Positive psychology emphasizes the importance of virtues and an individual’s signature  strengths, such as hope and gratitude (Lopez & Snyder, 2009; Peterson & Seligman, 2004). In contrast, existential positive psychology (Wong, 2011, 2019, 2021; Wong et al., in press) emphasis universal core values essential for survival and thriving, especially during difficult times, such as courage, compassion (Chen & Wu, 2021; Wong, 2020), and responsibility (Wong, 2008; 2016; 2019b).

In the era of COVID-19, the imperative of responsibility is thrust into the forefront of social consciousness. The public health policy of social distancing, wearing masks, and mass vaccination will not work for a determined minority who consider such government imposed mandate as an infringement against their personal liberty (Austen, 2021); such opposition may undermine scientists’ effort to end the spread of coronavirus (Ball, 2020). This is a serious public health issue, yet our proposal on the importance of responsibility during the pandemic (Wong et al., 2020) was rejected by a special issue on public psychology.

The positive psychology research on meaning also downplay the role of meaning. For example, George & Park, (2016) and Martela & Steger (2016) declared that a consensus had been reached on a tripartite definition of meaning in life, which excluded that responsibility as a factor, despite the vast literature on responsibility as important for meaningful living (e.g., Frankl, 1946/1985; Peck, 1978/2012; Peterson, 2018; Wong, 2008, 2019a). Even common sense would dictate that it is impossible for anyone to live a meaningful life in a free society with almost an infinite number of pathways, unless that person assumes the responsibility to pursue one’s ideal rather than letting others decide on one’s future. In fact, Wong (2019b) considers responsibility the key to living a meaningful and productive life (see Table 1).

Table 1

Wong’s (2019b) Responsibility Manifesto

Both research and clinical evidence support his view (e.g., Wong, 2010, 2012). The corollary is also true – the lack of responsibility may lead to addition (e.g., Rise & Halkjelsvik, 2019; Wong et al., 2013) and crimes (Trivedi-Bateman, 2020). Given the important role of responsibility in wellbeing and health, the main aim of the present paper is to validate the responsibility measure based on the Responsibility Checklist originally published in Wong (2019b).

Conceptualization of Responsibility

Specifically, responsibility is a broad construct characterized in different ways (e.g., moral, social, or personal responsibility; Sheldon et al., 2018) and is conceptualized as “feeling accountable for one’s decisions and actions, reliable and dependable to others, and empowered to act on issues within one’s control.” (Wray-Lake & Syvertsen, 2011, pp. 12). However, from the perspective of existential positive psychology (Wong, 2019a), responsibility is inextricably linked to the phenomenon of freedom and living an authentic life. From the existential perspective (Guignon, 1998; Jean-Paul Sartre 1973; Moeinikorbekandi, 2013; Wong, 2017), responsibility is predicated on individual freedom. We can transcend our genes, environment, and history only through the responsible exercise of freedom of choice.

According to Frankl (1946/1985), responsibility is conceptualized as “responsibleness” to the demands of life, and essential for living a meaningful life. Without responsibility, society will degenerate into chaos. To Viktor Frankl (1946/1985) and Wong (2012), responsibility equals meaning. Following Frankl, Wong defines meaning in terms of Purpose, Understanding, Responsibility, and Enjoyment (PURE). Frankl’s genius was identifying responsibility as an important half of the freedom-responsibility pair, essential for launching the quest for meaning and self-transcendence.

As the first tenet of logotherapy, freedom-of-will is a necessary step for a meaningful life; it hypothesizes that people who believe in the inherent human capacity for freedom and responsibility, regardless of circumstances, will show higher autonomy, authenticity, and live more meaningful lives than those without such beliefs (Wong, 2014). Thus, apart from a deep sense of responsibility, all the activities designed to achieve meaning in life would not survive the inescapable trials and temptations in human life.

Although responsibility is conceptualized in many different ways, researchers have commonly focused on the personal and social aspects of it (e.g., Giacalone et al., 2008; Sheldon et al., 2018; Wong, 2019b; Wray-Lake & Syvertsen, 2011). Personal responsibility means that an individual is accountable to oneself and the needs or wellbeing of others (Ruyter, 2002), and this construct emphasizes self-accountability, representing an individual’s behaviors and choices that can impact oneself and others (Mergler & Shield, 2016). In addition to personal responsibility, social responsibility is related to values that promote moral, prosocial, and civic behaviors of individuals (Wray-Lake & Syvertsen, 2011) and involves decisions and actions that serve to benefit others and society (Martins et al., 2015; Wong, 2019b).

 Individuals with high social responsibility are more likely to help and respect others and contribute to society (Mergler, 2017; Wray-Lake & Syvertsen, 2011). Wray-Lake and Syvertsen (2011) have proposed that social responsibility comprises of two key constructs including the social relationships and a moral sense of care and justice, and this structure refers to be a value-orientation that promotes an individual’s positive behaviors (e.g., moral, prosocial), which serve to benefit others and contribute to society (Wray-Lake & Syvertsen, 2011). Given these conceptualizations, social and personal responsibility can be characterized by separate, but related, aspects of overall responsibility (Mergler, 2017).

The literature has also suggested that responsibility is a powerful resource for healthy development and the wellbeing of individuals (Martins et al., 2015, 2017; Wong, 2019b; Wong & Wong, 2012). Specifically, previous studies have focused on the importance of personal responsibility in the education field and empathized that the sense of responsibility plays a vital role in promoting positive youth development and wellbeing (Martins et al., 2015; Mergler, 2017; Wray-Lake & Syvertsen, 2011; Zimmerman & Kitsantas, 2005). Responsibility promotes youth’s healthy functioning and skills and promotes positive behaviors; thereby, it provides a successful transition from adolescence to adulthood (Martins et al., 2017).

According to Frankl (1946/1985) and Wong (2012), responsibility equals meaning, and following Frankl, Wong defines meaning in terms of the Purpose, Understanding, Responsibility, and Enjoyment (PURE) model. According to the PURE model that has been developed with the context of the meaning–management theory (Wong, 2016; Wong & Wong, 2012), adaptive choices and decisions do not only provide satisfying and successful outcomes for the individuals but also contribute to others’ wellbeing and meet moral requirements. Specifically, responsibility is identified as a key construct of flourishing, resilience, and healthy functioning of individuals in this framework (Wong, 2010; Wong & Wong, 2012).

Responsible action refers to the behavioral component of the model, which includes doing what is right, finding the right solutions, making mends, appropriate reactions and actions, and taking actions that are congruent with the highest values (Wong, 2010). Wong (2010) has also emphasized the importance of responsible actions to live a purposefully and meaningfully life, and he has argued that responsibility is an important resource to thriving and surviving of individuals and society in difficult times (Wong, 2019b).

In addition, given the self-determination theory (Deci & Ryan, 2000), individuals with high self-determination feel responsible for their decisions and behaviors. Autonomous functioning helps individuals to make better decisions and choices (Ryan et al., 2012). This framework instantiates the concept of autonomy within the self-regulation approach and proposes that individuals who use self-regulated strategies are more likely to feel themselves responsible for their own choices and behaviors (Ryan et al., 2012). Self-determination may thus provide important implications for understanding of the development of responsibility constructs. The literature has theoretically and empirically suggested that responsibility is one of the keys to wellbeing and the good life, and the ability to make adaptive choices and decisions is paramount. Considering the important of this construct for individual and society, it is critical to develop a new effective scale for assessing of responsibility construct. Although many questionnaires are available to measure the sense of responsibility, most of them have  been developed and validated for use with students (for the review, see Mergler, 2017). Moreover, these measures have commonly been used to assess personal responsibility, and the social aspect of responsibility remains under-researched. Compared to the available instruments, the development of a valid and reliable screener for both personal and social responsibility can provide efficient measurement for researchers and practitioners. The purpose of the current study is to investigate the initial development and validation of the Responsibility Scale (RS) to measure the sense of responsibility of individuals.

Method

Participants

Participants included two independent samples, comprising of 284 adults, ranging in age between 18 and 84 years. Sample 1 was used to conducted the exploratory factor analysis and comprised of 152 adults (65% female), ranging in age from 18 to 82 years (M = 43.18, SD = 14.68). The participants reported their socioeconomic status (SES) reported as follow: <10000$= 22%, 10000–30000$= 25%, 30000–40000$= 15%, and 40000$ and upper = 18%. Sample 2 was used to conducted the confirmatory factor analysis. The sample consisted of 132 adults (56% female), ranging in age from 18 to 84 years (M = 29.08, SD = 12.45). The participants reported their socioeconomic status (SES) reported as follow: <10000$= 35%, 10000–30000$= 24%, 30000–40000$= 7%, 40000$ and upper = 11%, and prefer not to say= 23%. An online survey was created using Google Forms and presented to the participants who volunteered to participate in the study.

Measures

Responsibility Measure (RM). The scale item pool was generated by recommendations from literature on scale development (DeVellis, 2017; Tay & Jebb, 2017; Worthington & Whittaker, 2006). After outlining a rationale for developing a new scale for measuring responsibility, a two-step process was performed to create the pilot items of the RM. Literature and existing measures that have used to assess the personal and social responsibility were first reviewed to produce pilot items for the measure (Mergler, 2017; Mergler & Shield, 2016; Sheldon et al., 2018; Wong, 2019b). The scale’s items were created to reflect the hypothesized responsibility dimensions “responsibility to self” and “responsibility to others and society” as they are represented in the literature. After reviewing the theoretical and empirical literature, a total of 24 pilot items were drafted, and two experts, who were tenured professors in the fields of clinical psychology, were invited to review the RM item pool. Following the expert feedback, the item pool was then edited into a revised form, and the final draft was created containing 17 items.  All items were rated on a 5-point Likert-type scale (1= strongly disagree, 2= disagree, 3= undecided, 4= agree, and 5= strongly agree).

Personal Responsibility Scale (PRS). The PRS is an 8–item self-report measure developed to assess the personal responsibility of individuals (Sheldon et al., 2018). All items of the scale are scored using a 5-point Likert scale, ranging from 1= strongly disagree to 5= strongly agree (e.g., ‘‘I can almost always be counted on to be on time and to keep my promises”). Previous research provided evidence supporting that the scale had adequate internal reliability estimates (Sheldon et al., 2018). Similarly, the internal reliability estimate of the scale with the present sample was adequate, see Table 2.

Satisfaction with Life Scale (SWLS). Well-being was measured using the SWLS, which is a 5–item self-report measure developed to assess individuals’ cognitive judgments of life (Diener et al., 1985). The scale’s items scored based on a 7-point Likert scale, ranging from 1= strongly disagree to 7= strongly agree (e.g., “In most ways my life is close to my ideal”). Research supported that the scale had the internal and test-retest reliability estimates (Pavot & Diener, 1993). In the present study, the internal reliability estimate of the measure was strong, see Table 2.

Patient Health Questionnaire (PHQ). Psychological distress was assessed using the PHQ–4 designed to measure the depression and anxiety of individuals (Kroenke et al., 2009). All items of the PHQ are scored using a 3-point Likert scale, ranging between 0= not at all and 3= nearly every day (e.g., “Feeling nervous, anxious, or on edge”). Past research supported that the scale was a brief and effective measure to assess psychological distress and had strong internal reliability estimates (Kroenke et al., 2009). Findings from this study showed that the internal reliability estimate of the scale was strong, see Table 2.

Conscientiousness Scale. Conscientiousness was measured using the items from the Big Five Inventory (BFI) (John & Srivastava, 1999). Conscientiousness includes nine items scoring based on a 5–point Likert scale, ranging from 1= disagree strongly to 5= agree strongly (e.g., “Is a reliable worker”, “Tends to be lazy”). Research provided evidence supporting that the scale had strong internal reliability estimates (John & Srivastava, 1999). Findings from this study also showed that the scale had strong internal reliability estimate, see Table 2.

Table 2

Bivariate Correlations between the RM and Concurrent Validity Scales

Note. ^**p< .001. SS= Self-Responsibility Scale, SRS= Social Responsibility Scale, ORM= Overall Responsibility Scale.

Data Analyses

All statistical analyses were performed with SPSS version 25 and AMOS version 24. Exploratory factor analysis (EFA) was first conduct to examine the factor structure of the measure using the principal axis factoring method with a promax rotation with the first sample of the study. Prior to examining the EFA output, the number of indices were checked including the Kaiser-Meyer-Olkin (KMO> .50), Bartlett’s Test of Sphericity (<.05) the matrix determinant (> 0), and item communalities (<.30) (Field, 2009; Tabachnick & Fidell, 2013). After checking these criteria, the EFA output were interpreted using factor loading ≥ .40, cross-loading ≥ .32, and theoretical interpretability (Field, 2009; Stevens, 2009; Tabachnick & Fidell, 2013). As a second step, confirmatory factor analysis (CFA) was conducted to test the latent structure of the measure with the second sample of the study. The CFA results were interpreted using data-model fit statistics and their cut-of points: SRMR (standardized root mean square residual) and RMSEA (root mean square error of approximation) scores between .05 and .08 indicated adequate data–model fit, while values < .05 were accepted good data–model fit. CFI (comparative fit index) and TLI (Tucker Lewis index) scores > .90 were accepted as adequate and those exceeding .95 were interpreted as good data–model fit (Hooper et al., 2008; Kline, 2015; Mueller & Hancock, 2008). Additionally, the latent construct reliability (H) estimate was examined for the RM latent structures, and the estimate ≥ .70 indicated adequate (Mueller & Hancock, 2008). In the final phases of the analyses, concurrent validity was conducted. Bivariate correlation was first performed to explore the relationship between the observed RM scales and criterion variables. Findings from this analysis were interpreted using traditional standards: .10–.29 = small, .30–.49 = moderate, and ≥ .50 = large. Subsequently, a pair of latent variable path analyses (LVPA) was conducted to examine the predictive effect of the second and first-order measurement model on concurrent variables with the second sample of the study. Results of the LVPA were interpreted using the standardized regression estimate (β) scores and squared-multiple correlations (R²): .01–.059 = small, .06–.139 = moderate, ≥ .14 = large (Cohen, 1988).

Results

Factor structure

Findings from the first EFA indicated that the 17 RM items provided five factors with eigenvalues > 1 that accounted for approximately 46% of the variance and were characterized by lack of singularity (Bartlett’s χ² = 761.11, df = 136, p< .001), an adequate sample size (Kaiser-Meyer-Olkin [KMO] = .76), no multicollinearity (Matrix determinant>0), and moderate-to–strong communality estimates (h² range = .32–.71). Results of the pattern matrix indicated two non-loading items (λ ≤ .40) and six cross-loading items (λ > .32 across more than one factor). Moreover, findings from the parallel analysis and visual inspection of the scree plot suggested a two-factor solution would be a better fit to the data. Considering these outcomes, the two-factor structure was ultimately examined as the preferred model. Following excluding the non-loading and cross-loading items, the EFA was rerun constraining a two–factor solution. Further findings revealed the two–factor solution, comprising of 8 items accounted for 46% of the variance, with eigenvalues of 1.54 and 3.15. All factor loadings on two factors were strong exceeding .50, with no cross-loadings. Cross-loadings for all items were negligible (λ range = −.22–.25). Factor loadings of the scale were robust, with values ranging from .50 to .82, as shown in Table 3.

Table 3

Scale Items, Responsibility Domains, and Factor Loadings

Note. EFA = exploratory factor analyses; CFA = confirmatory factor analysis. λ₁ = item loadings for first-order factors; ℓ²₁ = indicator reliability for first-order factor items; λ₂ = first-order factor loading for second-order factor; ℓ²₂ = indicator reliability for second-order factor indicators; H = latent construct reliability for first-order and second-order factors, SS= Self-Responsibility Scale, SRS= Social Responsibility Scale.

Subsequently, the high-order CFA was conducted to confirm the latent structure of the scale using the second sample of the study. Findings from the CFA provided the good data-model fit statistics to the two-factor measurement model with 8 items, which were indicators of two latent constructs (i.e., self-responsibility and social responsibility)–χ² = 34.86, df = 19, p < .05, CFI = .96, TLI = .94, SRMR= .045, RMSEA [95% CI] = .080 [.035, .12]—that was characterized by robust latent structure reliability estimates (H range = .80–.90). Further results also showed that the scales had strong factor loadings, ranging from .55 to .80, as shown in Table 3.

Lastly, descriptive statistics and the internal reliability estimate of scale was performed. Findings from descriptive statistics indicated that the final 8-item measure and its subscales were relatively normally distributed (skewness and kurtosis < |2|), and item-total correlation estimates were large (r range = .63 to .84). reliability analyses also indicated that the scale and it sub–dimensions had strong internal reliability coefficients (α range = .86–to–.94), as shown in Table 4.

Table 4

Descriptive Statistics for the RM and Concurrent Validity Variables

Concurrent Validity

Bivariate correlation analysis was first performed to investigate the correlation between the RM and criterion variables (i.e., life satisfaction, psychological distress, personal responsibility, conscientiousness) selected for concurrent validity. Findings from correlation analysis revealed that the scale was moderately associated with psychological distress (r = –.34, p< .001) as well as had large correlations with life satisfaction (r = .58, p< .001), personal responsibility (r = .82, p< .001), and conscientiousness (r = .50, p< .001), as shown in Table 4. Subsequently, the latent variables path model, which preferred to investigate the predictive power of the first and second-order RM measurement model on life satisfaction and psychological distress, revealed that the high–order measurement had positive and significant predictive effect on life satisfaction (β = –.57, p< .001) and negative predictive effect on psychological distress (β = –.47, p< .001), accounting for approximately large of variance in the variables (R² = .14-.36), see Figure 1.

Figure 1

The Predictive Effect of Responsibility on Life Satisfaction and Psychological Distress

Note. ^**p<.001

Figure 2

Confirmatory Factor Analysis Results

Discussion

Responsibility is an important construct to mental health and wellbeing and helps individuals make better decisions and choices (Ryan et al., 2012). Responsibility provides satisfying and successful outcomes for people and contribute to others’ wellbeing (Wong, 2016; Wong & Wong, 2012). Therefore, this structure is a key to improve flourishing, resilience, and healthy functioning of individuals (Wong, 2010; Wong & Wong, 2012). Given the literature supporting the importance of responsibility to living a meaningful and happy life, it is critical to develop a scale for assessing of responsibility construct. The purpose of the present study is to investigate the initial development and validation of the Responsibility Scale (RS) to assess the sense of responsibility of individuals. Findings from exploratory factor analysis revealed the RS provided a two–factor solution comprising of 8 items that accounted for 46% of the variance, with equal items targeting characteristics of both personal and social responsibility. After examining the factor structure of the measure, confirmatory factor analysis was used to confirm the latent structure of the measure (Figure 2). Findings of this analysis supported the two–factor latent structure, providing provided good data-model fit statistics and by robust latent structure reliability estimates. Further results also showed that the internal reliability of the scale and its subscales were strong.  Finally, the correlation analysis and latent path model were conducted to examine the concurrent validity of the measure. Findings from correlation analysis first indicated the moderate–to–large association between responsibility and life satisfaction, personal responsibility, and psychological distress. Subsequently, the latent path model revealed that the first– and high–order measurement model had positive and significant predictive effect on life satisfaction and negative predictive effect on psychological distress, accounting for approximately large of variance in the variables.

Responsibility plays an essential role in the lives of people, influencing the choices they make and the willingness they show to hold themselves accountable for these choices (Mergler & Shield, 2016; Wong, 2016). Measuring this construct might help support people who may be making inappropriate decisions and choices and contribute to developing prevention and intervention services targeting to promote their mental health and wellbeing. Therefore, the present study has provided a number of methodological and empirical implications for research and practice.

At this particular juncture of history, many young people are suffering from both too much freedom and too little responsibility. On the one hand, Western culture encourages young people to hold the attitude that: “I want to do whatever I want to do”; thus, anything that interfere with such egotistic pursuit would be opposed as an infringement of their personal freedom. On the other hand, the current de-emphasis of personal responsibility means that when young people do whatever they want, they do not realize that they are not free from the consequences of their behavior. Often, their parents have to step in to bail them out, but it is unlikely that they can create a preferred future for themselves without a sense of responsibility.

That is why Wong (2019b) emphasizes that responsibility is the key to surviving and thriving in difficult time. No one is an isolated island. Whatever we do affect others; thus, personal responsibility involves social responsibility. We are more likely to prosper as individuals only when our society is safe, healthy, and prosperous. From the existential positive psychology perspective, responsibility is a master virtue which make all other virtues possible:

1. Since I am responsible for my life, I need courage to pursue my authentic path.
2. Since life is complex, competition is strong. I am responsible to develop knowledge and wisdom.
3. Since I am just one member of the human family, I am responsible to treat others with humanity.
4. Since I am the citizen of a society, I responsible to advocate social justice for all.
5. I am responsible to regulate my own emotions and desires and seek temperance to avoid harming myself and others.
6. Since my existence depends on the support from many people and a higher power, I need to seek transcendence and serve some entity greater than myself.

In his book Fear and Trembling, Kierkegaard (1843/1983) used the Biblical story of Abraham’s sacrifice of Isaac to argue that personal responsibility is taking a leap of faith to rise above one’s own desires and feelings in order to fulfill one’s “telos”. One’s life becomes meaningful when one takes the courageous action to become what one ought to be. According to Hanson (2020), “There is meaning in life for Kierkegaard precisely because there is a constitutive gap between the person I would be and the person I currently am.” Meaning is found in the struggle to become one’s ideal self.

In sum, the existential view adds a dimension of depth of understanding of responsibility and meaning in life as conceptualized by American positive psychology. The present Responsibility Scale commands itself as a useful tool to research the role of responsibility not only in meaningful living, but in various virtues as suggested above.