Abstract
This study investigated the relations between altruism, empathy, and spirituality in a sample of 186 university students. Zero-order and partial correlations controlling for age, sex, and social desirability indicated that, although altruism and empathy are related to each other in a manner consistent with previous research, the association of both of these to spirituality is complex and multidirectional. In particular, empathy was found to be significantly positively related to nonreligious spiritual cognitions, religiousness, and spiritual experiences and negatively associated with existential well-being. Altruism, on the other hand, was most strongly linked to spiritual experiences, followed by spiritual cognitions. Regression analyses revealed that nonreligious spiritual cognitions and spiritual experiences are the most potent predictors of both empathy and altruism, respectively. Based on the findings and literature suggesting that spirituality may have an impact on empathy and altruism, a path model was developed wherein spiritual experiences had a direct effect on altruism and spiritual cognitions had an indirect effect on altruism through empathy. A test of the model indicated that it demonstrates satisfactory goodness of fit. The study concludes with a discussion of the meaning and implications of the findings for future research.
In both the scientific and religious literature (e.g., such as “sacred” texts of virtually all world religions including Judaism, Christianity, Islam, Hinduism, and Buddhism), spiritual awareness or awakening as part of a devout religious practice (e.g., contemplation, meditation, prayer) has been linked to the expression of positive traits such as compassion, love, altruism, and empathy (Benson et al., 1980; DiLalla, Hull, & Dorsey, 2004; Kennedy, Kanthamani, & Palmer, 1994; L. B. Koenig, McGue, Krueger, & Bouchard, 2007; Morgan, 1983). In fact, the link is so ubiquitous that altruism has been likened to self-transcendence, a key component of spirituality in general and spiritual experience in particular (Krebs, 1991; Kristeller & Johnson, 2005; Post, 2003; Post, Underwood, Schloss, & Hurlbut, 2002), and at least one contemporary model of spirituality developed from a content analysis of the available psychological literature explicitly incorporates altruism as a core dimension of the construct (Elkins, Hedstrom, Hughes, Leaf, & Saunders, 1988).
Nevertheless, the empirical research has not produced wholly consistent findings regarding a spirituality–altruism or spirituality–empathy association. For instance, if spirituality is defined in terms of religiousness, some investigations suggest that religious individuals demonstrate less compassion toward racial injustices and are less likely to provide help to strangers (e.g., Jackson & Esses, 1997; Rokeach, 1970). Shariff and Norenzayan (2007) found that self-reported religiosity was unrelated to prosocial behavior, and Duriez (2004) found that religiousness itself was unrelated to empathy, whereas the manner in which religious content is mentally processed and interpreted (e.g., whether the Bible is understood literally vs. symbolically) was associated (with symbolic interpretation linked to higher levels of empathy). Notwithstanding the divergent findings, the overall literature is supportive of further investigation into the links between altruism, empathy, and spirituality.
Problems With How Spirituality Has Been Measured
When one considers the available research findings in their totality, what becomes apparent is that the manner in which spirituality is operationalized seems to influence the type of relationship found with other variables. For example, when religiousness was isolated from spirituality, Saroglou, Pichon, Trompette, Verschueren, and Dernelle (2005) found a difference in the types of prosocial behavior, with people high on religiousness engaging in helping behaviors toward loved ones but not strangers, whereas those high on spirituality showed helping behaviors toward both loved ones and strangers. In light of this, it would seem that greater consideration should be given to how spirituality is defined and measured before definitive conclusions about its relation and possible causal role in facilitating prosocial phenomenon, specifically altruism and empathy, are reached.
Over the past 30 years, numerous theories and measurement models of spirituality and related constructs have been proposed, but there has been little systematic development of a cumulative body of empirical research because of lack of adherence to and utilization of a well-standardized, scientifically rigorous model by investigators (MacDonald & Friedman, 2002; MacDonald, Friedman, & Kuentzel, 1999). Consistent with this, an examination of the extant measurement models reveals that there is an extremely muddled understanding of what does and does not constitute spirituality within the scientific community. In fact, problems with conceptualization are so rampant that some recognized authorities in the field of religion and health have called into question the very value of studying spirituality at all (e.g., H. G. Koenig, 2008).
Recognizing the problems with the literature and the need for an organizing framework to make sense of the plethora of models and tests, MacDonald (2000a, 2000b) developed a model of spirituality through the conjoint factor analyses of about 18 different available measures of spirituality and related constructs using data obtained from 1,472 university students. This model, which advances spirituality as an integral domain of human functioning but which, because of its inherent ineffability, can only be understood descriptively through its cognitive, affective, experiential, and behavioral expressions, comprises five dimensions and includes nonreligious spiritual cognitions (i.e., beliefs and attitudes about the validity and relevance of spirituality to daily functioning without alliance to any specific religious system), spiritual experiences, paranormal beliefs, existential well-being, and intrinsic religiousness (i.e., engaging in religious belief and practice for its own sake). This multidimensional model has not only been found to be comprehensive in its coverage of spirituality (e.g., all but one of the 18 measures contributed to at least one but not all five dimensions in the factor analysis) but has also been shown to hold promise for clarifying the relation of spirituality to a variety of aspects of health and human functioning. For instance, MacDonald and Friedman (2002) used the five dimensions to review the available literature and concluded that, contrary to the apparent perception that spirituality has a generally positive relation to health, its association to physical and psychological functioning is complex and multidirectional. In particular, they reported that existential well-being followed by religiousness and nonreligious spiritual cognitions appeared to be most consistently related to positive health outcomes whereas the findings for spiritual experience are mixed and those for paranormal beliefs more generally negative. To date, no research has been done examining the relation of MacDonald’s (2000a) model to altruism and empathy.
In this vein, the purpose of the present study was to empirically investigate the relation of spirituality, defined as per MacDonald’s (2000a) model, to both altruism (i.e., prosocial behavior toward another person that is aimed at enhancing the well-being of the other) and empathy (i.e., prosocial emotion that includes awareness of another’s suffering and affective participation in the other’s feelings). Both altruism and empathy were examined concurrently because of the research that has found a robust positive association between these constructs (Batson, Ahmad, & Stocks, 2004; Batson, Duncan, Ackerman, Buckley, & Birch, 1981; Batson, O’Quin, Fultz, Vanderplas, & Isen, 1983; Piliavin & Charng, 1990; Toi & Batson, 1982). In fact, it has even been concluded that “empathy is causally related to prosocial behavior. . . . Empirical studies have consistently supported the empathy-altruism prediction” (Piliavin & Charng, 1990, pp. 36-37), a position echoed in more recent publications (e.g., Batson et al., 2004). Based on the available research, it was generally expected that spirituality would be significantly associated to both altruism and empathy. To be more specific, given the findings of studies such as Duriez (2004), Saroglou et al. (2005), and Kristeller and Johnson (2005), it was hypothesized that spirituality as defined in terms of religiousness, nonreligious spiritual cognitions, and spiritual experiences would be significantly related to altruism and empathy. It was also expected that altruism and empathy would demonstrate a significant association.
Method
Participants
The sample consisted of 186 undergraduate psychology students at a Catholic-affiliated university located in the Midwestern United States. It comprised 45 males and 141 females, with a mean age of 23.03 years (SD = 7.37 years, range = 18-62). The sample was predominantly Caucasian (61.3% of the sample) and Christian (85.5% of the total sample).
Measures
Expressions of Spirituality Inventory (ESI-Revised; MacDonald, 2000a, 2000b)
The ESI-Revised is a 30-item self-report questionnaire derived from a longer 98-item parent version that uses a 5-point Likert-type scale (MacDonald, 2000a, 2000b) ranging from 0 (strongly disagree) to 4 (strongly agree). The ESI measures five distinct dimensions of spirituality including (a) Cognitive Orientation toward Spirituality (COS; i.e., nonreligious beliefs about the nature and importance of spirituality; example item—“Spirituality is an essential part of human existence”), (b) Existential Well-Being (EWB; i.e., sense of meaning, purpose, and ability to effectively cope with adversity; example of reverse-scored item—“I am an unhappy person”), (c) Experiential/Phenomenological Dimension (EPD; i.e., spiritual experience; example item—“I have had an experience in which I seemed to be deeply connected to everything”), (d) Paranormal Beliefs (PAR; i.e., belief in the existence of parapsychological phenomena; example item—“It is possible to communicate with the dead”), and (e) Religiousness (REL; i.e., intrinsic religious orientation and religious behavior such as meditation; example item—“I believe that God or a Higher Power is responsible for my existence”). The ESI-Revised has satisfactory interitem reliability, with α coefficients ranging from .80 to .89 for the different scales. All five scales of the ESI (both parent and revised versions) also have satisfactory convergent, discriminant, criterion (i.e., known groups), and factorial validity.
Self-Report Altruism Scale (SRAS; Rushton, Chrisjohn, & Fekken, 1981)
The SRAS is a 20-item survey where participants rate their answers on a 5-point Likert-type scale ranging from 1 (never) to 5 (very often). This tool was designed to measure altruism, or prosocial behaviors, which may be motivated by either egotism or empathy. Examples of item content include “I have delayed an elevator and held the door open for a stranger,” “I have given money to a charity,” and “I have offered to help a handicapped or elder stranger cross the street.” The SRAS has satisfactory reliability (e.g., α= .89) and satisfactory validity as seen by correlations with peer ratings and associations to conceptually similar tests.
Questionnaire Measure of Emotional Empathy (QMEE; Mehrabian & Epstein, 1972)
The QMEE is a 33-item survey where participants rate their answers on a 9-point Likert-type scale ranging from −4 (very strong disagreement) to +4 (very strong agreement). This tool was designed to measure emotional empathy (i.e., awareness of another’s suffering and affective participation with the other in suffering). Examples of item content include “The people around me have a great influence on my moods,” “I cannot continue to feel okay if people around me are depressed,” and “I tend to get emotionally involved in a friend’s problems.” The QMEE has adequate split-half reliability (r = .84) and some support for its convergent and discriminant validity (Crandall, 1980; Crandall & Harris, 1976).
Balanced Inventory of Desirable Responding (BIDR; Paulhus, 1984, 1991)
The BIDR is a 40-item survey designed to measure Paulhus’s (1984) two-factor model (i.e, impression management and self-deceptive enhancement) of social desirability. Participants rate their answers on a 7-point Likert-type scale ranging from 1 (not true) to 7 (very true). The BIDR has a satisfactory reliability (e.g., α = .83, retest correlations from .65 to .69) and evidence of reasonably good concurrent, convergent, and discriminant validity.
Procedure
Participants were recruited through a short presentation given to undergraduate-level psychology courses. As an incentive for participation, potential recruits were offered course extra credit from the instructor. All participants were given informed consent forms and questionnaires and were instructed to return the materials to the researchers when completed.
Results
Zero-order correlations were calculated to examine the interrelation of the measures of spirituality, altruism, and empathy as well as their associations to age, gender, and social desirability scores (see Table 1).
Zero-Order and Partial Correlations Between Age, Gender, Measures of Spirituality, Altruism, Empathy, and Social Desirability
Note: N = 186. Decimals omitted. Coefficients to the right of the diagonal are partial correlations wherein age, gender, and social desirability were controlled; df = 181. ESI = Expressions of Spirituality Inventory; COS = Cognitive Orientation toward Spirituality; EWB = Existential Well-Being; EPD = Experiential/Phenomenological Dimension; PAR = Paranormal Beliefs; REL = Religiousness; QMEE = Questionnaire Measure of Emotional Empathy; SRAS = Self-Report Altruism Scale; BIDR = Balanced Inventory of Desirable Responding. For gender, 0 = male and 1 = female.
p < .05. **p < .01. ***p < .001.
Inspection of the table reveals a fairly large number of statistically significant correlations. Both altruism and empathy are significantly correlated to each other (r = .27, p < .001) and both were found to be significantly associated with multiple ESI dimensions. In the case of altruism, significant correlations were generated with ESI COS (r = .30, p < .001), REL (r = .19, p < .05), and the EPD (r = .36, p < .001). For empathy, significant correlations were obtained with four of the five ESI dimensions (i.e., all but PAR) and include COS (r = .30, p < .001), REL (r = .28, p < .001), EPD (r = .16, p < .05), and a negative correlation with EWB (r = −.21, p < .01).
Turning to the demographic variables, age was found to significantly correlate with three ESI dimensions—COS (r = .22, p < .01), REL (r = .25, p < .001), and EPD (r = .23, p < .01)—and with altruism (r = .31, p < .001). Gender (where male was coded 0 and female was coded 1) correlated significantly with empathy (r = .44, p < .001).
Last, overall social desirability scores were found to significantly correlate with ESI COS (r = .20, p < .01), ESI REL (r = .17, p < .05), ESI EWB (r = .34, p < .001), and altruism (r = .19, p < .05).
Given the significant correlations of age, gender, and social desirability and the available literature suggesting that these may present as confounding variables in spirituality and religion research (e.g., H. G. Koenig, 2006; Wink & Dillon, 2002), partial correlations were calculated while controlling for them to better understand how spirituality, altruism, and empathy relate (see Table 1).
Examination of the partial correlations reveals a pattern of significant coefficients very similar to the zero-order correlations. In particular, altruism was found to produce significant partial correlations with ESI COS (partial r = .23, p < .05) and ESI EPD (partial r = .32, p < .001). Empathy generated significant positive partial correlations with ESI COS (partial r = .33, p < .001), ESI EPD (partial r = .18, p < .05), and ESI REL (partial r = .26, p < .001). A significant negative partial correlation was obtained between empathy and ESI EWB (partial r = −.24, p < .001).
Regression Analyses
To better determine how the various dimensions of spirituality as measured by the ESI both individually and collectively relate to empathy and altruism, two standard multiple regression analyses were conducted wherein the five ESI dimensions served as predictors and altruism and empathy scores, respectively, were used as the criterion variables (see Table 2).
Standard Multiple Regression Results Using ESI Dimensions as Predictors of Empathy and Altruism
Note: N = 186. sr2 = squared semipartial correlation.
p < .05. **p < .01. ***p < .001.
Considering first the regression involving empathy as the dependent variable, analyses indicated that all five ESI dimensions significantly predict empathy scores, F(5, 180) = 5.79, p < .001, R = .37, R2 = .14. ESI COS (Β = 1.14, β = .28, p < .05, sr2 = .02) and ESI EWB (Β = −1.13, β = −.22, p < .01, sr2 = .04) emerged as significant predictors. Squared semipartial correlations (sr2) for these two dimensions reveal that ESI COS uniquely accounts for 14.29% of the total explained variance (i.e., R2), whereas ESI EWB uniquely accounts for 28.57% of the total explained variance.
Turning next to the regression with altruism as the criterion variable, all five ESI dimensions collectively significantly predicted altruism scores, F(5, 180) = 6.76, p < .001; R = .40, R2 = .16. ESI COS (Β = 0.62, β = .34, p < .05, sr2 = .03) and ESI EPD (Β = 0.57, β = .28, p < .001, sr2 = .05) were found to be significant predictors, each uniquely accounting for 18.75% and 31.25% of the total explained variance, respectively.
Path Analysis
Given the results of the correlational and regression analyses, it appears possible to identify a directional path model in which specific aspects of spirituality can be used to predict empathy and altruism, respectively. Figure 1 presents a potential path model, whereby spiritual experience (ESI EPD) directly predicts altruism, and spiritual cognition (ESI COS) is linked to altruism through empathy. ESI COS and ESI EPD are correlated in the model.

Path model of spirituality predicting empathy and altruism
Analysis of model fit using structural equation modeling software (i.e., AMOS) revealed that all parameters (i.e., regression weights and correlations) are significant at p < .05 (standardized regression weights are reported in Figure 1). Statistics pertaining to overall fit suggest that the model fits the data well (e.g., χ2 = 1.15, df = 2, p > .05; χ2/df = .58; goodness-of-fit index = 1.0; adjusted goodness-of-fit index = 0.99; nonnormed fit index = 0.99; Tucker–Lewis index = 1.0).
Discussion
Results indicate that spirituality, altruism, and empathy are all significantly related to one another. Specific associations with the dimensions of spirituality in MacDonald’s (2000a) model indicate that nonreligious spiritual cognitions, spiritual experience, and less so religiousness demonstrate positive relations with both altruism and empathy, whereas existential well-being was found to be significantly and inversely related to empathy and nonsignificantly related to altruism. Paranormal beliefs were found to be unrelated to both altruism and empathy.
These findings are generally in line with existing research that supports the link between empathy, altruism, and their association with spirituality (Batson et al., 2004; L. B. Koenig et al., 2007; Kennedy et al., 1994; Saroglou et al., 2005). With regard to spiritual experience, the current findings bolster the view of altruism as being an expression of self-transcendence (Krebs, 1991; Kristeller & Johnson, 2005; Piliavin & Charng, 1990; Post, 2003; Post et al., 2002). Furthermore, the way in which people think about spirituality (i.e., spiritual cognitions) rather than mere religiousness is more potently related to empathy. This finding seems consistent with Duriez (2004), who found that the manner in which people think about religious ideas, rather than how people act in religious practices, is associated with higher levels of empathy.
Correlational and regression findings led to the development of a path model that strongly fits the data and points to the potential of using spirituality as a vehicle for cultivating empathy and altruism. In particular, the path model suggests it may be possible to enhance prosocial attitudes and behaviors by introducing people to certain activities (e.g., meditation) that facilitate the experience of self-transcendence and nurture the development of more nonliteral ways of thinking about spirituality or religious concepts. This use of spiritual methodology for personal transformation has been previously shown in the empirical literature to have positive effects on both empathy and altruism (DiLalla et al., 2004; Kristeller & Johnson, 2005).
Notwithstanding these findings, the results involving religiousness and existential well-being indicate that the overall relation of spirituality to prosocial attitudes and behaviors is not wholly positive. With religiousness, although our analyses showed it is significantly linked to empathy, notably weaker relations with altruism were found. Although some investigations have indicated that religiousness is tied to altruism (e.g., Benson et al., 1980; Morgan, 1983), other studies refute such an association (e.g., Shariff & Norenzayan, 2007). The present finding is consistent with portions of the complex research suggesting that religious people may feel more empathy and prosocial emotions for others (Saroglou et al., 2005) but may not act on these feelings as manifested in prosocial behaviors.
The findings involving existential well-being, wherein it is significantly negatively associated with empathy and nonsignificantly related to altruism, come somewhat as a surprise. Existential well-being, at least as conceived of by MacDonald (2000a), relates to the extent to which a person feels good about himself or herself and his or her ability to effectively cope with existential adversities in life (e.g., deriving a sense of meaning and purpose). In past research, existential well-being has been found to be inversely associated with antisocial attitudes and behaviors (e.g., MacDonald & Holland, 2003). Given this, it might be expected that people who feel efficacious and generally good about themselves and their situation in life would be more inclined to be helpful toward others. Instead, the present findings suggest that existential well-being has no impact on whether or not a person behaves in an altruistic manner, and it is even indicated that people with high existential well-being may be limited in their ability to empathize with the suffering of others.
These findings seem to be best explained in terms of the controversy around whether or not altruism is motivated by egotism. Though some researchers maintain that genuine altruism is possible (e.g., Batson & Shaw, 1991a, 1991b), others argue that altruism is motivated by an egoic need to minimize one’s vicarious suffering or maximize one’s feeling of goodness and superiority rather than a desire to truly be of help to another person (e.g., Cialdini, 1991; Hornstein, 1991). This argument can be taken even further. As it relates to empathy, existential well-being has been found in the present study and in past research (MacDonald, 2000a) to be significantly related to measures of social desirability and self-deceptive enhancement. Extending from this, and considering the research that shows that mental health is at least partially tied to exaggerated positive views of one’s self and one’s capabilities (e.g., Taylor & Brown, 1988; Taylor, Brown, Colvin, Block, & Funder, 2007), it may be that people higher in existential well-being show diminished empathy because they are motivated (whether consciously or unconsciously) to protect their own sense of competence and wellness from stressors or events that could introduce suffering into their self-consciousness. That is, the ability to empathize with others may be hindered by high levels of existential well-being because people do not want to feel suffering and do not want to face the possibility that they cannot ameliorate the suffering of themselves or others (thereby challenging their sense of self-efficacy). Clearly, more research is needed to explore this possibility.
Although the results of this investigation demonstrate that altruism, empathy, and spirituality are interrelated, some degree of caution is warranted in the interpretation of these findings. First, the sample consisted of mostly traditional undergraduate students at a religiously affiliated university. It is possible that the relation of spirituality to empathy and altruism may change as a function of age, maturation, and religious background (e.g., H. G. Koenig, 2006). Second, the path model devised and tested herein was based on the correlational and regression analyses rather than on a priori assumptions about how spirituality and empathy may influence levels of altruism. Since it was known that all parameters in the model would emerge as statistically significant, it thereby stands to reason that the current model would demonstrate satisfactory fit. Future research is required to verify the robustness of this directional model. Third and finally, the manner in which constructs were operationalized may not give a wholly accurate picture as to how they actually relate. For instance, empathy has been shown to consist of a variety of components, including fantasy (unconscious identification), empathic concern (genuine caring), personal distress (vicarious suffering), and perspective-taking (intellectual understanding). Two of these components—empathic concern and perspective-taking—have been found to be moderately correlated to altruism (Litvak-Miller & McDougall, 1997). The measure of emotional empathy used in this study did not clearly tease apart the various component elements in a way that permitted a more detailed analysis of the relation of empathy to both spirituality and altruism.
In conclusion, the present findings are generally consistent and supportive of the literature advocating associations between spirituality, empathy, and altruism. However, the use of a scientifically rigorous multidimensional model of spirituality has revealed that the relations of these constructs are complicated, a finding congruent with MacDonald and Friedman’s (2002) observation that the relation of spirituality to health is also multidirectional. It is hoped that this study will spur future research, which will give due consideration to the richness and complexity of spirituality when investigating its relation to human psychosocial functioning.
Footnotes
The author(s) declared no potential conflicts of interest with respect to the authorship and/or publication of this article.
The author(s) received no financial support for the research and/or authorship of this article.
