Family caregivers of older adults often suffer from fatigue and burnout (Day and Anderson, Reference Day and Anderson2011). Recent studies have found that self-compassion improved caregivers’ mental health and promoted their well-being (Delaney, Reference Delaney2018; Raab, Reference Raab2014). By contrast, with a lack of self-compassion, family caregivers ignore their own needs for a long time, and their capacity to provide compassionate care to their loved ones decreases over time. This is known as compassion fatigue (Day and Anderson, Reference Day and Anderson2011). Therefore, before developing a compassion program for caregivers to improve their well-being and reduce compassion fatigue, it is necessary to develop a good psychometric compassion scale to measure the effectiveness of compassion programs.
Murfield et al. (Reference Murfield, Moyle, O’Donovan and Ware2021) explore the validity of the compassionate engagement and action scales (CEAS) with family carers of older adults. These three scales were developed by Gilbert et al. (Reference Gilbert2017) to measure the competencies of an individual engaging in suffering and taking action to alleviate and prevent suffering. The items of the compassionate engagement scale represent key attributes: care for well-being, sensitivity, distress tolerance, sympathy, empathy, and non-judgment. The items of the compassionate action scale describe transformative skills: attention (mindfulness practice), reasoning (cognitive reappraisal), behavior (gratitude exercises), feeling (showing emotions), imagery (guided imagery exercises), and sensory (breathing practices, posture). These attributes reflect the five elements of compassion proposed in Strauss et al.’s (Reference Strauss2016) review study, namely, recognizing suffering, understanding the university of human suffering, being emotionally connected with the suffering, tolerating uncomfortable feelings, and acting or feeling motivated to alleviate suffering. Developing compassion intervention may include these skills. CEAS can examine the impact of such actions on improving competency in compassionate engagement and action.
The three CEAS scales are based on Gilbert’s (Reference Gilbert2014) hypothesis of the three-way flow of compassion for self, compassion for others, and compassion from others, which are involved in emotional regulation. People experience emotional distress when their thinking brain is hijacked by the emotional limbic system, and, subsequently, they move into the primitive drives of fight or flight (threat system). Cultivating and developing individuals’ competences in self-compassion, receiving compassion from others, and showing compassion to others can enhance their emotional soothing system to regulate the threat system (Gilbert, Reference Gilbert2014). Currently, the most widely used compassion measure is Neff’s (Reference Neff2003) Self-Compassion Scale. Gilbert’s three-directional flows of compassion have not been examined well. Murfield et al.’s validation of these three scales helps future researchers using these scales to examine hypotheses on the abilities to be compassionate toward oneself, to feel compassion for others, and to receive compassion from others in emotional regulation studies.
Murfield et al. use an international sample of 171 family caregivers to validate the three compassion scales through confirmatory factor analysis (CFA). The strength of CFA is that it can confirm the construct validity of a measure through verifying the underlying factors and the pattern of item-factor relationships (i.e. factor loadings) (Brown, Reference Brown2006). In this study, the results of the CFA support a single higher-order factor; that is, the factor structure of the CEAS measures involves a single, broader construct representing the concept of compassion and two dimensions representing engagement and action that support Gilbert’s compassion theory (Gilbert et al., Reference Gilbert2017). However, the use of a single higher-order factor model could not demonstrate the discriminant validity of the subscales. Discriminant validity is supported when the relationship between distinct factors (engagement and action) is small to moderate (Brown, Reference Brown2006). To develop and validate higher-order constructs, the following procedure is useful: (1) confirm a conceptually valid first-order CFA model; (2) report a strong correlation among first-order factors; and (3) examine an alternative or competing models (Brown, Reference Brown2006; Koufteros et al., Reference Koufteros, Babbar and Kaighobadi2009). It is also important to identify the preferable model based on the appropriate fit statistics (i.e. chi-square difference test for nested models), parsimony, and relevant theory (Cabrera-Nguyen, Reference Cabrera-Nguyen2010).
Consistent with Gilbert et al. (Reference Gilbert2017), in Murfield et al.’s study, CFA showed a more complicated higher-order model for the Self-Compassion Scale. The construct of self-compassion is conceptualized as two dimensions of engagement and action. Moreover, the dimension of engagement consists of two subdimensions, namely, sensitivity to suffering and engagement with suffering. However, their small sample size (<200) may limit the analysis of a complex model. More theoretical and empirical justification and a sufficient sample size are also necessary here.
In Murfield et al.’s study, Item 5 in the CFA model of compassion for self had a comparatively low factor loading (<0.40). This item assesses competency in distress tolerance, which is a key element of compassion (Strauss et al., Reference Strauss2016). Similarly, Gu et al. (Reference Gu, Cavanagh, Baer and Strauss2017) conducted a validity study on the five-element compassion test proposed by Strauss et al. (Reference Strauss2016) using a combination of existing compassion measures. They also found that the factor of “tolerating uncomfortable feelings” may be problematic as a subcomponent of compassion. However, compassion theory (Gilbert et al., Reference Gilbert2017) proposes that individuals who lack distress tolerance are more likely to become frightened or overwhelmed by their own or others’ suffering as a result of engaging suffering by avoidance or dissociation. In addition to Item 5, Item 4, which assesses being emotionally moved (sympathy), had a low loading (<0.40) on the engagement factor of self-compassion. Further analysis, such as an item analysis and an exploratory factor analysis, may be useful to find an internally consistent set of items.
According to Gilbert’s compassion theory (Gilbert et al., Reference Gilbert2017), the three directional flows of compassion have moderate associations with each other. Gilbert et al. (Reference Gilbert, McEwan, Catarino and Baião2014) found that fear of receiving compassion from others had a strong correlation with fear of being compassionate to oneself, but a weaker correlation with fear of giving compassion to others among depressed patients. However, Lopez et al. (Reference Lopez, Sanderman, Ranchor and Schroevers2018) revealed that in a community sample, compassion for self and compassion for others did not have a significant relationship. Muirfield et al. did not examine the correlations among the three flows of compassion. Establishing the validity of three CEAS measures will help to explore the associations between these different flows of compassion further.
In the CEAS original validation research (Gilbert et al., Reference Gilbert2017), self-compassion and compassion from others had significant correlations with depression, anxiety, stress, and psychological well-being, while compassion for others only had a relationship with well-being. Other researchers found associations between self-compassion and lower levels of depressive symptoms and higher levels of positive affect, while compassion for others had no significant relationship with psychological well-being (Lopez et al., Reference Lopez, Sanderman, Ranchor and Schroevers2018). In the transactional model of stress and coping, self-compassion is a coping resource for people to recover from stressful events (Allen and Leary, Reference Allen and Leary2010). A study of 72 patient-caregiver dyads found that self-compassion serves as a protective factor to mitigate the relationship between caregiving stress and depressive symptoms (Hsieh et al., Reference Hsieh, Yu, Chen, Chen, Chang and Hsiao2019). Compassionate caregivers are more likely to use positive cognitive restructuring and less likely to use avoidance coping strategies, thereby promoting psychological adjustment (Allen and Leary, Reference Allen and Leary2010). In addition to self-compassion, being open to receive compassion from others could buffer against the impact of self-criticism on depression (Hermanto et al., Reference Hermanto2016). According to these findings, future studies can explore the role of the three flows of compassion as a coping resource in the transactional model of stress and coping and their correlations with psychological distress and well-being in clinical and non-clinical samples.
Gilbert’s CEAS was originally developed using the general population. Murfield et al.’s study validated the CEAS measures among family caregivers of older adults. Because Gilbert’s CEAS of self-compassion, compassion for others, and compassion from others scales are still new, future validation research is necessary in more diverse populations (e.g. professional healthcare providers, clinical populations). Furthermore, the original validation research found gender differences on compassion (Gilbert et al., Reference Gilbert2017). The results reveal that women report higher scores on the compassion for others scale than men. Lopez et al. (Reference Lopez, Sanderman, Ranchor and Schroevers2018) reported that women and lower-educated individuals consistently have more compassion for others than their counterparts. Given that the Murfield et al.’s study sampled mainly white females (88%), future research should assess measurement invariance by gender. Moreover, as noted, although Murfield et al.’s study was conducted with an international sample, it contained no further discussion on the variations among ethnic groups. Therefore, the CEAS measures would benefit from cross validation in different populations and cultures.
Murfield et al.’s study presents important results in measuring and understanding the theoretical constructs of compassion measures. The measures of CEAS capturing the three orientations of compassion have made a significant contribution to the field of compassion research and practice. It is important that future studies continue to validate this scale.
You have
Access
Family caregivers of older adults often suffer from fatigue and burnout (Day and Anderson, Reference Day and Anderson2011). Recent studies have found that self-compassion improved caregivers’ mental health and promoted their well-being (Delaney, Reference Delaney2018; Raab, Reference Raab2014). By contrast, with a lack of self-compassion, family caregivers ignore their own needs for a long time, and their capacity to provide compassionate care to their loved ones decreases over time. This is known as compassion fatigue (Day and Anderson, Reference Day and Anderson2011). Therefore, before developing a compassion program for caregivers to improve their well-being and reduce compassion fatigue, it is necessary to develop a good psychometric compassion scale to measure the effectiveness of compassion programs.
Murfield et al. (Reference Murfield, Moyle, O’Donovan and Ware2021) explore the validity of the compassionate engagement and action scales (CEAS) with family carers of older adults. These three scales were developed by Gilbert et al. (Reference Gilbert2017) to measure the competencies of an individual engaging in suffering and taking action to alleviate and prevent suffering. The items of the compassionate engagement scale represent key attributes: care for well-being, sensitivity, distress tolerance, sympathy, empathy, and non-judgment. The items of the compassionate action scale describe transformative skills: attention (mindfulness practice), reasoning (cognitive reappraisal), behavior (gratitude exercises), feeling (showing emotions), imagery (guided imagery exercises), and sensory (breathing practices, posture). These attributes reflect the five elements of compassion proposed in Strauss et al.’s (Reference Strauss2016) review study, namely, recognizing suffering, understanding the university of human suffering, being emotionally connected with the suffering, tolerating uncomfortable feelings, and acting or feeling motivated to alleviate suffering. Developing compassion intervention may include these skills. CEAS can examine the impact of such actions on improving competency in compassionate engagement and action.
The three CEAS scales are based on Gilbert’s (Reference Gilbert2014) hypothesis of the three-way flow of compassion for self, compassion for others, and compassion from others, which are involved in emotional regulation. People experience emotional distress when their thinking brain is hijacked by the emotional limbic system, and, subsequently, they move into the primitive drives of fight or flight (threat system). Cultivating and developing individuals’ competences in self-compassion, receiving compassion from others, and showing compassion to others can enhance their emotional soothing system to regulate the threat system (Gilbert, Reference Gilbert2014). Currently, the most widely used compassion measure is Neff’s (Reference Neff2003) Self-Compassion Scale. Gilbert’s three-directional flows of compassion have not been examined well. Murfield et al.’s validation of these three scales helps future researchers using these scales to examine hypotheses on the abilities to be compassionate toward oneself, to feel compassion for others, and to receive compassion from others in emotional regulation studies.
Murfield et al. use an international sample of 171 family caregivers to validate the three compassion scales through confirmatory factor analysis (CFA). The strength of CFA is that it can confirm the construct validity of a measure through verifying the underlying factors and the pattern of item-factor relationships (i.e. factor loadings) (Brown, Reference Brown2006). In this study, the results of the CFA support a single higher-order factor; that is, the factor structure of the CEAS measures involves a single, broader construct representing the concept of compassion and two dimensions representing engagement and action that support Gilbert’s compassion theory (Gilbert et al., Reference Gilbert2017). However, the use of a single higher-order factor model could not demonstrate the discriminant validity of the subscales. Discriminant validity is supported when the relationship between distinct factors (engagement and action) is small to moderate (Brown, Reference Brown2006). To develop and validate higher-order constructs, the following procedure is useful: (1) confirm a conceptually valid first-order CFA model; (2) report a strong correlation among first-order factors; and (3) examine an alternative or competing models (Brown, Reference Brown2006; Koufteros et al., Reference Koufteros, Babbar and Kaighobadi2009). It is also important to identify the preferable model based on the appropriate fit statistics (i.e. chi-square difference test for nested models), parsimony, and relevant theory (Cabrera-Nguyen, Reference Cabrera-Nguyen2010).
Consistent with Gilbert et al. (Reference Gilbert2017), in Murfield et al.’s study, CFA showed a more complicated higher-order model for the Self-Compassion Scale. The construct of self-compassion is conceptualized as two dimensions of engagement and action. Moreover, the dimension of engagement consists of two subdimensions, namely, sensitivity to suffering and engagement with suffering. However, their small sample size (<200) may limit the analysis of a complex model. More theoretical and empirical justification and a sufficient sample size are also necessary here.
In Murfield et al.’s study, Item 5 in the CFA model of compassion for self had a comparatively low factor loading (<0.40). This item assesses competency in distress tolerance, which is a key element of compassion (Strauss et al., Reference Strauss2016). Similarly, Gu et al. (Reference Gu, Cavanagh, Baer and Strauss2017) conducted a validity study on the five-element compassion test proposed by Strauss et al. (Reference Strauss2016) using a combination of existing compassion measures. They also found that the factor of “tolerating uncomfortable feelings” may be problematic as a subcomponent of compassion. However, compassion theory (Gilbert et al., Reference Gilbert2017) proposes that individuals who lack distress tolerance are more likely to become frightened or overwhelmed by their own or others’ suffering as a result of engaging suffering by avoidance or dissociation. In addition to Item 5, Item 4, which assesses being emotionally moved (sympathy), had a low loading (<0.40) on the engagement factor of self-compassion. Further analysis, such as an item analysis and an exploratory factor analysis, may be useful to find an internally consistent set of items.
According to Gilbert’s compassion theory (Gilbert et al., Reference Gilbert2017), the three directional flows of compassion have moderate associations with each other. Gilbert et al. (Reference Gilbert, McEwan, Catarino and Baião2014) found that fear of receiving compassion from others had a strong correlation with fear of being compassionate to oneself, but a weaker correlation with fear of giving compassion to others among depressed patients. However, Lopez et al. (Reference Lopez, Sanderman, Ranchor and Schroevers2018) revealed that in a community sample, compassion for self and compassion for others did not have a significant relationship. Muirfield et al. did not examine the correlations among the three flows of compassion. Establishing the validity of three CEAS measures will help to explore the associations between these different flows of compassion further.
In the CEAS original validation research (Gilbert et al., Reference Gilbert2017), self-compassion and compassion from others had significant correlations with depression, anxiety, stress, and psychological well-being, while compassion for others only had a relationship with well-being. Other researchers found associations between self-compassion and lower levels of depressive symptoms and higher levels of positive affect, while compassion for others had no significant relationship with psychological well-being (Lopez et al., Reference Lopez, Sanderman, Ranchor and Schroevers2018). In the transactional model of stress and coping, self-compassion is a coping resource for people to recover from stressful events (Allen and Leary, Reference Allen and Leary2010). A study of 72 patient-caregiver dyads found that self-compassion serves as a protective factor to mitigate the relationship between caregiving stress and depressive symptoms (Hsieh et al., Reference Hsieh, Yu, Chen, Chen, Chang and Hsiao2019). Compassionate caregivers are more likely to use positive cognitive restructuring and less likely to use avoidance coping strategies, thereby promoting psychological adjustment (Allen and Leary, Reference Allen and Leary2010). In addition to self-compassion, being open to receive compassion from others could buffer against the impact of self-criticism on depression (Hermanto et al., Reference Hermanto2016). According to these findings, future studies can explore the role of the three flows of compassion as a coping resource in the transactional model of stress and coping and their correlations with psychological distress and well-being in clinical and non-clinical samples.
Gilbert’s CEAS was originally developed using the general population. Murfield et al.’s study validated the CEAS measures among family caregivers of older adults. Because Gilbert’s CEAS of self-compassion, compassion for others, and compassion from others scales are still new, future validation research is necessary in more diverse populations (e.g. professional healthcare providers, clinical populations). Furthermore, the original validation research found gender differences on compassion (Gilbert et al., Reference Gilbert2017). The results reveal that women report higher scores on the compassion for others scale than men. Lopez et al. (Reference Lopez, Sanderman, Ranchor and Schroevers2018) reported that women and lower-educated individuals consistently have more compassion for others than their counterparts. Given that the Murfield et al.’s study sampled mainly white females (88%), future research should assess measurement invariance by gender. Moreover, as noted, although Murfield et al.’s study was conducted with an international sample, it contained no further discussion on the variations among ethnic groups. Therefore, the CEAS measures would benefit from cross validation in different populations and cultures.
Murfield et al.’s study presents important results in measuring and understanding the theoretical constructs of compassion measures. The measures of CEAS capturing the three orientations of compassion have made a significant contribution to the field of compassion research and practice. It is important that future studies continue to validate this scale.
Conflict of interest
None.