaction throughout EFT. This finding indicates that women with traditional ideologies and values benefited more from EFT treatment than women who were less traditional. Johnson and Talitman (1997) also examined traditional values as a predictor of success of EFT but found that traditional values among couples did not predict EFT outcomes. A possible cultural explanation for this result is that several research studies have found that women in Taiwan who endorse more traditional gender roles experience greater relationship satisfaction in their relationships (e.g., Lin, 1996; Yu, 2006). Taiwan has undergone significant societal and economic changes in the last 50 years, and women are now encouraged to pursue higher education; however, traditional values regarding gender roles in relationships are still prevalent (Fan & Qian, 2022). Female traditionalism is considered an important factor for the maintenance and health of a relationship because women who endorse traditional values, traditional femininity ideology, and masculinity ideology meet societal expectations of 69 womanhood (Thompson & Walker, 1989). Women in Taiwan who are more traditional may benefit from EFT because EFT encourages women to share the emotions and desires they might otherwise suppress due to cultural expectations. Within traditional gender roles, women who express their needs might be seen as demanding because women are traditionally expected to accommodate the needs of others (Zeytinoglu-Saydam, 2018). EFT provides a safe space for women to explore their attachment needs, and therapists may encourage women clients to express their inner needs, which in turn can help them reconnect with their partners. Emotional Expressivity The study findings suggested that participants with higher emotional expressivity at intake experienced an increase in relationship satisfaction throughout EFT than participants with lower emotional expressivity. In addition, participants with higher emotional expressivity at intake reported a decrease in depressive symptoms throughout EFT than participants with lower emotional expressivity. This finding conflicts with prior study findings that partners who suppressed negative emotion at intake had greater changes in relationship satisfaction throughout EFT than those who were emotionally expressive at intake (Dalgleish et al., 2015). This finding highlighted the important role of emotion in EFT. Emotional accessibility and responsiveness are an important part of EFT (Johnson, 2019). In EFT, changes occur when partners can access and reprocess their emotions, leading to the creation of new corrective emotional experiences and new ways to connect with their partners (Johnson, 2019). In Taiwan, tolerance and suppression of one’s emotions are culturally appreciated because emotional control is considered a means to maintaining harmonious relationships (Chang et al., 2020; Chen & Li, 2007). Couples in Taiwan may consider therapy as a last resort and therefore may be more open to exploring their needs and wants, even if they are not used to being expressive due to cultural 70 expectations (Zeytinoglu-Saydam, 2018). Based on this perspective, EFT may be a particularly useful tool for couples in Taiwan if they are ready to be emotionally expressive in therapy. Though Taiwanese individuals and couples are not encouraged to express negative emotions in their cultural context (Chang et al., 2020), therapy can provide a safe place for them to identify and process their underlying emotions. Therapists working with couples in Taiwan should bear in mind the specific cultural perspective on tolerance and suppression of emotions. However, they should also consider that these cultural characteristics may enable therapy to provide an opportunity for couples to discover and express an emotional connection in a safe environment. Limitations There are some limitations to this study. First, the sample size was very small. Small sample sizes are expected in clinical studies (Bartle-Haring et al., 2020); however, a larger sample size would provide more statistical power and more confidence in the findings. Second, previous study findings have indicated that assessment of sample characteristics at baseline may not accurately predict outcomes and growth if these characteristics evolve over the course of treatment (Atkins et al., 2005). Future studies should examine the trajectories of changes in these characteristics over time as time-varying co-variates and assess whether change in personal characteristics are associated with change in clinical outcomes in EFT (Dalgleish et al., 2015). Conclusion This is the first study to examine predictors of change among couples in Taiwan receiving EFT. The study findings suggest that EFT may be particularly effective at increasing relationship satisfaction for women with high traditionalism at intake and for women and men with high emotional expressivity at intake. EFT was successful at decreasing depression for men with high attachment avoidance at intake and for men and women with high attachment anxiety 71 and emotional expressivity at intake. Future research is needed to explore who benefits most from EFT. 72 APPENDICES 73 APPENDIX A Tables 74 Table 1. Descriptive Information for Relationship Satisfaction and Depression Over Time Sample Size Relationship Satisfaction Depression M SD Alpha(α) M SD Alpha(α) Men Week 1 17 83.75 6.09 .94 10.13 1.68 .92 Week 2 17 84.93 6.61 .95 9.86 1.59 .88 Week 3 12 87.72 7.48 .95 6.63 1.63 .93 Week 4 11 91.90 7.12 .92 6.60 1.85 .95 Week 5 9 93.44 9.17 .95 7.89 2.21 .93 Week 6 9 84.00 8.67 .94 6.13 3.03 .98 Week 7 7 81.00 13.18 .97 7.17 3.94 .98 Week 8 7 83.86 11.92 .98 7.00 2.53 .96 Week 9 6 91.67 14.51 .97 7.83 2.96 .96 Week 10 6 89.80 11.18 .96 9.20 4.63 .98 Week 11 4 70.67 14.26 .97 12.25 4.07 .95 Week 12 3 78.50 20.50 .95 12.00 7.00 .95 Week 13 3 86.00 13.00 .94 12.00 5.00 .95 Week 14a 2 96.00 - - 4.00 - - Week 15a 2 99.00 - - 4.00 - - Week 16a 1 95.00 - - 7.00 - - Week 17a 1 99.00 - - 4.00 - - Women Week 1 17 86.60 4.85 .90 11.00 1.76 .94 Week 2 16 84.50 5.74 .87 9.64 1.48 .90 Week 3 12 100.50 4.87 .90 6.80 1.56 .92 Week 4 11 95.00 7.77 .95 7.50 1.76 .89 Week 5 9 95.22 5.03 .87 8.44 2.22 .96 Week 6 9 93.71 7.18 .91 7.28 1.67 .89 Week 7 7 87.29 7.99 .95 5.86 1.92 .92 Week 8 7 93.14 7.18 .95 4.42 1.56 .87 Week 9 6 87.17 14.37 .95 5.50 1.77 .91 Week 10 6 85.00 10.87 .98 7.33 2.47 .95 Week 11 4 76.00 20.07 .97 2.67 1.76 .88 Week 12 3 80.33 14.38 .93 8.33 2.96 .75 Week 13a 3 78.00 31.00 - 6.50 2.50 .90 Week 14a 2 86.00 22.00 - 6.00 3.00 .94 Week 15a 2 97.50 15.50 - 6.00 1.00 .56 Week 16a 1 109.00 - - 3.00 - - Week 17a 1 107.00 - - 2.00 - - 75 Table 2. Dyadic Multilevel Growth Model of Relationship Satisfaction and Depression Indistinguishable Dyadic Model of Indistinguishable Dyadic Model of Relationship Satisfaction Depression b SE CI p b SE CI p Fixed Effects Intercept 88.26* 4.80 78.11, 98.42 <.001 8.57* 1.32 5.80, 11.34 <.001 Linear Slope -.09 .30 -.70, .51 .76 -.15 .09 -0.32, 0.02 .089 Quadratic Slope .03* .01 0.00, 0.06 <.05 Random Effects Intercept Variance 430.16* 138.61 228.74, 808.93 <.05 31.21* 10.14 16.51, 59.00 <.05 Intercept Covariance .65* .16 0.24, 0.86 <.001 .67* .16 0.22, 0.88 <.001 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 76 Table 3. Means and Standard Deviations of Predictor Variables Variable N M SD Men Experiences in Close Relationships- 17 3.50 0.33 Attachment Avoidance Experiences in Close Relationships- 17 4.25 0.34 Attachment Anxiety Femininity Ideology 17 26.76 1.68 Male Role Norms 17 66.76 4.01 Traditional Values 17 41.06 2.87 Modern Values 17 66.65 1.44 Emotional Expressivity 16 62.69 2.18 Women Experiences in Close Relationships- 17 3.25 0.20 Attachment Avoidance Experiences in Close Relationships- 17 5.20 0.38 Attachment Anxiety Femininity Ideology 17 25.53 1.42 Male Role Norms 17 55.00 4.42 Traditional Values 17 32.47 3.28 Modern Values 17 72.65 1.96 Emotional Expressivity 15 60.80 2.78 77 Table 4. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Attachment Avoidance Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 88.42* 3.37 10.03 26.22 80.91, 95.93 .00 Rates of Change -.10 .32 97.65 -.31 -0.73, 0.53 .75 Gender -.59 2.69 8.98 -.22 -6.67, 5.48 .83 Rates of Change *Gender .11 .18 79.00 .62 -0.24, 0.46 .54 Rates of Change *Attachment Avoidance -.31 .30 132.76 -1.07 -0.90, 0.27 .29 Attachment Avoidance * Gender -.43 3.50 31.12 -.12 -7.57, 6.71 .90 Rates of Change *Attachment Avoidance*Gender .21 .31 145.78 .69 -0.40, 0.82 .49 Random Effect Initial Variance 245.68* 76.50 133.45, 452.30 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 78 Table 5. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Attachment Anxiety Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 87.51* 5.14 17.00 17.02 76.66, 98.36 .00 Rates of Change -.03 .31 97.67 -.11 -0.65, 0.58 .91 Gender -.42 2.65 15.05 -.16 -6.06, 5.22 .88 Rates of Change *Gender .11 .19 85.28 .57 -0.26, 0.48 .57 Rates of Change *Attachment Anxiety .12 .16 107.63 .78 -0.19, 0.44 .44 Attachment Anxiety * Gender 1.35 2.19 20.60 .61 -3.22, 5.91 .55 Rates of Change *Attachment Anxiety*Gender .03 .19 172.528 .14 -0.35, 0.41 .89 Random Effect Initial Variance 451.89* 152.32 233.41, 874.89 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 79 Table 6. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Femininity Ideology Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 89.29* 5.00 17.07 17.85 78.74, 99.85 .00 Rates of Change -.13 .28 86.72 -.47 -0.68, 0.43 .64 Gender .95 2.40 16.46 .40 -4.12, 6.02 .70 Rates of Change *Gender .11 .18 79.12 .60 -0.25, 0.46 .55 Rates of Change *Femininity Ideology .08* .03 101.67 2.65 0.02, 0.14 .01 Femininity Ideology * Gender -.07 .50 23.20 -.15 -1.12, 0.97 .89 Rates of Change *Femininity Ideology*Gender .12* .04 175.39 3.26 0.05, 0.19 .00 Random Effect Initial Variance 454.00* 147.78 239.87, 859.28 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 80 Table 7. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Male Role Norms Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 88.28* 4.96 17.20 17.79 77.82, 98.74 .00 Rates of Change .17 .31 107.67 .56 -0.44, 0.78 .58 Gender -.25 2.84 17.77 -.09 -6.22, 5.72 .93 Rates of Change *Gender .44* .21 106.38 2.16 0.04, 0.85 .03 Rates of Change *Male Role Norms .05* .02 151.50 2.97 0.02, 0.08 .00 Male Role Norms * Gender -.00 .18 19.68 -.01 -0.37, 0.37 .10 Rates of Change *Male Role Norms*Gender .04* .01 118.42 2.65 0.01, 0.07 .01 Random Effect Initial Variance 447.80* 144.40 238.02, 842.48 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 81 Table 8. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Traditional Values Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 88.85* 4.96 18.30 17.93 78.45, 99.25 .00 Rates of Change .13 .32 100.63 .40 -0.51, 0.77 .69 Gender .99 2.66 17.68 .37 -4.61, 6.59 .71 Rates of Change *Gender .14 .18 80.84 .74 -0.23, 0.50 .46 Rates of Change *Traditional Values .05 .02 130.62 1.99 .0.00, 0.10 .05 Traditional Values * Gender .14 .22 18.75 .65 -0.31, 0.59 .53 Rates of Change *Traditional Values*Gender .06* .02 112.43 2.56 0.01, 0.10 .01 Random Effect Initial Variance 413.48 137.06 215.93, 791.78 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 82 Table 9. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Modern Values Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 88.79* 4.90 18.56 18.12 78.51, 99.06 .00 Rates of Change -.07 .31 94.71 -.23 -0.68, 0.54 .82 Gender 1.90 2.84 17.55 .67 -4.07, 7.88 .51 Rates of Change *Gender .06 .18 81.73 .34 -0.29, 0.42 .73 Rates of Change *Modern Values .01 .05 160.50 .24 -0.08, 0.11 .81 Modern Values * Gender .22 .50 25.50 .45 -0.81, 1.26 .66 Rates of Change *Modern Values*Gender -.05 .04 126.42 -1.23 -0.14, 0.03 .22 Random Effect Initial Variance 417.42 134.40 222.07, 784.59 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 83 Table 10. Dyadic Multilevel Growth Model of Relationship Satisfaction Trajectories: Estimated Effects of Predictor Emotional Expressivity Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 87.55* 4.48 12.08 19.53 77.80, 97.31 .00 Rates of Change .03 .32 92.63 .10 -0.60, 0.66 .92 Gender .43 2.87 10.45 .15 -5.92, 6.78 .88 Rates of Change *Gender -.02 .17 79.31 -.12 -0.36, 0.32 .91 Rates of Change *Emotional Expressivity .08* .03 139.59 3.12 0.03, 0.14 .00 Emotional Expressivity * Gender .54 .37 16.58 1.46 -0.24, 1.33 .16 Rates of Change *Emotional Expressivity*Gender -.02 .02 103.72 -.87 -0.07, 0.03 .39 Random Effect Initial Variance 373.20 125.41 193.15, 721.07 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 84 Table 11. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Attachment Avoidance Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.32* 1.25 14.81 7.48 6.66, 11.98 .00 Rates of Change -.06 .07 101.42 -.81 -0.20, 0.09 .42 Gender .75 .69 12.89 1.09 -0.74, 2.24 .30 Rates of Change *Gender -.09 .07 104.18 -1.33 -0.22, 0.04 .19 Rates of Change *Attachment Avoidance -.22* .09 182.16 -2.41 -0.41, -0.04 .02 Attachment Avoidance * Gender -.65 .92 19.18 -.71 -2.57, 1.26 .48 Rates of Change *Attachment Avoidance*Gender .21* .09 185.29 2.23 0.02, 0.39 .03 Random Effect Initial Variance 29.37* 9.80 15.28, 56.45 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 85 Table 12. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Attachment Anxiety Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 8.83* 1.24 17.14 7.11 6.21, 11.44 .00 Rates of Change -.11 .07 102.14 -1.49 -0.25, 0.03 .14 Gender .14 .76 13.75 .18 -1.48, 1.76 .86 Rates of Change *Gender -.15* .06 103.48 -2.28 -0.28, -0.02 .03 Rates of Change *Attachment Anxiety -.21* .05 158.15 -4.30 -0.31, -0.12 .00 Attachment Anxiety * Gender .40 .61 19.99 .66 -0.87, 1.67 .52 Rates of Change *Attachment Anxiety*Gender -.01 .05 170.03 -.16 -0.11, 0.09 .88 Random Effect Initial Variance 27.47* 8.94 14.51, 52.00 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 86 Table 13. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Femininity Ideology Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.65* 1.40 16.42 6.90 6.69, 12.60 .00 Rates of Change -.04 .07 97.62 -.60 -0.18, 0.10 .55 Gender .97 .65 16.62 1.48 -0.41, 2.34 .16 Rates of Change *Gender -.12 .07 104.49 -1.78 -0.25, 0.01 .08 Rates of Change *Femininity Ideology -.00 .01 156.38 -.28 -0.02, 0.02 .78 Femininity Ideology * Gender .26 .14 20.85 1.93 -0.02, 0.55 .07 Rates of Change *Femininity Ideology*Gender -.02 .01 166.46 -1.90 -0.04, 0.00 .06 Random Effect Initial Variance 34.77* 11.72 17.96, 67.34 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 87 Table 14. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Male Role Norms Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.81* 1.42 16.78 6.92 6.82, 12.81 .00 Rates of Change -.04 .08 122.97 -.48 -0.20, 0.12 .63 Gender .68 .69 15.15 .99 -0.78, 2.15 .34 Rates of Change *Gender -.13 .07 134.40 -1.76 -0.28, 0.17 .08 Rates of Change *Male Role Norms -.00 .00 177.95 -.16 -0.01, 0.09 .87 Male Role Norms * Gender .08 .04 15.43 1.80 -0.01, 0.18 .09 Rates of Change *Male Role Norms*Gender -.00 .00 173.51 -.16 -0.01, 0.01 .88 Random Effects Initial Variance 34.40* 11.84 17.52, 67.52 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 88 Table 15. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Traditional Values Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.31* 1.36 17.46 6.86 6.45, 12.16 .00 Rates of Change -.06 .08 109.99 -.78 -0.21, 0.09 .44 Gender .53 .83 16.34 .65 -1.21, 2.28 .53 Rates of Change *Gender -.12 .07 110.36 -1.74 -0.27, 0.02 .09 Rates of Change *Traditional Values -.00 .01 181.78 -.60 -0.02, 0.01 .55 Traditional Values * Gender .03 .07 17.41 .52 -0.12, 0.18 .61 Rates of Change *Traditional Values*Gender -.01 .01 176.51 -1.02 -0.02, 0.01 .31 Random Effects Initial Variance 32.50* 10.52 17.24, 61.30 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 89 Table 16. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Modern Values Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.45* 1.27 17.46 7.45 6.78, 12.11 .00 Rates of Change -.04 .07 100.52 -.56 -0.19, 0.10 .58 Gender .06 .77 16.14 .08 -1.57, 1.70 .94 Rates of Change *Gender -.11 .07 101.58 -1.62 -0.24, 0.02 .11 Rates of Change *Modern Values -.01 .01 186.05 -.59 -0.04, 0.02 .56 Modern Values * Gender -.15 .14 24.39 -1.10 -0.43, 0.13 .28 Rates of Change *Modern Values*Gender .01 .01 176.23 1.02 -0.01, 0.04 .31 Random Effects Initial Variance 27.89* 9.25 14.56, 53.43 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 90 Table 17. Dyadic Multilevel Growth Model of Depressive Symptoms Trajectories: Estimated Effects of Predictor Emotional Expressivity Undistinguishable Dyad Model b SE df t CI p Fixed Effects Intercept 9.42* 1.22 14.58 7.72 6.82, 12.03 .00 Rates of Change -.06 .07 98.69 -.80 -0.20, 0.08 .43 Gender .46 .64 17.33 .71 -0.90, 1.82 .49 Rates of Change *Gender -.11 .06 104.40 -1.76 -0.24, 0.01 .08 Rates of Change *Emotional Expressivity -.02* .01 172.84 -2.71 -0.04, -0.01 .01 Emotional Expressivity * Gender -.04 .08 19.55 -.54 -0.21, 0.13 .59 Rates of Change *Emotional Expressivity*Gender -.01 .01 166.58 -1.39 -0.03, 0.00 .17 Random Effect Initial Variance 25.22 8.85 12.67, 50.17 .00 Notes: SE = Standard error. CI = 95% Confidence intervals. *p < .05 91 APPENDIX B Figures 92 Figure 1. Relationship Satisfaction Trajectories Over the Course of EFT. 120 100 80 60 40 20 0 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Observed Men Relationship Satisfaction Observed Women Relationship Satisfaction Predicted Men Relationship Satisfaction Predicted Women Relationshhip Satisfaction Overall Observed Relationship Satisfaction Overall Predicted Relationship Satisfaction Note: There is no sex difference in the DAS model since this is an indistinguishable model. This graph still shows what men and women’s trajectories of relationship satisfaction look like over time. 93 Figure 2. Depressive Symptoms Trajectories Over the Course of EFT. 16 14 12 10 8 6 4 2 0 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Observed Men Depression Observed Women Depression Overall Observed Depression Overall Predicted Depression Quadratic Slope Note: There is no sex difference in the PHQ-9 model since this is an indistinguishable model. 94 Figure 3. Femininity Ideology Association with Rate of Change in Relationship Satisfaction Trajectory. 94 93 92 91 90 89 88 87 86 85 84 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Predicted Relationship Satisfaction for Men Predicted Relationship Satisfaction for Women Note: The rate of change in relationship satisfaction is only significant for women. 95 Figure 4. Male Role Norms Association with Rate of Change in Relationship Satisfaction Trajectory. 90 89.5 89 88.5 88 87.5 87 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Predicted Relationship Satisfaction for Men Predicted Relationship Satisfaction for Women Note: The rate of change in relationship satisfaction is only significant for women. 96 Figure 5. Traditional Values Association with Rate of Change in Relationship Satisfaction Trajectory. 92 91 90 89 88 87 86 85 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Predicted Relationship Satisfaction for Men Predicted Relationship Satisfaction for Women Note: The rate of change in relationship satisfaction is only significant for women. 97 Figure 6. Attachment Avoidance Association with Rate of Change in Depressive Symptoms Trajectory. 12 10 8 6 4 2 0 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Predicted Depressive Symptoms for Men Predicted Depressive Symptoms for Women Note: The rate of change in depression is only significant for men. 98 Figure 7. Attachment Anxiety Association with Rate of Change in Depressive Symptoms Trajectory. 10 9 8 7 6 5 4 3 2 1 0 Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Predicted Depressive Symptoms for Men Predicted Depressive Symptoms for Women Note: The rate of change in depression is significant for both men and women. 99 BIBLIOGRAPHY 100 BIBLIOGRAPHY Ahmadi, F. S., Zarei, E., & Fallahchai, S. R. (2014). The effectiveness of emotionally-focused couple therapy in resolution of marital conflicts between the couples who visited the consultation centers. Journal of Educational and Management Studies, 4(1), 118–123. Alan, R., & Zeidlhack, D. (2021). Treatment: Emotionally Focused Couples Therapy. https://div12.org/treatment/emotionally-focused-couples-therapy/ Asia EFT. (n.d.). ICCEFT Certified EFT Therapist in Asia. https://www.asia- eft.com/therapist_profile/ Atkins, D. C., Dimidjian, S., Bedics, J. D., & Christensen, A. (2009). Couple discord and depression in couples during couple therapy and in depressed individuals during depression treatment. Journal of Consulting and Clinical Psychology, 77(6), 1089–1099. https://doi.org/10.1037/a0017119 Atkins, D. C., Doss, B. D., Berns, S. B., George, W. H., Gattis, K., & Christensen, A. (2005). Prediction of response to treatment in a randomized clinical trial of marital therapy. Journal of Consulting and Clinical Psychology, 73(5), 893–903. https://doi.org/10.1037/0022- 006X.73.5.893 Arnault, D. S. (2009). Cultural determinants of help seeking: a model for research and practice. Research and Theory for Nursing Practice, 23(4), 259-278. Barbato, A., D’Avanzo, B., & Parabiaghi, A. (2018). Couple therapy for depression (Review). In Cochrane Database of Systematic Reviews (Vol. 6). https://doi.org/10.1080/01612840.2019.1688439 Bartholomew, K., & Horowitz, L. (1991). Attachment styles among young adults. Journal of Personality and Social Psychology, 61, 226–244. Bartle-Haring, S., Washburn-Busk, M., & VanBergen, A. (2020). A primer for the use of multilevel models in couple and family therapy research. Journal of Marital and Family Therapy, 46(4), 582-602. Beach, S. R. H., Katz, J., Kim, S., & Brody, G. H. (2003). Prospective effects of marital satisfaction on depressive symptoms in established marriages: A dyadic model. Journal of Social and Personal Relationships, 20(3), 355–371. https://doi.org/10.1177/0265407503020003005 Beach, S. R. H., & O’Leary, K. D. (1993). Dysphoria and marital discord: Are dysphoric individuals at risk for marital maladjustment? Journal of Marital and Family Therapy, 19(4), 355–368. https://doi.org/10.1111/j.1752-0606.1993.tb00998.x 101 Beasley, C. C., & Ager, R. (2019). Emotionally Focused Couples Therapy: A systematic review of its effectiveness over the past 19 Years. Journal of Evidence-Informed Social Work, 16(2), 144–159. https://doi.org/10.1080/23761407.2018.1563013 Behling, O., & Law, K. S. (2000). Translating questionnaires and other research instruments: Problems and solutions. SAGE Publications, Inc. Bell, B.A., Schoeneberger, J.A., Leighton, E., Haines, S.K., Ene, M., Smiley, W., & Kromrey, J. D. (2013, May). Doubly-diminishing returns: An empirical investigation of the impact of sample size and predictor prevalence on point and interval estimates from two- level linear models. Presented at the Modern Modeling Methods Conference. Storrs, CT, United States. Bernal, G., Jiménez-Chafey, M. I., & Domenech Rodríguez, M. M. (2009). Cultural adaptation of treatments: A resource for considering culture in evidence-based practice. Professional Psychology: Research and Practice, 40(4), 361–368. https://doi.org/10.1037/a0016401 Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books. Bowlby, J. (1988). A secure base. Basic Books. Bowlby, J. (1991). Postscript. In C. M. Parkes, J. Stevenson-Hinde, & P. Marris (Eds.), Attachment across the lifespan (pp. 293–297). Routledge. Burgess Moser, M., Johnson, S. M., Dalgleish, T. L., Lafontaine, M. F., Wiebe, S. A., & Tasca, G. A. (2016). Changes in relationship-specific attachment in Emotionally Focused Couple Therapy. Journal of Marital and Family Therapy, 42(2), 231–245. https://doi.org/10.1111/jmft.12139 Centers for Disease Control and Prevention (2018). Leading causes of death—Males—United States, 2018. https://www.cdc.gov/healthequity/lcod/men/2018/index.html Chan, R. C. K., Wang, Y., Li, H., Shi, Y., Wang, Y., Liu, W., & Huang, J. (2010). A 2-stage factor analysis of the Emotional Expressivity Scale in the Chinese context. Psychologia, 53, 44–50. https://doi.org/10.2117/psysoc.2010.44 Chang, S. C., Chang, J. H. Y., Low, M. Y., Chen, T. C., & Kuo, S. H. (2020). Self-regulation of the newlyweds in Taiwan: Goals and strategies. Journal of Social and Personal Relationships, 37(8–9), 2674–2690. https://doi.org/10.1177/0265407520929762 Chen, F. M., & Li, T. S. (2007). Marital enqing: An examination of its relationship to spousal contributions, sacrifices, and family stress in Chinese marriages. Journal of Social Psychology, 147(4), 393–412. https://doi.org/10.3200/SOCP.147.4.393-412 Cheng, Y. A., & Yang, C. W. (2021). Continuity and changes in attitudes toward marriage in contemporary Taiwan. Journal of Population Research, 38(2), 139–167. 102 https://doi.org/10.1007/s12546-021-09259-z Chuick, C. D., Greenfeld, J. M., Greenberg, S. T., Shepard, S. J., Cochran, S. V., & Haley, J. T. (2009). A qualitative investigation of depression in men. Psychology of Men & Masculinity, 10(4), 302–313. https://doi.org/10.1037/a0016672 Cloutier, P. F., Manion, I. G., Walker, J. G., & Johnson, S. M. (2002). Emotionally focused interventions for couples with chronically ill children: A 2-year follow-up. Journal of Marital and Family Therapy, 28(4), 391–398. https://doi.org/10.1111/j.1752- 0606.2002.tb00364.x Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Dalgleish, T. L., Johnson, S. M., Burgess Moser, M., Lafontaine, M. F., Wiebe, S. A., & Tasca, G. A. (2015). Predicting change in marital satisfaction throughout Emotionally Focused Couple Therapy. Journal of Marital and Family Therapy, 41(3), 276–291. https://doi.org/10.1111/jmft.12077 Dalton, E. J., Greenman, P. S., Classen, C. C., & Johnson, S. M. (2013). Nurturing connections in the aftermath of childhood trauma: A randomized controlled trial of emotionally focused couple therapy for female survivors of childhood abuse. Couple and Family Psychology: Research and Practice, 2(3), 209–221. https://doi.org/10.1037/a0032772 Dandeneau, M. L., & Johnson, S. M. (1994). Facilitating intimacy: Interventions and effects. Journal of Marital and Family Therapy, 20(1), 17- 33. https://doi.org/10.1111/j.1752-0606.1994.tb01008.x Davis, S. N., & Greenstein, T. N. (2009). Gender ideology: Components, predictors, and consequences. Annual Review of Sociology, 35(1), 87–105. https://doi.org/10.1146/annurev- soc-070308-115920 Denton, W. H., Burleson, B. R., Clark, T. E., Rodriguez, C. P., & Hobbs, B. V. (2000). A randomized trial of emotion-focused therapy for couples in a training clinic. Journal of Marital and Family Therapy, 26(1), 65–78. https://doi.org/10.1111/j.1752- 0606.2000.tb00277.x Denton, W. H., Wittenborn, A. K., & Golden, R. N. (2012). Augmenting antidepressant medication treatment of depressed women with Emotionally Focused Therapy for couples: A randomized pilot study. Journal of Marital and Family Therapy, 38(s1), 23–38. https://doi.org/10.1111/j.1752-0606..2012.00291.x Dessaulles, A., Johnson, S. M., & Denton, W. H. (2003). Emotion-focused therapy for couples in the treatment of depression: A pilot study. American Journal of Family Therapy, 31(5), 345–353. https://doi.org/10.1080/01926180390232266 103 Ditommaso, E., Brannen, C., & Burgess, M. (2005). The universality of relationship characteristics: a cross-cultural comparison of different types of attachment and loneliness in Canadian and visiting Chinese students. Social Behavior and Personality, 33(1), 57–68. Doss, B. D., Roddy, M. K., Wiebe, S. A., & Johnson, S. M. (2022). A review of the research during 2010-2019 on evidence-based treatments for couple relationship distress. Journal of Marital and Family Therapy, 48(1), 283-306. Ekman, P. (2007). Emotions revealed (2nd ed.). Henry Holt. Ekman, P. (2018). Are facial expressions universal or culturally specific. https://www.paulekman.com/blog/facial-expressions-universal-culturally-specific/ Fan, W. & Qian, Y. (2022) Constellations of gender ideology, earnings arrangements, and marital satisfaction: a comparison across four East Asian societies. Asian Population Studies, 18(1), 24-40, doi: 10.1080/17441730.2021.1932067 Foran, H. M., Whisman, M. A., & Beach, S. R. H. (2015). Intimate partner relationship distress in the DSM-5. Family Process, 54(1), 48–63. Fraley, R. C., Heffernan, M. E., Vicary, A. M., & Brumbaugh, C. C. (2011). The Experiences in Close Relationships-Relationship Structures Questionnaire: A method for assessing attachment orientations across relationships. Psychological Assessment, 23(3), 615–625. https://doi.org/10.1037/a0022898 Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self- report measures of adult attachment. Journal of Personality and Social Psychology, 78(2), 350–365. https://doi.org/10.1037/0022-3514.78.2.350 Frijda, N. H. (1986). The emotions. Cambridge University Press. Geisinger, K. F. (1994). Cross-cultural normative assessment: Translation and adaptation issues influencing the normative interpretation of assessment instruments. Psychological Assessment, 6(4), 304–312. Greenman, P. S., Faller, G., & Johnson, S. M. (2012). Finding the words: Working with men in Emotionally Focused Therapy (EFT) for couples. In D. S. Shepard & M. Harway (Eds.), Engaging men in couples therapy (pp. 129–151). Routledge/Taylor & Francis Group. Grimm, K. J., Ram, N., & Estabrook, R. (2017). Growth modeling: structural equation and multilevel modeling approaches. The Guilford Press. Guo, Y., Wang, S., Combs, D., Lin, Y., & Johnson, V. (2013). Professional counseling in Taiwan: past to future. Journal of Counseling and Development, 91(3), 331-335. 104 Gupta, S. (2011). Intention-to-treat concept: A review. Perspectives in Clinical Research, 2(3), 109–112. https://doi.org/10.4103/2229-3485.83221 Heritier, S. R., Gebski, V. J., & Keech, A. C. (2003). Inclusion of patients in clinical trial analysis: The intention-to-treat principle. Medical Journal of Australia, 179(8), 438–440. https://doi.org/10.5694/j.1326-5377.2003.tb05627.x Horowitz, L. M., Rosenberg, S. E., & Bartholomew, K. (1993). Interpersonal problems, attachment styles, and outcome in brief dynamic psychotherapy. Journal of Consulting and Clinical Psychology, 61(4), 549-560. doi: 10.1037//0022-006x.61.4.549. Huang, K. K. (2009). Confucius Relationships: Philosophy, theory construction, and evidence research. Taipei: Psychology Publisher. Hwang, K. K., and Han, K. H. (2010). Face and morality in Confucian society, in Oxford Handbook of Chinese Psychology, M. H. Bond (Hong Kong: Oxford University Press), 479–498. Inglehart, R., & Baker, W. E. (2000). Modernization , cultural change, and the persistence of traditional values. American Sociological Review, 65, 19–51. Institute of Medicine (US) Committee on Strategies for Small-Number-Participant Clinical Research Trials. Small Clinical Trials: Issues and Challenges. Evans C. Jr, Ildstad S. T., editors. Washington (DC): National Academies Press (US); 2001. PMID: 25057552. Jacobson, N. S., Follette, W. C., & Revenstorf, D. (1984). Toward a standard definition of clinically significant change. Behavior Therapy, 17, 308-311. Jacobson, N. S., Follette, W. C., & Pagel, M. (1986). Predicting who will benefit from Behavioral Marital Therapy. Journal of Consulting and Clinical Psychology, 54(4), 518– 522. https://doi.org/10.1037/0022-006X.54.4.518 Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12-19. Johnson, S. M., & Greenberg, L. S. (1985). Emotionally Focused Couples Therapy: an outcome study. Journal of Marital and Family Therapy, 11(3), 313–317. https://doi.org/10.1111/j.1752-0606.1985.tb00624.x Johnson, S. M., & Talitman, E. (1997). Predictors of success in Emotionally Focused Marital Therapy. Journal of Marital and Family Therapy, 23(2), 135–152. https://doi.org/10.1111/j.1752-0606.1997.tb00239.x Johnson, S. M., Hunsley, J., Greenberg, L., & Schindler, D. (1999). Emotionally Focused Couples Therapy: status and challenges. Clinical Psychology: Science and Practice, 6(1), 105 67–79. https://doi.org/10.1093/clipsy/6.1.67 Johnson, S. M. (2019). The practice of Emotionally Focused Couple Therapy (3nd ed.). New York: Routledge. Johnson, S. M. (2019). Attachment theory in practice: Emotionally Focused Therapy in individuals, couples, and families. The Guilford Press. Kao, S.-F., & Lu, L. (2006). The relationship of congruence in psychological traditionality/modernity to marital adjustment (夫妻傳統性⁄現代性的契合與婚姻適應之 關聯). Indigenous Psychological Research in Chinese Societies, 25, 47–100. Kashy, D. A., & Donnellan, M. B. (2008). Comparing MLM and SEM approaches to analyzing developmental dyadic data: Growth curve models of hostility in families. In N. A. Card, J. P. Selig, & T. Little (Eds.), Modeling Dyadic and Interdependent Data in the Developmental and Behavioral Sciences (pp. 165–190). Routledge Taylor & Francis Group. https://doi.org/10.1016/j.paid.2011.03.037 Kashy, D. A., Donnellan, M. B., Burt, S. A., & McGue, M. (2008). Growth curve models for Indistinguishable dyads using multilevel modeling and structural equation modeling: The case of adolescent twins’ conflict with their mothers. Developmental Psychology, 44(2), 316–329. https://doi.org/10.1037/0012-1649.44.2.316 Kenny, D. A., Kashy, D. A., Cook, W. L., & Simpson, J. A. (2006). Dyadic data analysis. The Guilford Press. Kenward, M. G., & Roger, J. H. (1997). Small sample inference for fixed effects from restricted maximum likelihood. Biometrics, 53, 983–997. Kring, A. M., Smith, D. A., & Neale, J. M. (1994). Individual differences in dispositional expressiveness: Development and validation of the Emotional Expressivity Scale. Journal of Personality and Social Psychology, 66(5), 934–949. https://doi.org/10.1037/0022- 3514.66.5.934 Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of Gen Intern Med, 16, 606–613. Kruse, R. L., Alper, B. S., Reust, C., Stevermer, J. J., Shannon, S., & Williams, R. (2002). Intention-to-treat analysis: Who is in? Who is out. The Journal of Family Practice, 51(11), 969–971. Kudva, K. G., El Hayek, S., Gupta, A. K., Kurokawa, S., Bangshan, L., Armas-Villavicencio, M. V. C., Oishi, K., Mishra, S., Tiensuntisook, S., & Sartorius, N. (2020). Stigma in mental illness: Perspective from eight Asian nations. Asia-Pacific Psychiatry, 12(2). https://doi.org/10.1111/appy.12380 106 Lauber, C., & Rössler, W. (2007). Stigma towards people with mental illness in developing countries in Asia. International Review of Psychiatry, 19(2), 157–178. https://doi.org/10.1080/09540260701278903 Lebow, J. L., Chambers, A. L., Christensen, A., & Johnson, S. M. (2012). Research on the treatment of couple distress. Journal of Marital and Family Therapy, 38(1), 145–168. https://doi.org/10.1111/j.1752-0606.2011.00249.x Ledermann, T., & Kenny, D. A. (2017). Analyzing dyadic data with multilevel modeling versus structural equation modeling: A tale of two methods. Journal of Family Psychology, 31(4), 442–452. https://doi.org/10.1037/fam0000290 Lee, N. A., Spengler, P. M., Mitchell, A. M., Spengler, E. S., & Spiker, D. A. (2017). Facilitating withdrawer re‐engagement in emotionally focused couple therapy: A modified task analysis. Couple and Family Psychology: Research and Practice, 6, 205–225. https://doi.org/10.1037/cfp0000084 Levant, R. F., Alto, K. M., McKelvey, D. K., Richmond, K. A., & McDermott, R. C. (2017). Variance composition, measurement invariance by gender, and construct validity of the femininity ideology scale-short form. Journal of Counseling Psychology, 64(6), 708–723. https://doi.org/10.1037/cou0000230 Levant, R. F., Hall, R. J., & Rankin, T. J. (2013). Male role norms inventory-short form (mrni- sf): Development, confirmatory factor analytic investigation of structure, and measurement invariance across gender. Journal of Counseling Psychology, 60(2), 228–238. https://doi.org/10.1037/a0031545 Levant, R. F., & Richmond, K. (2007). A review of research on masculinity ideologies using the Male Role Norms Inventory. Journal of Men’s Studies, 15, 130–146. doi:10.3149/jms.1502.130 Leyland, A. H. (2004). A review of multilevel modelling in SPSS. https://www.bristol.ac.uk/media-library/sites/cmm/migrated/documents/reviewspss.pdf Li, T. S. (2006). Chinese marital resilience: Conceptual clarification and theory construction (華 人婚姻韌性的形成與變化：概念釐清與理論建構). Indigenous Psychological Research in Chinese Societies, 25, 101–137. https://doi.org/10.6254/2006.25.101 Li, T. S. (2012). Ren (Forbearance) in couple relationship and how it is related to marital satisfaction. Formosa Journal of Mental Health, 25(3), 447–475. Li, T. S., & Hsiao, Y. (2016). Implicit affection in Taiwanese couples: gratitude, forbearance, and marital satisfaction (夫妻內隱情感經驗：感恩、忍與婚姻滿意度). Indigenous Psychological Research in Chinese Societies, 45, 93–128. https://doi.org/10.6254/2016.45.93 107 Liao, S. C., Chen, W. J., Lee, M. B., Lung, F. W., Lai, T. J., Liu, C. Y., Lin, C. Y., Yang, M. J., & Chen, C. C. (2012). Low prevalence of major depressive disorder in Taiwanese adults: Possible explanations and implications. Psychological Medicine, 42(6), 1227–1237. https://doi.org/10.1017/S0033291711002364 Lim, N. (2016). Cultural differences in emotion: differences in emotional arousal level between the East and the West. Integrative Medicine Research, 5(2), 105–109. https://doi.org/10.1016/j.imr.2016.03.004 Liu, F., Li, C., & Gao, X. (2018). Reliability and validity of the Chinese Version of the Male Role Norms Inventory-Revised. Advances in Psychology, 8(1), 120–130. https://doi.org/10.12677/ap.2018.81015 Liu, H., & Waite, L. (2014). Bad marriage, broken heart? Age and gender differences in the link between marital quality and cardiovascular risks among older adults. Journal of Health and Social Behavior, 55(4), 403–423. https://doi.org/10.1038/jid.2014.371 Liu, T., & Wittenborn, A. K. (2011). Emotionally Focused Therapy with culturall diversed couples. In & B. A. B. J. L. Furrow, S. M. Johnson (Ed.), The Emotionally Focused casebook (pp. 295–316). Routledge. Liu, Ting, & Hung, S.-T. (2019). The clinical practice of Emotionally Focused Couple Therapy with an in-expressive partner: A case example (情緒取向伴侶治療的實務應用：以一方 為情緒表達含蓄之伴侶個案為例). The Archive of Guidance and Counseling, 41(2), 57– 77. Matsumoto, D., Yoo, S. H., Fontaine, J., Anguas-Wong, A. M., Arriola, M., Ataca, B., Bond, M. H., Boratav, H. B., Breugelmans, S. M., Cabecinhas, R., Chae, J., Chin, W. H., Comunian, A. L., Degere, D. N., Djunaidi, A., Fok, H. K., Friedlmeier, W., Ghosh, A., Glamcevski, M., … Grossi, E. (2008). Mapping expressive differences around the world: The relationship between emotional display rules and individualism versus collectivism. Journal of Cross- Cultural Psychology, 39(1), 55–74. https://doi.org/10.1177/0022022107311854 McFarlane, A. C., & van der Kolk, B. A. (1996). Trauma and its challenge to society. In B. A. van der Kolk, A. C. Mcfarlane, & L. Weisaeth (Eds.), Traumatic stress (pp. 211–215). Guilford Press. McLean, L. M., Walton, T., Rodin, G., Esplen, M. J., & Jones, J. M. (2013). A couple-based intervention for patients and caregivers facing end-stage cancer: Outcomes of a randomized controlled trial. Psycho-Oncology, 22(1), 28–38. https://doi.org/10.1002/pon.2046 McNeish, D. M., & Stapleton, L. M. (2016). The effect of small sample size on two-level model estimates: A review and illustration. Educational Psychology Review, 28(2), 295–314. https://doi.org/10.1007/s10648-014-9287-x Miller, R. B. (2020). An introduction to a new generation of clinical statistical approaches. 108 Journal of Marital and Family Therapy, 46(4), 577-581. Ministry of Health and Welfare. (2004). Health knowledge, attitudes, and behaviors among Taiwanese. https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=364&pid=6539 Ministry of Health and Welfare. (2020). Antidepressants usage statistics. https://www.mohw.gov.tw/dl-16605-eee5f1f6-b653-46ce-9931-9d8fa54959d7.html Ministry of the Interior Taiwan Department of Statistics. (2020). 2021 Divorce rate. https://www.moi.gov.tw/CL.aspx?n=3920 Ministry of Health and Welfare (2021). Including counseling and therapy into the coverage of National Health Insurance. https://www.mohw.gov.tw/cp-5021-63448-1.html Motaharinasab, A., Zare Bahramabadi, M., & Ahmadi, K. (2016). The effect of Emotionally Focused Therapy on emotional expression styles among married women. International Journal of Mental Health and Addiction, 14(4), 505–513. https://doi.org/10.1007/s11469- 015-9614-6 Myung, H. S., Furrow, J. L., & Lee, N. A. (2022). Understanding the emotional landscape in withdrawer re-engagement and blamer softening EFCT change events. Journal of Marital and Family Therapy. Naaman, S. C. (2009). Evaluation of the clinical efficacy of Emotionally Focused Therapy on psychological adjustment and natural killer cell cytotoxicity in early breast cancer. Dissertation Abstracts International, 69, 5044. National Institute of Mental Health. (2019). Depression. https://www.nimh.nih.gov/health/topics/depression National Institute for Health and Care Excellence (2009). Depression in adults: Recognition and management. https://www.nice.org.uk/guidance/cg90 Oliver, M. I., Pearson, N., Coe, N., Gunnell, D. (2005). Help-seeking behavior in men and women with common mental health problems: cross-sectional study. British Journal of Psychiatry, 186-297. Onken, L. S., Carroll, K. M., Shoham, V., Cuthbert, B. N., & Riddle, M. (2014). Reenvisioning clinical science: Unifying the discipline to improve the public health. Clinical Psychological Science, 2(1), 22–34. https://doi.org/10.1177/2167702613497932 Online Etymology Dictionary. (n.d.). emotion (n.). https://www.etymonline.com/word/emotion Pan, Y., & Puente, M. D. L. (2005). Census bureau guideline for the translation of data collection instruments and supporting materials: documentation on how the guideline was developed. https://www.census.gov/srd/papers/pdf/rsm2005-06.pdf 109 Planalp, E. M., Du, H., Braungart-Rieker, J. M., & Wang, L. (2017). Growth curve modeling to studying change: A comparison of approaches using longitudinal dyadic data with distinguishable dyads. Structural Equation Modeling, 24(1), 129–147. https://doi.org/10.1080/10705511.2016.1224088 Qian, Y. & Sayer, L. C. (2016). Division of labor, gender ideology, and marital satisfaction in East Asia. Journal of Marriage and Family, 78(2), 383-400. https://doi.org/10.1111/jomf.12274 Rathgeber, M., Bürkner, P.- C., Schiller, E.- M., & Holling, H. (2019). The efficacy of emotionally focused couples therapy and behavioral couples therapy: A meta-analysis. Journal of Marital and Family Therapy, 45(3), 447– 463. https://doi.org/10.1111/jmft.12336 Raudenbush, S.W., & Bryk, A.S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Robles, T. F., Slatcher, R. B., Trombello, J. M., & McGinn, M. M. (2014). Marital quality and health: A meta-analytic review. Psychological Bulletin, 140(1), 140–187. https://doi.org/10.1037/a0031859 Rothbaum, F., Weisz, J., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security in the United States and Japan. American Psychologist, 55(10), 1093–1104. https://doi.org/10.1037/0003-066X.55.10.1093 Ross, T. (2004). Attachment representation, attachment style, or attachment pattern? Usage of terminolog in attachment theory. In F. Pfäfflin & G. Adshead (Eds.), A matter of security: The application of attachment theory to forensic psychiatry and psychotherapy (pp. 57–84). Jessica Kinsgley, London Roth, A., Fonagy, P., & Parry, G. (2005). What works for whom? A critical review of psychotherapy research (2nd ed.). Guilford Publications. Ryder, A. G., Yang, J., Zhu, X., Yao, S., Yi, J., Heine, S. J., & Bagby, R. M. (2008). The cultural shaping of depression: Somatic symptoms in China, psychological symptoms in North America? Journal of Abnormal Psychology, 117(2), 300–313. https://doi.org/10.1037/0021- 843X.117.2.300 Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177. https://doi.org/10.1037/1082-989X.7.2.147 Schmitt, D. P., Alcalay, L., Allensworth, M., Allik, J., Ault, L., Austers, I., Bennett, K. L., Bianchi, G., Boholst, F., Cunen, M. A. B., Braeckman, J., Brainerd, E. G., Caral, L. G. A., Caron, G., Casullo, M. M., Cunningham, M., Daibo, I., De Backer, C., De Souza, E., … Shackelford, T. K. (2004). Patterns and universals of adult romantic attachment across 62 110 cultural regions: Are models of self and of other pancultural constructs? Journal of Cross- Cultural Psychology, 35(4), 367-402. https://doi.org/10.1177/0022022104266105 Shek, D. T. L. (1995). The Chinese version of the Dyadic Adjustment Scale: Does language make a difference? Journal of Clinical Psychology, 51(6), 802–811. https://doi.org/10.1002/1097-4679 Shek, D. T. L., & Cheung, C. K. (2008). Dimensionality of the Chinese Dyadic Adjustment Scale based on confirmatory factor analysis. Soc Indic Res, 86, 201–212. Shek, D. T. L., Lam, M. C., Tsoi, K. W., & Lam, C. M. (1993). Marital assessment in a Chinese context: an evaluation of three instruments. Hong Kong Journal of Social Work, 27, 80–81. Simon, G. E., Coleman, K. J., Rossom, R. C., Beck, A., Oliver, M., Johnson, E., Whiteside, U., Operskalski, B., Penfold, R. B., Shortreed, S. M. (2016). Risk of suicide attempt and suicide death following completion of the patient health questionnaire depression module in community practice. Journal of Clinical Psychiatry, 77, 221–227. Slatcher, R. B., & Selcuk, E. (2017). A social psychological perspective on the links between close relationships and health. Current Directions in Psychological Science, 26, 16– 21. https://doi.org/10.1177/09637 21416 667444 Snyder, D. K., Mangrum, L. F., & Wills, R. M. (1993). Predicting couples’ response to marital therapy: A comparison of short- and long-term predictors. Journal of Consulting and Clinical Psychology, 61(1), 61–69. https://doi.org/10.1037//0022-006x.61.1.61 Soleimani, A. A., Najafi, M., Ahmadi, K., Javidi, N., Kamkar, E. H., & Mahboubi, M. (2015). The effectiveness of emotionally focused couples therapy on sexual satisfaction and marital adjustment of infertile couples with marital conflicts. International Journal of Fertility and Sterility, 9(3), 393–402. https://doi.org/10.22074/ijfs.2015.4556 Soltani, A., Molazadeh, J., Mahmoodi, M., & Hosseini, S. (2013). A study on the effectiveness of Emotional Focused Couple Therapy on intimacy of couples. Procedia Social and Behavioral Sciences, 461–465. https://doi.org/10.1016/j.sbspro.2013.06.293 Soltani, M., Shairi, M. R., Roshan, R., & Rahimi, C. (2014). The impact of Emotionally Focused Therapy on emotional distress in infertile couples. International Journal of Fertility and Sterility, 7(4), 337–344. Son, H., Friedmann, E, & Thomas, S. A. (2012). Application of pattern mixture models to address missing data in longitudinal data analysis using SPSS. Nursing Research, 61(3), 195-203. https://doi.org/10.1097/NNR.0b013e3182541d8c Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family, 38(1), 15–28. https://doi.org/10.2307/350547 111 Spengler, P. M., Lee, N. A., Wiebe, S. A., & Wittenborn, A. K. (in press). A comprehensive meta-analysis on the efficacy of emotionally focused couple therapy. Couple and Family Psychology: Research and Practice. Sun, S.-H., & Liu, T. (2015). Hold Me Tight: An Emotionally Focused psychoeducational group for unmarried couples (情緒取向未婚伴侶心理教育團體). Journal of Educational Practice and Research, 28(2), 213–242. Sun, S.-H. (2019). The experiences of Emotionally Focused Couple Therapy for unmarried couples (愛情練習曲：未婚伴侶在情緒取向伴侶諮商的經驗研究). Research in Applied Psychology, 71, 137–182. https://doi.org/10.3966/156092512019120071005 Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). New York: Pearson. Taiwan Counseling Psychologist Union (2011). Suggestions for counseling session fees. https://www.tcpu.org.tw/about/electronic-communications.html Taiwan Association Against Depression (2020). Current Help Seeking Behaviors in Taiwan. http://www.depression.org.tw/knowledge/info.asp?/71.html Thompson, L., & Walker, A. J. (1989). Gender in families: Women and men in marriage, work, and parenthood. Journal of Marriage and the Family, 51(4), 845- 871. https://doi.org/10.2307/353201 Thorpe, K. E., Zwarenstein, M., Oxman, A. D., Treweek, S., Furberg, C. D., Altman, D. G., Tunis, S., Bergel, E., Harvey, I., Magid, D. J., & Chalkidou. (2009). A pragmatic- explanatory continuum indicator summary (PRECIS): a tool to help trial designers. Journal of Clinical Epidemiology, 62, 464-475. Tran, T. V. (2009). Developing Cross-Cultural Measurement. Oxford University Press. https://doi.org/10.1017/CBO9781107415324.004 Tsai, J. L., & Levenson, R. W. (1997). Cultural influences on emotional responding: Chinese American and European American dating couples during interpersonal conflict. Journal of Cross-Cultural Psychology, 28(5), 600–625. https://doi.org/10.1177/0022022197285006 Tsai, M. C., Peng, S. C., & Kuo, W. Ben. (2021). Singlehood and childlessness: an age-period- cohort analysis of changing attitudes toward family in Taiwan (2005-2020). Journal of Family Studies. https://doi.org/10.1080/13229400.2021.2004912 Tsuya, N. O., & Bumpass, L. L. (2004). Introduction. in N. O. Tsuya, & L. L. Bumpass (Eds.), Marriage, work and family life in comparative perspective: Japan, South Korea, and the United States (pp.1-18). University of Hawaii Press. 112 United Nations (2014). Country classification. https://www.un.org/en/development/desa/policy/wesp/wesp_current/2014wesp_country_cla ssification.pdf van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J. (1996). Attachment representations in mothers, fathers, adolescents, and clinical groups: A meta-analytic search for normative data. Journal of Consulting and Clinical Psychology, 64(1), 8–21. https://doi.org/10.1037/0022-006X.64.1.8 van IJzendoorn, M. H., & Sagi-Schwartz, A. (2008). Cross-cultural patterns of attachment: Universal and contextual dimensions. In J. Cassidy & P. R. Shaver (Eds.), Handbook of Attachment: Theory, research, and clinical applications (pp. 880–905). The Guilford Press. https://doi.org/10.4088/jcp.v61n1011c Walsh, S. B. (2002). Emotion-focused couples therapy as a treatment of somatoform disorders: An outcome study. Dissertation Abstracts International, 63, 1579. Wang, C. D. C., & Mallinckrodt, B. S. (2006). Differences between Taiwanese and U.S. cultural beliefs about ideal adult attachment. Journal of Counseling Psychology, 53(2), 192–204. https://doi.org/10.1037/0022-0167.53.2.192 Wang, W.-H., & Wang, C.-H. (2015). Exploring the attachment injury after experiencing Extramarital affair with EFT couple therapy (外遇風暴後－以情緒取向伴侶治療探究外 遇所引發的依附創傷). Guidance Quarterly, 51(2), 14–20. http://www.airitilibrary.com/Publication/Index/19966822-201506-201508280009- 201508280009-14-20 Wang, W., Bian, Q., Zhao, Y., Li, X., Wang, W., Du, J., Zhang, G., Zhou, Q., & Zhao, M. (2014). Reliability and validity of the Chinese version of the Patient Health Questionnaire (PHQ-9) in the general population. General Hospital Psychiatry, 36(5), 539–544. https://doi.org/10.1016/j.genhosppsych.2014.05.021 Wang, Y. T., & Yang, W. S. (2019). Changes and trends in family structure in Taiwan, 1990 to 2010. Journal of Family Issues, 40(14), 1896–1911. https://doi.org/10.1177/0192513X19863203 Wei, M., Russell, D. W., Mallinckrodt, B., & Zakalik, R. A. (2004). Cultural equivalence of adult attachment across four ethnic groups: Factor structure, structured means, and associations with negative mood. Journal of Counseling Psychology, 51(4), 408– 417. https://doi.org/10.1037/0022-0167.51.4.408 Weissman, M. M. (1987). Advances in psychiatric epidemiology: Rates and risks for major depression. American Journal of Public Health, 77(4), 445–451. https://doi.org/10.2105/AJPH.77.4.445 Wertz, R. T. (1995). Intention to treat: Once randomized, always analyzed. Clinical Aphasiology, 113 23, 57–74. http://aphasiology.pitt.edu/archive/00000188/ Whisman, M. A., & Bruce, M. L. (1999). Marital dissatisfaction and incidence of major depressive episode in a community sample. Journal of Abnormal Psychology, 108(4), 674– 678. https://doi.org/10.1037/0021-843X.108.4.674 Whisman, M. A., Uebelacker, L. A., & Weinstock, L. M. (2004). Psychopathology and marital satisfaction: The importance of evaluating both partners. Journal of Consulting and Clinical Psychology, 72(5), 830–838. https://doi.org/10.1037/0022-006X.72.5.830 Wiebe, S. A., & Johnson, S. M. (2016). A review of the research in Emotionally Focused Therapy for couples. Family Process, 55(3), 390–407. https://doi.org/10.1111/famp.12229 Wiebe, S. A., Johnson, S. M., Lafontaine, M. F., Burgess Moser, M., Dalgleish, T. L., & Tasca, G. A. (2017). Two- year follow- up outcomes in emotionally focused couple therapy: An investigation of relationship satisfaction and attachment trajectories. Journal of Marital and Family Therapy, 43(2), 227– 244. https://doi.org/10.1111/ jmft.12206 Wittenborn, A. K., Liu, T., Ridenour, T. A., Lachmar, E. M., Mitchell, E. A., & Seedall, R. B. (2019). Randomized controlled trial of emotionally focused couple therapy compared to treatment as usual for depression: Outcomes and mechanisms of change. Journal of Marital and Family Therapy, 45(3), 395–409. https://doi.org/10.1111/jmft.12350 Wood, N. D., Crane, D. R., Schaalje, G. B., & Law, D. D. (2005). What works for whom: A meta-analytic review of marital and couples therapy in reference to marital distress. American Journal of Family Therapy, 33(4), 273–287. https://doi.org/10.1080/01926180590962147 World Health Organization (2017). Depression and other common mental disorders: global health estimates. Geneva, Switzerland: WHO Document Production Services. World Health Organization. (2021). Depression. https://www.who.int/news-room/fact- sheets/detail/depression. Yang, K. S. (1996). The psychological transformation of the Chinese people as a result of societal modernization. In The handbook of Chinese psychology (pp. 479–498). Oxford University Press. Yang, K. S. (1998). Chinese responses to modernization: A psychological analysis. Asian Journal of Social Psychology, 1, 75–97. https://doi.org/10.1111/1467-839X.00006 Yu, Y. (2006). From new gender roles of Chinese immigrant wives. International Journal of Diversity in Organizations, Communities, and Nations, 5(4), 101-109. Zhang, Q., Hou, Z., Fraley, R. C., Hu, Y., Zhang, X., Zhang, J., & Guo, X. (2022). Validating the Experiences in Close Relationships-Relationship Structures Scale among Chinese children 114 and adolescents. Journal of Personality Assessment, 104(3), 347-358. Zeytinoglu-Saydam, S. (2018). Cross-culturally responsive training of emotionally focused couple therapy: International experiences. In S. Singh Poulsen & R. Allan (Eds.), Cross cultural responsiveness and systemic therapy: Personal and clinical narratives (pp. 53–67). Springer. Zheng, L., Luo, Y., & Chen, X. (2020). Different effects of attachment anxiety and attachment avoidance on depressive symptoms: A meta-analysis. Journal of Social and Personal Relationships, 37(12), 3028-3050. 115