How to measure ego-resiliency in the face of various life-changing crises: Measurement invariance, convergent and discriminant validity and reliability of the Polish version of the Revised Ego-Resiliency Scale (ER89-R12)

Introduction

In recent years, there has been a noticeable increase in interest in resilience and a surge in the number of studies devoted to this issue. Many of these studies have aimed to establish the definitional or conceptual meaning of resilience and to review the available research tools (Arat, 2014; Fletcher & Sarkar, 2013; Windle, Bennett & Noyes, 2011). This area of research is characterized by high complexity in terms of definitions, with variously-defined concepts used throughout the literature: ‘resilience’, ‘resiliency’, and the most semantically clarified concept, ‘ego-resiliency’. Definitions of the terms ‘resilience’ and ‘resiliency’ are most often built around two key constructs: adversity and positive adaptation (Fletcher & Sarkar, 2013). While the term ‘resilience’ refers to the dynamic process responsible for effective adaptation in unfavourable life circumstances (Bonanno, 2004; Luthar, Cicchetti & Becker, 2000), ‘resiliency’ refers to a relatively constant personality trait that facilitates coping with stress effectively (Bonanno & Mancini, 2012). This characteristic allows for flexible adaptation and the mobilization of resources, which increases tolerance to negative emotions (Ong et al., 2006). People with a high level of resilience are characterized by emotional stability, a tendency towards the positive over the negative, and a greater sense of their own efficacy (Tugade & Fredrickson, 2004). They tend to interpret unfavourable circumstances in terms of challenges and opportunities for new experiences, rather than burdens (Tugade & Fredrickson, 2004; Ogińska-Bulik, 2013). Above all, the features of resilience are revealed in the way a person adjusts following exposure to a highly stressful event. This is manifested through stable behavioural patterns and the ability to maintain balance (Bonanno, 2004). The key concepts of resilience are flexible adaptation and effective coping. A person who is resilient shows a significantly higher level of capacity for positive and resourceful adaptation to external and internal stressors (Klohnen, 1996; Alessandri et al., 2012). Therefore, he does not experience symptoms of post-traumatic stress due to the protective nature of his or her resilience resilience (Bonanno, 2004; Bonanno & Mancini, 2012). Moreover, resilient people return to their pre-crisis state more quickly, and have a better mental health (Lepore & Revenson, 2006).

The definitive and conceptual complexity of the analysed concepts, and thus the tools used to measure the studied variables, results mainly from the fact that there are three different issues involved in resilience (Luthar, 2015; Windle, Bennett & Noyes, 2011). Firstly, resilience is a process requiring specific resources or protective factors, which leads to a specific result. Secondly, resilience can be seen as the actual result of this process. In this case, researchers are interested in the resources—personal and family, environmental and, more broadly, cultural—that facilitate resilience. Thirdly, resilience is also the individual ability to return to balance (bounce back), cope with adversities, adapt flexibly, and so on (Windle, Bennett & Noyes, 2011). In the latter sense, the terms ‘resiliency’ or ‘ego-resiliency’ are often used, although some researchers use the term ‘resilience’ to mean the disposition, not the process or its result (Bonanno, 2004; Kowalczyk et al., 2015; Perrig-Chiello, Hutchison & Morselli, 2015; Quinney & Fouts, 2004). Interestingly, for Bonanno (2004), resilience is more an individual trait than a process.

Materials and Methods

Results

The MGCFA was used to assess the measurement invariance of the ER89-R12 Scale across the three groups. As a first step, we tested the factor structure of the scale within each group separately. This model adopted by Kołodziej-Zaleska & Przybyła-Basista (2018a), shown in Fig. 1, was consistent within two factors: Ego-Resiliency Optimal Regulation (ER-OR) and Ego-Resiliency Openness to Life Experience (ER-OL) proposed Alessandri et al. (2007; 2012).

Before comparing the groups, it is important to make sure that the structure provides good fit for all groups. Thus, the first step is to test whether the proposed two-factor model of ER fits the empirical data from each group. Three CFAs were conducted for sample A, B, and sample C separately. The results show an excellent model fit for each sample, indicating that the two-factor model of ego-resiliency is supported in all samples. The CFA model with the unconstrained factor loadings and intercepts is shown in Table 1.

The adequacy of these models can also be determined in relation to the parameter estimates. Standardized factor loadings for the two-factor CFA are included in Table 2.

As the results of the ER Scale are derived from three samples, before they could be compared, an MGCFA was performed to demonstrate the equivalence of the measurement in the three samples (Beaujean, 2014). In the first stage of the study, an analysis of the equivalence of the ER measurements was carried out in the group of parents of children with Down’s syndrome, in the group of women with breast cancer, and in the group of people after divorce. The purpose of these analyses was to show that the ER Scale has comparable psychometric properties, regardless of the type of crisis situation people are facing. We tested for measurement invariance (MGCFA) in order to cross-validate the two-factor model across the three samples; see Table 3 for the fit indices.

Four measurement invariance steps were considered: (1) configural, equivalence of model form; (2) metric (weak factorial), equivalence of factor loadings; (3) scalar (strong factorial), equivalence of item intercepts or thresholds; and (4) residual (strict or invariant uniqueness), equivalence of item residuals or unique variances (see: Chen, 2007; Putnick & Bornstein, 2016). As the excellent fit of the two-factor structure had been established independently for each group in the first step, it could be expected that configural invariance would be supported; indeed, this was confirmed by the fit indices, as shown in Table 3. Model 1 provides an excellent fit to the data, indicating that the factorial structure of the construct is equal across groups. As the configural invariance was supported, the factor pattern coefficients were then constrained to be equal to the test for metric invariance. Model 2 had good fit indices (see Table 3). The comparative fit indices (ΔCFI = 0.005) indicate that the viability of constraining the factor loading was the same across samples (ΔRMSEA is 0.026, but ΔSRMR is good, it equals 0.013). The scalar invariance model (Model 3) and the residual invariance model (Model 4) provide an excellent fit to the data. Support for scalar invariance indicates that the latent means can be meaningfully compared across samples. Support for residual invariance indicates that the sum of specific variance and error variance is similar across groups. The analysis above supports the measurement invariance of the two-factor model of ego-resiliency across the three groups.

The obtained results indicate the scalar equivalence of the measurements in both samples (the CFI value for the scalar equivalence model was 0.995 compared to the CFI value for the metric equivalence model, which was also 0.995; ΔSRMR was not greater than 0.01, the RMSEA difference for the compared models was 0.003). This result enables the conclusion that the ER Scale maintains comparable psychometric properties in studies of parents of children with Down’s syndrome, women with breast cancer and people after divorce. It is therefore possible to compare the mean results for all three groups.

The one-way ANOVA carried out at a later stage of the study for comparative analysis of the three groups shows statistically significant differences for the main ER score (F(3, 546) = 3.02; p < 0.05) and for Optimal Regulation (F(3, 546) = 3.29; p < 0.05). With an in-depth analysis including assessment between the three groups, the differences remained significant between the following groups: ER: sample A and sample B (95% CI [3.38; −0.008]); OR: sample A and sample B (95% CI <−2.41; −0.03>). The highest ER was found in the women with breast cancer (M = 35.26; SD = 6.78), followed by the divorced individuals (M = 35.09; SD = 6.71), and the lowest ER was found in parents of children with Down’s syndrome (M = 33.53; SD = 6.25). See Table 4 for details.

In the next stage of the analysis, the correlation coefficients between ER and the other tested variables were calculated in order to assess the construct validation. The Pearson correlation was calculated between the individual mean scores on the ER89-R12 and the some measures of mental health: psychological well-being, perceived social support, post-traumatic growth, personality traits like self-esteem, and perceived stress. As expected, higher levels of ego-resiliency were strongly associated with psychological well-being. The size of the coefficient was fairly high in each sample. Ego-resiliency showed a positive correlation with perceived social support, self-esteem, and post-traumatic growth. Furthermore, ego-resiliency was negatively correlated with perceived stress (Table 5).

Discussion

The aim of our work was to confirm the structure of the tool in groups of people facing various psychological crises. For this purpose, a measurement invariance analysis was used. Measurement invariance assesses the psychometric equivalence of a construct across groups or measures and shows whether the construct has the same meaning in these groups or repeated measures. The conducted measurement equivalence analysis, composed of several levels, showed that the representatives of the various groups we studied have the same understanding of the ego-resiliency construct. They assign the same weighting to the test items of the ER89-R12 Scale in their responses, meaning that the answers obtained in all three groups are comparable with one another. In our research four levels of equivalence were tested: configural, metric, scalar and residual invariance. Therefore, as the presented results of the analysis show, the ER89-R12 tool is an accurate measure of ego-resiliency in groups of people facing various crises. In other words, the latent construct is measured identically across groups and, as in the Italian version of the tool (Alessandri et al., 2007; Alessandri et al., 2012), it includes two subscales: optimal regulation and openness to life experience. The presented results of the analysis make it possible to compare the mean results on the ER89-R12 Scale and its subscales between the studied smples. The conducted comparative analysis showed that the level of ego-resiliency differed depending on the studied group. However, differences between groups were small referring to effect size. Despite the differences in the ego-resiliency levels between the groups, the analysis of the construct validity of the ER89-R12 in the three groups indicated the same patterns of dependencies between variables. The results show that ego-resiliency was positively correlated with measures of mental health: psychological well-being, post-traumatic growth, self-esteem and perceived social support, and negatively correlated with perceived stress in each of the analysed groups. The correlations are consistent with the predictions, theoretical assumptions and results of other similar analyses (Lepore & Revenson, 2006). The results confirm the conclusion that ego-resiliency is a construct related to the notions of adjustment, maladjustment and psychological health (Alessandri et al., 2012; Block & Block, 1980; Letzring, Block & Funder, 2005; Vecchione et al., 2010). As Block & Kremen (1996) noted, individuals with high ego-resiliency are better adapted and have the ability to deal more effectively with both changing everyday events and difficult situations than those with low ego-resiliency. Research shows that ego-resiliency is a key personal resource that might be able to buffer the detrimental effects of daily stressors on individuals’ negative emotional inertia (Alessnadri et al., 2020). Ego-resiliency explains the functioning of people who cope well with stress (Kaczmarek, 2011). Therefore, it can also be expected that ego-resiliency is related to those aspects of functioning that enable effective adaptation, coping, flexibility, and a committed and active approach to the world (Alessandri et al., 2007).

Ego-resiliency is a trait that plays a protective role in the adaptation process (Ilska, Brandt-Salmeri & Kołodziej-Zaleska, 2020; Kołodziej-Zaleska & Przybyła-Basista, 2020; Izydorczyk et al., 2018) Used in conjunction with the variables determining psychological well-being, support, coping with stress and trauma, and other personality traits which are protective factors in the process of adaptation to crises, the Ego-Resiliency Scale allows for a deeper understanding of how ego-resiliency works. Ego-resiliency can be considered a potential factor facilitating the process of adaptation (Shin et al., 2019). According to current knowledge, individuals with a high level of ego-resiliency find it easier to adapt to changing circumstances (Block & Kremen, 1996; Vecchione et al., 2010), which is important for people facing crises. Ego-resiliency can help people deal more quickly with loss, grief, guilt, sadness and other negative feelings, and focus on adaptive flexibility and coping effectively in life.

Research studies on resilience, understood as a personality trait, in life crisis situations such as post-divorce, illness or a child’s disability are not carried out frequently (Caples et al., 2018; Frisby et al., 2012; Izydorczyk et al., 2018; Perrig-Chiello, Hutchison & Morselli, 2015; Seiler & Jenewein, 2019; Van Riper, 2007). Bonanno, Westphal & Mancini (2011, 2012) discuss whether resiliency (understood as an individual’s disposition) plays a very important role in the process of adapting to various crisis events but rarely appears in research in this context. Klohnen (1996), meanwhile, notes that too little attention is paid to resilience, understood as a variable, in the field of individual differences (see also Ilska, Brandt-Salmeri & Kołodziej-Zaleska, 2020).

Most of the large-scale studies using the Ego-Resiliency Scale have been conducted with adolescents and young adults in the general population (Alessandri et al., 2014; Caprara et al., 2014; Block & Kremen, 1996; Letzring, Block & Funder, 2005; Vecchione et al., 2010). The analysis conducted here confirms that the tool can be used unreservedly in research with older or middle-aged people and, importantly, people who are struggling with various adversities. Due to the variety of crises experienced by the respondents (crisis related to a child’s illness, crisis related to losing one’s own health and struggling with a terminal illness, and family crisis related to the experience of divorce), it can also be assumed that the ER89-R12 will be an adequate measurement tool in the case of other life crises, allowing the results of people with various difficult life experiences to be compared.

An additional and interenting result of the analysis carried out is the difference in the level of ego-resiliency in the group of parents of children with Down’s syndrome. Compared to other groups the level of ego-resiliency was lower. Looking at the distinct components, the differences relate to optimal regulation, but they do not occur in the area of openness to life experience, which supports the process of accepting the state of affairs, taking action, and flexible focus on goals. Perhaps parents of children with Down’s syndrome have difficulty in adapting to the role of parents of a child with a mental disability. They experience a higher level of stress than parents of healthy children (Baker et al., 2003). In the light of the presented results, in the case of parents of children with Down’s syndrome, ego-resiliency is weaker than in the group of people facing a crisis related to their own terminal illness. Adequate modulation of emotion, flexible adaptation of responses, and readiness to face new and unexpected situations are of key importance in enduring cancer. Ego-resiliency promotes the mobilization of the patient to take various remedial actions; it may act as a resource that increases tolerance to the ambivalence and ambiguity that are part of the patient’s reality. For breast cancer patients, ego-resiliency has a regulatory function and may also facilitate the adoption of an open attitude towards the challenges accompanying the disease. This trait is a kind of meta-resource that optimizes the relationship between personality and environmental variables and remains superior to other resources. It provides flexibility as well as a greater ability to search for meaning in what is happening (Farkas & Orosz, 2015). It aids optimal adaptation, especially in the experience of somatic disease (Park, 2010).

Coping proactively with the effects of a crisis by mobilizing resources appears to be more likely, but also more effective, when the disease affects oneself rather than one’s child. It seems that a child’s disability could be a factor that disturbs the world of values, emotional balance and sense of meaning to a greater extent than a cancer diagnosis. Being the parent of a child with Down’s syndrome is undoubtedly a crisis of a permanent nature. In the case of women with breast cancer or divorced people, the crisis may be characterized by a narrow period of time, often limited to coping with the acute effects of the stressors experienced. Having a disabled child, on the other hand, is a crisis with an unlimited time spectrum. Moreover, it is a crisis the impact of which extends not only to the adult, but above all to the child. It can be assumed that this disturbs the coherent image of a valuable ‘I’, and the hope for favourable circumstances in the future, to a greater extent. The birth of a child with a disability usually induces in parents a sense of being trapped with no possibility of escape (Będkowska-Heine, 2007). According to researchers (Kaczmarek, Sęk & Ziarko, 2011), people with lower levels of ego-resiliency tend to focus on failures, have little emotional diversity and often suffer from a lack of meaning in life. Furthermore, the lower level of ego-resiliency of parents of children with Down’s syndrome may be related to higher levels of perceived stress and lower levels of psychological well-being (Kózka & Przybyła-Basista, 2017).

On the other hand, some research has provided support for the belief that many families of children with Down’s syndrome respond to challenges with resilience (Van Riper, 2007). According to certain studies, families of children with Down’s syndrome can adapt and become resilient; factors found to positively influence this process include family hardiness and affirming family communication (Caples et al., 2018). These studies, however, focus on family resilience, indicating the aspects of family life that enable adaptation to the circumstances of the child’s illness. They are much more about resilience as a process than a personality trait. They provide no information about the parents’ individual resources.

This study is an empirical attempt to examine the measurement invariance of the ego-resiliency measure across diverse groups of individuals experiencing life crises. Assessing the questionnaire’s capacity to measure the same construct in different contexts is important from theoretical and practical perspectives. Our findings suggest that the interpretation of items on the scale did not differ among individuals facing life crises. It means that individuals from the three groups interpret the items in an equivalent manner (Hussey & Hughes, 2020). The scores on the scale are likely to measure the same latent variable in a comparable way, regardless of the groups to which the individuals belong. Moreover, our findings show that ER89-R12 has essential properties: good test-retest reliability and clear factor structure in line with the previous validation studies on the ER89-R (Alessandri et al., 2007; Alessandri et al., 2012; Vecchione et al., 2010). Our results can be helpful for researchers investigating the predictors of resilience during the adaptation process after the crisis. Confidence in the comparability and replicability of research findings is a fundamental pillar for progress in research (Hussey & Hughes, 2020). Our analyzes also showed that further studies on the invariant and replicability of the scale structure in various populations, both general and clinical populations, are necessary. Our findings support the validity of the two-factor model of ego-resiliency construct among individuals facing a crisis. We recommend using the ER89-R-12 in research projects because of its good psychometric properties, easy administration, and applicability in the life-crises context.

Limitations

There are several limitations of the presented study. First, women predominated in all studied samples. It would be helpful in future research to include more men to evaluate the validity of the ER89-R12 Scale and measure invariance across gender. Second, due to the convenient samples used in the study, the generalization of these findings should be conducted with caution. Third, in our research, we did not always use the same instrument to measure psychological well-being and related to its constructs. So the Oxford Happiness Questionnaire (OHQ) was administered in two samples (among divorced people within sample A and parents of children with Down syndrome in sample C). Still, the Meaning in Life Questionnaire was distributed only in the study of women with breast cancer, sample B. We are aware that the hedonistic view of happiness or subjective view of well-being (e.g., Diener, 2000) differs from the eudaimonic view of well-being, emphasizing personal potential, virtue, and meaningful life (Ryff & Singer, 1998). Thus, the measurement tools of the OHQ (Hills & Argyle, 2002) and MLQ (Steger, Oishi & Kashdan, 2009) are not the same. However, in the broadest sense, the term “well-being” refers to people’s optimal functioning and positive experiences (Steger & Samman, 2012) and includes both a sense of subjective happiness and life meaning (Lopez, Pedrotti & Snyder, 2015). Therefore, we use the term well-being when describing the results obtained in these questionnaires.

Our study was part of a larger project on the determinants of adaptation effectiveness in the face of critical life events. One of the research tasks was to test the role of ego-resiliency. In this paper, we present only the part of our study that deals with validating the ER89-R12 measurement tool and its invariance. Ego-resiliency was measured with the ER89-R12 questionnaire in all samples. But the instruments measuring stress, perceived social support or growth after trauma/ crisis were already somewhat more differentiated and tailored to the needs of the study in the specific groups. It seems important that the correlations between ego-resiliency and the analysed constructs indicate the similarity of the patterns of dependencies between the distinct groups. Therefore, future studies need to be repeated using the same measurement tools. There are also limitations on the intergroup comparisons due to the group sizes being large enough that detected differences may be treated rather little in size.

Supplemental Information