A Construal Level Account of the Impact of Religion and God on Prosociality

Abstract

This research shows that the two most prevalent religious constructs—God and religion—differentially impact cognition. Activating thoughts about God (vs. religion) induces a relatively more abstract (vs. concrete) mindset (Studies 1a–1c). Consequently, time donation intentions (Study 2) and actual monetary donations (Study 3) after a God (vs. religion) prime increase when people are presented an abstractly (vs. concretely) framed donation appeal. Similarly, people donate more money to distant (vs. close) donation targets, which are construed relatively abstractly (vs. concretely), when a religious speech activates predominantly God-specific (vs. religion-specific) thoughts (Study 4). These effects are mediated by “feeling right” under construal level fit (Study 3). Overall, this research significantly advances extant knowledge on religious cognition and past research on the link between religion and prosociality.

Keywords

religion God prosociality donation construal level

It has long been debated in psychological research whether religion makes people more prosocial and altruistic (see, for reviews, Galen, 2012; Preston et al., 2010; Shariff et al., 2016). However, the empirical studies examining this link have provided puzzling results. While some studies have documented that religion and religious primes increase prosocial behavior compared with a neutral condition (Batson et al., 1989; Bekkers & Schuyt, 2008; Brooks, 2006; Sosis & Ruffle, 2004), other research has evinced no significant effect of religion on prosociality (Nowell & Laufer, 1997; Smith et al., 1975). Although relatively fewer in number, some studies have even reported antisocial effects of religion (Guttmann, 1984).

There are two common issues in past research on religious prosociality that might have resulted in these seemingly contradictory findings. First, the term “religion” has remained largely an umbrella term, one that includes all aspects of religious cognition from church membership and attendance in religious congregations to primes of supernatural agents. Second, extant studies measured prosocial intentions toward qualitatively distinct agents from anonymous strangers to close others. Following recent reformulations in the field of religious cognition suggesting that God and religion are distinct constructs with diverging cognitive and motivational corollaries (Ritter & Preston, 2013), we show in the present research that the cognitive fit among the specific religious construct being activated and the salient characteristics of the decision context moderates the impact of religion on prosociality. We show in five experiments and one field study that activating thoughts about God (vs. religion) leads to relatively higher (vs. lower) abstraction in thinking. This, in turn, influences prosocial intentions and actual behavior when other characteristics of the context fit, or misfit, this level of construal. More specifically, prosocial intentions and actual donations are higher for psychologically distant (vs. proximate) targets or for abstractly (vs. concretely) framed donation requests when decision makers are primed with God (vs. religion). We further show that this effect is mediated by an experience of “feeling right,” which occurs under construal fit.

The current research significantly advances previous research on religious cognition and the link between religion and prosociality. By developing a general framework for explaining differential cognitive consequences of the two most prevalent religious concepts, we help explain the general pattern in previous research on the impact of religious primes on prosocial tendencies. This framework also opens up a new avenue of research by providing new insights and novel predictions.

In what follows, we first review past research on religious prosociality. This is followed by a brief discussion of construal level theory (CLT; Trope & Liberman, 2010) and construal level definitions of abstraction and concreteness, which sets the basis for our premises. Next, drawing on past conceptualizations of the concepts of God and religion, we present our rationale regarding our prediction that priming God (vs. religion) results in relatively higher (vs. lower) abstraction in thinking, and messages or targets that are construed more abstractly (vs. concretely) lead to higher prosociality when people are reminded of God (vs. religion). Finally, we present empirical evidence from five experiments and one field study that support our predictions, and we discuss the theoretical implications of our findings and avenues for future research.

Past Research on Religious Prosociality

The link between religion and prosociality has long been investigated by researchers in different fields. Early research found a positive link between self-reported religiosity and self-reported altruistic behaviors (for a review, see Batson et al., 1993). In one study using frequency of prayer as an indication of personal religiosity, for example, Morgan (1983) found that more religious people are more friendly and cooperative than their less religious counterparts. American participants who frequently pray and attend religious services also reported more charitable giving and volunteering (Brooks, 2006). Similarly, more religious people rated themselves as being agreeable (Saroglou, 2002) and ready to forgive (McCullough & Worthington, 1999). Although this stream of research is useful in understanding how religion influences prosocial tendencies, they have two methodological limitations. First, unlike experimental research, they do not allow for causal interpretation. Second, they rely heavily on self-reports of prosocial tendencies, which people often exaggerate for social approval reasons (Gervais & Norenzayan, 2012).

To address these limitations and find a causal link between religion and prosocial behavior, more recent research has employed experimental methodology. Several of these studies found that priming people with religious concepts increases cooperation (Ahmed & Salas, 2011; Shariff & Norenzayan, 2007), generosity (Ahmed & Salas, 2011), and contribution to public goods (Ahmed & Hammarstedt, 2011) in economic games. These prosocial effects of religion were also reported by other research that showed that religious primes lead to increased prosocial tendencies in the form of taking more pamphlets from a local charity for distribution (Pichon et al., 2007) and to increased honesty and lower cheating on a subsequent task that provided the opportunity to claim more money than one actually deserved (Randolph-Seng & Nielsen, 2007). These effects occurred even when religious thoughts were simply activated through the salience of immediate environmental contexts such that people intended to contribute more to a common pot in a public-goods game when they were in a church rather than a lecture hall (Ahmed & Salas, 2013).

On the contrary, several other studies have suggested a causal link between religion and anti-social behaviors. M. K. Johnson et al. (2010), for example, showed that a subliminal prime of religious words increases prejudice against value-inconsistent others. In another empirical investigation, McKay et al. (2011) asked participants to play a two-stage economic game. The first stage involved allocating money between one’s self and the opponent, and the second stage involved one’s opponent’s decision in the first stage. In other words, participants were given the chance to retaliate in the second stage. The authors found that, among participants who made donations to a religious charity in the past, a religious prime increased retaliation and stricter punishment of opponents who allocated less money previously.

In summary, causal investigations of the link between religion and prosociality have yielded mixed results. Notably, one problem of this stream of research is the use of “religion” and “religious prime” as an umbrella term, as is priming religion with various concepts and words from Christian to God (Preston et al., 2010). This raises the possibility that divergent effects of religious primes on prosocial behavior might be partly explained by the nature of priming (Ritter & Preston, 2013). In fact, using multidimensional scaling, property fitting, and cluster analysis, Ritter and Preston (2013) found that different religious concepts are represented under different categories, each of which is likely to be associated with different cognitive characteristics. This suggests that God and religion might be different cognitive constructs with differential cognitive, affective, and behavioral outcomes. The possibility of distinct impacts of different religious concepts on human cognition, however, has not yet been investigated.

In the recently growing experimental tradition of research on the relationship between religion and prosociality, only a limited number of studies have attempted to distinguish between the effects of the two most common religious concepts—religion and God—by manipulating either one or the other to prevent any confound between the two constructs (Batara et al., 2016; Preston & Ritter, 2013; Ramsay et al., 2016). In one of these studies, Preston and Ritter (2013) found that activating religion leads to a greater likelihood of helping an ingroup, while reminding people of God increases prosocial intentions for outgroups because of different motivations associated with these two constructs. According to this motivational explanation, religion primarily activates ingroup protection motives whereas God heightens impression-management concerns, leading to a greater likelihood of prosociality toward ingroup (vs. outgroup) members under religion (vs. God) prime. In the current research, we examine another possibility that these constructs are also associated with different levels of mental abstraction, which could possibly explain why the impact of the two concepts on prosociality varies with respect to contextual factors. In the next section, we summarize CLT, which provides a general framework to help understand and explain the corollaries of mindset abstraction.

CLT

CLT (Trope & Liberman, 2010) proposes that any behavior, person, object, or concept could be categorized at varying levels of abstractness (Vallacher & Wegner, 1987, 1989) from relatively decontextualized and superordinate high-level representations to contextualized and subordinate low-level representations (Medin & Ortony, 1989). High-level construals convey the essential and superordinate meanings of the targets of perception. In contrast, low-level construals are contextualized representations that reflect the subordinate secondary features (Trope & Liberman, 2010). In line with this, one critical distinction between high and low levels of construal is that the former refers to why a certain action is performed, while the latter refers to how the action is completed (Trope & Liberman, 2003). For example, the action of “voting” might be construed at a higher, more abstract level of “influencing the election,” which refers to the superordinate goal of the action. Alternatively, it could be construed at a relatively lower, more concrete level of “marking a ballot,” which reflects the subordinate process of voting. This distinction is not limited to descriptions of concrete behaviors or actions; rather, it could be generalized to abstract concepts. The concept of democracy, for instance, could be construed either at an abstract level through its ultimate goal of “establishing the power of people” or at a concrete level through its subordinate process of “voting in elections.”

This distinction implicates a difference between focusing on the end (or the purpose) versus the means (or the process) associated with a given action, event, or concept (Vallacher et al., 1992). The purpose or the end associated with a concept or an action conveys the decontextualized underlying meaning of it, whereas the means are concrete and contextualized manners that help achieve an end. As such, an ample amount of empirical evidence has demonstrated that individuals who focus on the purpose exhibit an abstract level of mental construal, whereas those who focus on the process consequently adopt a more concrete level of mental construal (Freitas et al., 2004; Liberman et al., 2007; Stephan et al., 2011). This is because ends (or purposes) refer to superordinate, abstract goals reflecting why something is achieved, while means (or processes) refer to the subordinate, concrete actions that concern how an end is achieved.

In a similar vein, different levels of mental construal are associated with a difference in relative preference between desirability and feasibility concerns. Desirability concerns, which are related to the valence of an end state, are prioritized under abstract, higher levels of mental construal. On the contrary, feasibility concerns which involve the ease or the availability of necessary means to reach an end state are heightened under concrete, lower levels of mental construal (Vallacher & Wegner, 1987). Past research has documented that decision makers attend to high-level features of desirability and adopt abstract thinking when contextual factors make them think of superordinate meanings and form a high-level view of a situation; however, they adopt a concrete mindset, prioritizing low-level feasibility concerns when contextual factors remind them of low-level, concrete actions or manners (Trope & Liberman, 2003).

In summary, past research in the CLT tradition has shown that an important corollary of mindset abstraction is whether a salient concept makes people think of general, high-level, abstract characteristics of it that convey ultimate goals or whether it makes them attend to relatively low-level, subordinate, concrete actions that act as the means of achieving an ultimate end.

God, Religion, and the Level of Mental Construal

A prominent feature of all large-scale religious cultures is that God(s), or deities, are construed as a supernatural, moralizing agent that ask for human submission (Burkert, 1996) to higher order moral rules of what is right or wrong (Boyer, 2001; Preston et al., 2010). In other words, God refers to a concept that sets high-order moral rules that involve the values of what is right and what is wrong (Gray & Wegner, 2010). The moral values reflect ends, or purposes, in the sense that they provide believers a sense of the ultimate goals to be reached (e.g., “it is wrong to steal”). People use simple, intuitive heuristics while processing these values (Sunstein, 2005), and they generally ignore contextual information that would change the meaning of these values (Baron, 1994; Haidt, 2001). Thus, the moral values are relatively decontextualized, as their meanings and the ultimate moral principles they imply are hard to change with respect to the context (Eyal et al., 2008).

The moralizing nature of God, or other deities, that is common across large-scale religions relates to another characteristic of God(s) that is of important interest in the current research. The moralizing God is an ideal agent that not only sets higher rules but also possesses abstract, supernatural characteristics that would allow this agent to monitor the people’s deference to these rules (Gray & Wegner, 2010). Thus, particularly in Abrahamic religions (which constitute a significant portion of all believers, with more than four-billion adherents worldwide), God is attributed abstract, supernatural characteristics such as omniscience (i.e., knowing and seeing everything) and omnipotence (i.e., having ultimate and unlimited power to do anything; D. Johnson & Kruger, 2004; Laurin et al., 2012). In other words, the God concept is associated with both decontextualized higher order moral values and abstract characteristics that go beyond the senses and concrete imagination of people.

Religion, on the contrary, is a system that transcends beyond the idea of a supernatural agent. In an early theorizing, Durkheim (1912/1995) argued that, despite putting speculations about divine beings at the center, religious systems are indeed inherently social. The religious system is characterized by both beliefs about divine beings and concrete acts and practices called “rites” or “rituals.” While beliefs about divine beings like God define the ultimate object of rituals, which is to bring individuals closer to God by being a good or virtuous person (Carroll et al., 1986), many religious rituals which are specified in detail, such as religious congregations and prayers, act as the “pathways” to reach the “destination” (Pergament, 1997). To state this differently, God represents the superordinate values that act as ultimate destinations or ends to be reached (Schweiker, 1969). These relatively decontextualized values achieve meanings by being located in context through the subordinate actions or rituals involved in religion (Bellah, 1964). Thus, rituals involve the concrete actions and practices that represent the means through which the ultimate destination is reached.

The aforementioned distinctions between God and religion make it clear that the concept of God, which itself is attributed imperceptible and intangible qualities, involves symbolic superordinate values that act as ends. The concept of religion, on the contrary, involves concrete and tangible rituals and other concrete institutional agents, such as buildings where these rituals take place (Saroglou, 2006), and these rituals represent subordinate means of achieving the values implied by God. Viewing the construal level definitions in conjunction with the very definitions of abstraction and concreteness that center around perceptibility by the senses and tangibility (Vallacher & Wegner, 1987; Yan et al., 2016), we formally hypothesize that priming people with God and religion results in differential levels of mindset abstraction. More specifically, we predict that thinking about God brings abstract values and characteristics into people’s mind, leading to a higher level of mental construal. In contrast, we predict that exposure to the concept of religion leads people to think in more concrete terms.

Construal Fit and Message Persuasiveness

A large body of research has documented that different types of psychological distance have a strong bidirectional link with the level of mental construal (Liberman & Trope, 2008; Trope & Liberman, 2010). Objects, events, or people that are temporally (Liberman et al., 2002), spatially (Fujita, Henderson, Eng, Trope, Liberman, 2006), or socially (Stephan et al., 2010) distant are perceived in more abstract terms than their psychologically close counterparts. Consequently, the salience of psychologically distant targets in a given decision context results in higher levels of construal, as shown by participants’ reliance on broader, superordinate categories (Liberman et al., 2002) or their tendency to identify actions in more abstract terms (Fujita et al., 2006).

This relationship has significant implications for the persuasiveness of appeals. Emerging research suggests that decision makers experience a heightened subjective engagement in a message when there is a fit between the levels of mental construal implied by salient factors in the decision environment (Higgins, 2006; Lee et al., 2010). This engagement intensifies the favorability of evaluations through an experience of “feeling right” (Cesario & Higgins, 2008). Stated differently, when the levels of construal induced by salient factors in a decision context fit each other, consumers “feel right” about the message, leading to higher likelihood of compliance with the message. For instance, in one study investigating the effectiveness of political messages, Kim et al. (2009) found that abstractly (vs. concretely) framed messages are more persuasive when the election is temporally distant (vs. imminent). This is because higher (vs. lower) temporal distance is associated with a relatively higher (vs. lower) level of mental abstraction, which fits the linguistic abstraction (vs. concreteness) in the appeal. This construal fit makes decision makers “feel right” about the message, which increases the persuasiveness of the message.

The Current Research

Based on the above discussion, our research predicts that activating thoughts about God (vs. religion) results in higher (vs. lower) mental abstraction, which leads to higher prosocial intentions and actual prosocial behavior when other salient factors in the decision context are also associated with a higher (vs. lower) level of mental construal. We further predict this effect to be a construal fit effect and therefore driven by an experience of “feeling right” that occurs under construal fit.

We tested these predictions in five controlled studies and in one field study across different sources of abstraction (i.e., the abstraction of language used in the verbal description of the donation request; verbal description of the target; and the psychological distance between the donor and the receiver). Studies 1a to 1c tested our basic prediction that exposure to the concept of God (vs. religion) results in a relatively higher level of construal among samples from predominantly Christian and Muslim populations. In Study 2, we show a downstream consequence of this effect by demonstrating that prompting thoughts about God (vs. religion) leads to higher intentions of donating time to an organization when the organization is described in abstract (vs. concrete) terms. Study 3 replicates this fit effect in people’s real monetary donations and provides process evidence that the effect is mediated by “feeling right” experience of donors. Finally, Study 4, which involves 60-week donation data obtained from four mosques in a small city in Turkey, shows that Friday sermons that activate predominantly God-specific (vs. religion-specific) thoughts result in higher monetary donations for spatially distant (vs. close) targets.

Note that a recent meta-analysis of religious priming research with a focus on prosociality reports an average effect size of g = .40 (Shariff et al., 2016). A power analysis assuming this effect size indicated that at least 52 participants are required for a study (with numerator degrees of freedom being 1) to be powered at 80%. All studies reported below sufficiently meet this minimum requirement.

Study 1a

Study 1a tests the basic prediction that activating thoughts about God (vs. religion) leads to a higher (vs. lower) level of mental construal among a sample from a predominantly Christian population—U.S. Americans.

Method

Sixty U.S. participants (M_age = 34.2; 34 female) were recruited on Amazon Mechanical Turk to participate in the study in return for monetary reward. Participants first wrote down in a couple of sentences the thoughts that came to their mind when they saw the word God or Christianity (see Supplemental Material for details). Past research has demonstrated that four linguistic categories—descriptive action verbs (DAVs), interpretive action verbs (IAVs), state verbs (SVs), and adjectives (ADJs)—reflect differential levels of mindset abstraction, DAVs reflecting the least abstraction and ADJs reflecting the most abstraction (Semin & Fiedler, 1988). Hence, we predicted that the texts in which participants described their thoughts upon exposure to the concept would involve more (vs. less) abstract linguistic categories in the God (vs. religion) prime condition. Following the coding schema developed by Semin and Fiedler (1988) and commonly used in subsequent research (e.g., Douglas & Sutton, 2003; Menegatti & Rubini, 2013), a judge blind to research hypotheses counted the number of DAVs, IAVs, SVs, and ADJs used in each participant’s response. We next calculated based on these values a weighted linguistic abstraction score using the following formula developed by Semin and Fiedler (1988): [1 × (DAVs) + 2 × (IAVs) + 3 × (SVs) + 4 × (ADJs)] ÷ [DAVs + IAVs + SVs + ADJs]. Note that weights in the formula increase as the abstraction implied by the linguistic category increases; hence, a higher linguistic abstraction score reflects a higher level of construal.

After participants wrote down their thoughts, we collected measures of three control variables that might influence the results: (a) religious self-identification (1: “believer/religious”; 2: “spiritual”; 3: “agnostic”; 4: “non-believer/atheistic”), (b) the religion with which they identify themselves (1: “Christian”; “2: Jewish”; 3: “Muslim”; 4: “None”; 5: “Other”), and (c) the strength of their belief in God measured on a 101-point scale (0: “Not at all”; 100: “Extremely”). Twenty-five participants identified themselves as believer/religious, 13 as spiritual, nine as agnostic, and 13 as atheistic. All participants who self-identified as believer or religious stated that they are Christians. Finally, participants indicated their age and gender, and were debriefed and thanked.

Results and Discussion

Consistent with our theorizing, participants in the God prime condition heavily used abstract qualities associated with God (e.g., powerful, eternal, and all-knowing). In contrast, participants in the religion prime condition used more concrete practices and objects in their response (e.g., prayer, churchgoing, Christmas, and Bible). More formally, we analyzed the impact of the activated religious concept on the abstraction of language use by including in an analysis of variance (ANOVA) the primed concept as the independent variable and the language abstraction score (computed by the above formula developed by Semin and Fiedler, 1988) as the dependent variable. We kept religious self-identification as a covariate because it significantly influenced language abstraction, F(1, 57) = 5.29, p = .025; confidence interval (CI_95%) = [0.0155, 0.2242], η² = .085, d = .62. The results supported our prediction, F(1, 57) = 8.28, p = .006; CI_95% = [0.1082, 0.6029], η² = .085, d = .81. Participants who were asked to write their thoughts about God used significantly more abstract language (M = 3.02, SD = 0.51) than those who wrote their thoughts about Christianity (M = 2.69, SD = 0.48), showing that God (vs. religion) activates more abstract (vs. concrete) cognition. Note that the impact of religious concept on language abstraction remains significant even when religious identification is not included in the model as a covariate, F(1, 58) = 6.47, p = .014; CI_95% = [0.0689, 0.5782], η² = .100, d = .71.

A separate analysis of the impact of religious self-identification on language abstraction showed that participants who identified themselves as agnostic used a significantly more abstract language (M = 3.07, SD = 0.55) than those who self-identified as believer/religious (M = 2.67, SD = 0.56; F = 5.66, p = .021; CI_95% = [−0.7453, −0.0639], η² = .092, d = .65). Those who identified themselves as being atheistic or spiritual, however, did not differ from believers or agnostics in the level of abstraction of the language they used (ps = n.s.). This finding is unsurprising because agnostics by definition believe in unknowability of religious claims whereas believers are certain in their belief in God or in the validity of other religious claims. This aspect of agnosticism is also what differentiates it from atheism as agnostics perceive the existence of a deity or deities as being probabilistic and they take an indifferent position. The fact that uncertainty is associated with a higher level of mental construal than certainty (Bar-Anan et al., 2006) might have driven this result such that agnostics might have used a relatively more abstract language when they are asked to think about concepts which prompt uncertainty or unknowability.

Study 1b

Study 1b aimed at heightening the validity and generalizability of the result obtained in Study 1a by replicating it among a sample from a predominantly Muslim population, and by using an additional measure of mindset abstraction.

Method

One hundred college students (M_age = 21.18; 68 female) participated in the study in return for course credit. Participants were first exposed to one of the two concepts (“Allah” or “Islam”), and they were asked to write down in a couple of sentences the thoughts that came to their mind when they first saw the given word. Next, we administered Behavioral Identification Form (BIF; Vallacher & Wegner, 1989), which is a 25-item dichotomous questionnaire that measures whether people identify actions at an abstract or concrete level. Finally, we collected the same control and demographic measures as in Study 1a. Forty-eight participants self-identified as religious/believer, 22 as agnostic, 19 as spiritual, and 11 as atheistic. All believer/religious participants indicated that they are Muslims.

Results

Action identification

We subjected participants’ responses to BIF to a binary coding (concrete description of the action = 0, abstract description of the action = 1). We computed a composite score of action identification by summing their responses to each item. In other words, a higher score indicated a higher level of construal. None of the covariates had a significant effect on BIF score; hence, they were dropped from analysis. A one-way ANOVA yielded a significant effect of concept on BIF score, F(1, 98) = 7.63, p = .007; CI_95% = [0.7776, 4.7423], η² = .072, d = .78. Supporting our predictions, exposure to the God concept resulted in significantly more abstract description of actions (M = 17.16, SD = 4.65) than exposure to the concept of religion (M = 14.40, SD = 5.32).

Language abstraction

As in Study 1a, we also analyzed the abstraction in language use upon exposure to either concept. Based on the coding of a judge blind to our hypotheses, we calculated the weighted linguistic abstraction score in the same way as in Study 1a. A one-way ANOVA on this score provided further support for our prediction. Participants used significantly more abstract language while describing the thoughts that came to their mind when they saw the word “Allah,” M_Allah = 2.79, SD = 0.51; M_Islam = 2.49, SD = 0.52; F(1, 98) = 8.23, p = .005; CI_95% = [0.0914, 0.5011], η² = .078, d = .81.

Discussion

Study 1b replicates our finding in Study 1a that activating thoughts about God (vs. religion) prompt a relatively higher (vs. lower) level of mental construal among a sample from a predominantly Muslim population by using an additional measure of mindset abstraction. Note, however, that the priming procedures in Studies 1a and 1b involved specific religious denominations (i.e., “Christianity” or “Islam”) rather than the religion concept itself. We address this issue in Study 1c below.

Study 1c

Two main objectives guide Study 1c. First, as explained above, we aim to test whether our prediction regarding the differential impacts of God and religion on mental abstraction hold also when participants are merely reminded of the religion concept rather than specific religious denominations. Second, we aim to heighten the validity of our findings by (a) priming concepts by using a relatively less supraliminal procedure and (b) measuring mindset abstraction via a different measure (i.e., breadth of object categorization; Liberman et al., 2002).

Method

Eighty Turkish college students (M_age = 21.53; 53 female) participated in the study in return for course credit. We primed religious concepts through a sentence unscrambling task. Participants were given six sets of five words and were asked to construct a meaningful four-word sentence. In other words, one word was redundant in each set. In God-concept condition, the redundant words in three sets were three Turkish words that are commonly used to refer to God (“Allah”; “Tanrı”; “Rab”). In the religion condition, we included three redundant religion-related Turkish words in three sets (“din”; “dinsel”; “dinî”).

Next, we administered an object categorization task (Liberman et al., 2002) by asking participants to categorize 35 objects. The idea behind this measure is that an abstract mindset prompts reliance on high-level, more inclusive categories; thus, those who adopt higher level of construal generate lower numbers of categories. Finally, we collected the same control and demographic measures as in Studies 1a and 1b. Forty-four participants self-identified as religious/believer, 16 as agnostic, 16 as atheistic, and four as spiritual. Of 44 believers, all the participants, except one participant who reported to be Armenian, indicated that they identify themselves with Islam. As control variables did not influence breadth of categorization, they were excluded from further analyses. Finally, participants were debriefed and thanked.

Results

We conducted an ANOVA with the number of categories as the dependent variable, and the religious concept as the independent variable. The analysis yielded a significant impact of primed concept on the number of categories generated by participants, F(1, 78) = 9.17, p = .003; CI_95% = [0.4710, 2.2791], η² = .105, d = .85. As predicted, participants who were primed with God generated a significantly lower number of categories (M = 5.45, SD = 1.89) than those who were primed with religion (M = 6.83, SD = 2.16). This shows that God prime prompts a significantly more abstract mindset than religion prime, leading people to think in terms of more inclusive, broader categories when they are reminded of God.

Discussion

Using different measures of mental construal, Studies 1a to 1c provided converging evidence that activating thoughts about God (vs. religion) results in relatively more abstract (vs. concrete) thinking. The effect replicated among samples from both predominantly Muslim and predominantly Christian populations. Moreover, the effect was obtained under different priming procedures, when both the religion concept in general and specific religious denominations were primed, and when different measures of mindset abstraction were used, heightening the validity of our prediction. Now, we turn in Studies 2 and 3 to investigating how this effect interacts with the level of abstraction associated with prosocial appeals in influencing prosociality.

Study 2

The objective of Study 2 was to empirically test the hypothesized fit effect. Specifically, we predict that activating thoughts about God (vs. religion) leads to higher compliance with prosocial requests when the receiver of the donation is framed abstractly (vs. concretely).

Method

One hundred eleven Turkish college students (M_age = 21.5 years; 77 female) participated in the study in return for course credit. The study involved a 2 (activated concept: Allah vs. Islam) × 2 (donation target description: Abstract vs. Concrete) between-subjects design. Participants were given the same manipulation as in Study 1b and wrote down the thoughts that come to their mind when they see the word “Islam” or “Allah.” Next, they were presented the description of a volunteer student club that undertakes projects for socioeconomically disadvantaged children. One of the descriptions involved more concrete statements, whereas the other description involved more abstract construal of the club:

The student club gives every weekend tutorials that supplement the school curriculum of elementary school students. The club offers classes of music and drama to elementary school students. [Concrete description condition]

The student club helps elementary school students improve their academic skills and succeed in their exams. The club contributes to artistic and creative development of elementary school students. [Abstract description condition]

Next, participants evaluated the description on three 7-point (1: “strongly disagree”; 7: “strongly agree”) statements to make sure that the framing does not change the perceived importance, informativeness, or interestingness of the student club or the message (“This description clearly explains that what this student club does is important”; “This description gives adequate information on what this student club does”; “I found this description interesting”). This evaluation was followed by the time donation intention whereby participants indicated in minutes how much time they would be willing to volunteer for the student club the coming weekend. They finally indicated their gender, age, religious self-identification, the religion with which they identify themselves with, and the strength of their belief in God. The items and scales for these variables were same as in above studies. Forty-six participants reported to be believer/religious, 34 atheistic, 19 agnostic, and 12 spiritual. All believer/religious participants were Muslim. As control variables did not significantly influence time donation intentions, they were not included in analyses.

Results

Manipulation check

There was no significant difference between the two messages in terms how informative participants thought the message was, M_abstract = 4.15, SD = 1.76; M_concrete = 4.23, SD = 1.64; F(1, 109) = 0.06, p > .8; how important they thought the student club was, M_abstract = 4.81, SD = 1.52; M_concrete = 4.74, SD = 1.58; F(1, 109) = 0.07, p > .79; and how interesting the message was, M_abstract = 3.52, SD = 1.94; M_concrete = 3.25, SD = 1.72; F(1, 109) = 0.61, p > 4.

Donation intentions

A two-way ANOVA on participants’ donation intention revealed a significant interaction of activated religious concept and description abstraction, F(1, 107) = 12.76, p < .001; CI_95% = [46.6389, 162.9809], η² = .106, d = .94. Planned contrasts yielded support for our fit predictions. When God-related thoughts were activated, participants indicated a significantly higher amount of time for volunteering when the message was abstract, M_God-abstract = 101.73, SD = 126.81, than when the message was concrete, M_God-concrete = 49.14, SD = 38.67; F(1, 107) = 7.42, p = .008; CI_95% = [14.3124, 90.8789], η² = .065, d = .77. In contrast, activating thoughts about religion resulted in higher volunteering intentions when the message was concrete, M_{religion-concrete} = 91.50, SD = 80.39; M_{religion-abstract} = 39.29, SD = 33.70; F(1, 107) = 5.58, p = .02; CI_95% = [8.4164, 96.0121], η² = .049, d = .65 (see Figure 1).

Figure 1.

Time donation intentions (in minutes) in Study 2 as a function of the activated religious concept and concreteness of the description of the student club.

Looking at these results from a different angle, the concrete description resulted in higher donation intentions following religion prime, M_{religion-concrete} = 91.50, than God prime, M_God-concrete = 49.14; F(1, 107) = 4.09, p = .046; CI_95% = [0.8458, 83.8840], η² = .037, d = .52. This pattern reversed for the abstract description, such that participants indicated higher intentions following God prime, M_God-abstract = 101.73, M_{religion-abstract} = 39.29; F(1, 107) = 9.23, p = .003; CI_95% = [21.7016, 103.1885], η² = .079, d = .85.

Discussion

Results from Study 2 provide empirical evidence for the hypothesized fit effect such that people’s compliance with a prosocial request increases when the requester is described abstractly (vs. concretely) following a God (vs. religion) reminder. Our theorizing posits that this effect occurs due to decision makers’ heightened engagement in the decision task under fit between the levels of abstraction implied by the message and the activated religious concept. We test this process prediction in Study 3.

Study 3

The primary objective of Study 3 is to test our process explanation by examining the mediating role of heightened engagement under fit. Moreover, we aim to heighten the external validity of the fit effect documented in Study 2 in two ways. First, we examine whether the effect translates into real monetary donations. Second, we investigate whether the effect obtained in Study 2 among a predominantly Muslim sample also occurs among U.S. Americans.

Method

One hundred ninety-three U.S. Americans (M_age = 37.2 years; 110 female) were recruited on Amazon Mechanical Turk for participation in the study in return for a monetary reward. The study employed a 2 (concept: God vs. Religion) × 2 (message: Abstract vs. Concrete) between-subjects design. We primed the religious concept in the same way as in Study 1a.

After the religious concept prime, participants were told that they would be given an additional reward of 50¢ to either add to their original reward for participation or donate fully or partly in the second part of the experiment. Next, participants were shown one of two messages of United Nations Children’s Fund (UNICEF) which asks for donation. One message described UNICEF concretely (“UNICEF works in more than 100 countries to distribute food to children. In 2015, UNICEF served breakfast and dinner to millions of children”) and gave a concrete donation request (“Help UNICEF serve more food to children”). The other message, on the contrary, involved an abstract description of UNICEF (“UNICEF works across the globe to address hunger of children. In 2015, UNICEF solved hunger problem of millions of children”), followed by an abstract donation request (“Help UNICEF make an impact”).

A separate pretest among a different sample from the same population (n = 85; M_age = 41.6; 48 female) asked participants to report the extent to which the message made them think about the specifics of UNICEF (vs. generalities of UNICEF) and the specifics of the campaign (vs. generalities of the campaign) on a 7-point scale (i.e., a higher score reflects higher focus on generalities; thus, higher abstraction). A one-way ANOVA on the composite score of these two items (Cronbach’s alpha = .87) yielded that the message in the abstract condition is perceived more abstractly, M = 4.80, SD = 1.92, than the message in the concrete condition, M = 3.44, SD = 1.93; F(1, 83) = 10.50, p = .002. Moreover, the pretest yielded that the two messages are perceived as equally important, M_abstract = 5.74, SD = 1.83; M_concrete = 5.26, SD = 2.34; F(1, 83) = 1.11, p > .2, and made participants feel equally confident that their contributions would be effective, M_abstract = 4.95, SD = 1.78; M_concrete = 5.21, SD = 1.79; F(1, 83) = 0.44, p > .5.

After reading the donation message, participants indicated how much of the additional 50¢ they wanted the researchers to donate to UNICEF on their behalf. Next, participants responded to the three-item “feeling right” scale (“I felt right while making the decision”; “I felt motivated while making the decision”; “I felt wrong while making the decision (reverse coded)”; Lee et al., 2010), which measures heightened engagement during decision making (Higgins, 2006). Finally, participants reported their gender and age, and were thanked. The demographic variables did not significantly influence the donation amount; thus, we did not include them in further analyses.

Results

Monetary donations

A two-way ANOVA on donation amount resulted in a significant interaction of religious concept and message concreteness, F(1, 189) = 15.91, p < .001; CI_95% = [9.0828, 26.8513], η² = .078, d = .98. We conducted planned contrast analyses to probe into this interaction. In support of our predictions, following God prime, participants donated significantly more money to the abstract request, M_God-abstract = 20.40, SD = 15.49; M_God-concrete = 13.39, SD = 14.50; F(1, 189) = 4.37, p = .038; CI_95% = [0.3947, 13.6248], η² = .023, d = .55. As shown in Figure 2, this effect reversed when participants were primed with religion, M_{religion-abstract} = 11.08, SD = 12.24; M_{religion-concrete} = 22.04, SD = 18.65; F(1, 189) = 13.28, p < .001; CI_95% = [5.0268, 16.8878], η² = .066, d = .95.

Figure 2.

Monetary donations (in U.S. cents) in Study 3 for abstract and concrete UNICEF appeals following religion and God primes.

From a different point of view, the concrete appeal resulted in higher donations following religion prime, M_{religion-concrete} = 22.04, than God prime, M_God-concrete = 13.39; F(1, 189) = 7.33, p = .007; CI_95% = [2.3478, 14.9432], η² = .037, d = .77. In contrast, when the donation appeal was abstract, participants who were primed with God donated significantly more, M_God-abstract = 20.40, than participants who were initially primed with religion, M_{religion-abstract} = 11.08; F(1, 189) = 8.61, p = .004; CI_95% = [3.0551, 15.5881], η² = .044, d = .83.

Moderated mediation analysis

We computed a composite score of feeling right by averaging the three items (Cronbach’s alpha = .74). Next, we estimated coefficients based on 10,000 bootstrapped samples for a moderated mediation model using PROCESS macro (Model 7; Hayes, 2017). The model included primed concept as the independent variable (1: religion; 2: God), the message abstraction as the moderator (1: concrete; 2: abstract), and the composite “feeling right” score as the process variable. The predicted moderated mediation model was significant, as demonstrated by a significant index of moderated mediation (β = 6.12, SE = 2.15; CI_95% = [2.2594, 10.7267]). More importantly, “feeling right” mediated the effects of primed religious concept on monetary donations in both concrete (β = −2.86, SE = 1.49; CI_95% = [−6.0163, −0.1723]) and abstract (β = 3.26, SE = 1.32; CI_95% = [0.8600, 6.1013]) message conditions.

Discussion

Study 3 showed that people’s actual monetary donations increase when they are presented an abstractly (vs. concretely) framed donation message following a God (vs. religion) prime. As hypothesized, this is a construal fit effect such that God (vs. religion) prime results in a heightened subjective engagement in the decision task when the request is framed abstractly (vs. concretely), and this heightened engagement makes people feel more right about engaging in prosocial behavior. Consequently, this feeling right experience leads to higher actual monetary donations.

Study 4

In Study 4, we examine whether the obtained effect occurs in the field. To this aim, we obtained donation data from four mosques in a small town in Turkey. In Turkey, religious sermons and activities are strictly regulated by the government. Hence, preachers of all mosques are given the same Friday sermon and the target of donation (if a donation is to be requested from prayers on that specific week) by governmental authorities. The preacher’s responsibilities include counting the donation collected and arranging a donation receipt that shows both the target of donation (e.g., another mosque in the neighborhood, another mosque in a different neighborhood of the same city, disaster victims in a different city, and victims of a war in a different country) and the donation amount. The preachers are legally responsible for keeping copies of these receipts for a specified amount of time, and our dataset involved 60 such receipts from each of the four mosques that were available at the time of data collection. The dataset covers a time range between November 2013 and October 2015.

Method

We first obtained each week’s Friday sermon from a governmental website. Two experts, who received formal theology education, and one novice, who lacks formal theology education but who is a practicing Muslim, rated each week’s sermon on a 5-point scale (1: “This sermon activates predominantly religion-specific thoughts”; 5: “This sermon activates predominantly God-related thoughts”). We were interested in how the level of abstraction induced by the sermon would interact with the abstraction level at which the donation target is construed to influence the amount of monetary donations. Based on past research showing a strong bidirectional link between spatial distance and abstract construal (Fujita et al., 2006), we operationalized target abstraction through the actual distance between the mosque and the target. We measured this distance using Google Maps.

We controlled for three factors that might have influenced the donation amount. First, we added a covariate that shows whether there was an important political or social event during the week before the Friday prayer. We identified nine significant such events such as terrorist attacks, political crisis, or a natural disaster. Second, we controlled for whether the prayer took place during Ramadan or not. The third covariate accounted for variation across four mosques. Only the mosque variable significantly influenced the donation amount; thus, we kept it as a covariate in further analyses.

Results

The coding of concept activation (God vs. religion) by three independent coders resulted in a high reliability (Cronbach’s alpha = .86), and we computed a composite concept score by averaging the three coders’ ratings. A lower (vs. higher) score in this variable reflects activation of predominantly religion-specific (vs. God-specific) thoughts. Note that ratings from the two expert coders also resulted in a high reliability (Cronbach’s alpha = .80), and the significance of the results summarized below do not change when the composite score is computed by using only the experts’ coding. For parsimony, we report only the results when the concept variable is computed by using all coders’ coding.

To meet linearity assumptions, we first log-transformed the spatial distance between the target and the mosque (in kilometers), and the donation amount. Next, we analyzed the interactive effect of religious concept and spatial distance on the donation amount based on coefficient estimates from 10,000 bootstrapped samples (version 3.1; Model 1; Hayes, 2017). The model yielded a significant interaction, β = .05, F(1, 235) = 15.01, p < .001; CI_95% = [0.0247, 0.0758].

Spotlight analyses supported our predictions. At 1 SD below the mean activated concept score (i.e., when the sermon activated predominantly religion-specific thoughts), the impact of spatial distance on donation was negative and significant (β = −.05, p = .006; CI_95% = [−0.0833, −0.0141]). This suggests that, when the sermon primed prayers with religion-specific thoughts, prayers donated significantly more to spatially closer targets than distant targets. In contrast, the coefficient was positive and significant (β = .06, p = .005; CI_95% = [0.0173, 0.0937) when the sermon activated relatively God-specific thoughts (i.e., at 1 SD above the mean concept score), suggesting that donations increased significantly after God prime as the distance of the donor increased (see Figure 3).

Figure 3.

Average monetary donations (in local currency) after Friday sermons in Study 4.

In real monetary terms, the prayers’ actual monetary donation after a relatively religion-specific sermon was 208 (in local currency) for a spatially distant target (i.e., for the target which was located at 1 SD above the mean distance of all targets). This amount significantly increased to 261. In contrast, a relatively God-specific sermon resulted in a donation of 192 for spatially closer targets, which significantly increased to 249 when the donor was spatially distant.

Discussion

The data we analyzed in Study 4 provided further support for our theorizing by showing that a religious sermon that activates predominantly religion-specific (vs. God-related) thoughts results in significantly higher donation to spatially close (vs. distant) targets, who are construed relatively concretely (vs. abstractly). This finding heightens the generalizability of the fit effect obtained in Studies 2 and 3 by showing the effect in the field, and when a different factor other than message framing—that is, psychological distance—is salient in the decision context.

General Discussion

The present research provides one of the initial scholarly attempts to investigate the possibly differential cognitive consequences of the two most prevalent religious concepts: God and religion. We showed in three initial studies (Studies 1a–1c) that activating God-specific (vs. religion-specific) thoughts leads to relatively higher (vs. lower) levels of mental construal. In three additional studies, we documented a significant downstream consequence of this finding. Study 2 showed that God (vs. religion) priming increases prosocial intentions toward abstractly (vs. concretely) framed targets. Study 3 extended this finding in two main ways. First, it showed that this effect spills over to actual behavior, such that people donate more money to an abstractly (vs. concretely) framed donation appeal after thoughts about God (vs. religion) are activated. Second, it provided empirical evidence for the underlying mechanism that the effect is driven by people’s “feeling right” experience, which occurs under construal fit. Finally, Study 4 presented results from field data that involved 60-week donation figures from four mosques in Turkey. The results indicated that worshippers gave more to spatially distant (vs. close) targets after hearing a sermon that activated relatively God-specific (vs. religion-specific) thoughts. Given that spatially distant targets are construed more abstractly than spatially closer targets (Fujita et al., 2006), this finding provides further support for our conceptualization.

The construal level account of the cognition of God and religion developed in the current research has important theoretical contributions. First, this study is one of the initial studies that has attempted to differentiate between behavioral consequences of the two most prevalent religious concepts (see also Batara et al., 2016; Preston & Ritter, 2013; Ramsay et al., 2016). Despite the Ritter and Preston’s (2013) argument that all religious concepts should not be treated equivalently, research in the domain of religious cognition has treated “religion” as an umbrella term, and it has used both God-specific and religion-specific concepts to prime religion (e.g., M. K. Johnson et al., 2010; Pichon et al., 2007; Shariff & Norenzayan, 2007). Only a limited number of studies have attempted to distinguish between religion and God as two constructs with distinct psychological impacts. One of these studies found that reminding people of God (vs. religious affiliation) increased prosocial intentions to outgroup (vs. ingroup) members (Preston & Ritter, 2013). This research proposed a motivational explanation for this pattern of findings: People are more motivated for impression management when God is salient, and for self and group protection when religion is salient. We contributed to this stream of research by showing that God and religion are also associated with distinct cognitive corollaries as evinced by different levels of mindset abstraction that people exhibit when they are reminded of God versus religion.

The framework that we developed in the current research does not contradict the motivational account developed by Preston and Ritter (2013). One type of psychological distance that influences the level of mental construal is social distance (Bar-Anan et al., 2006). People tend to perceive socially distant others more abstractly than socially closer others. This finding on the relationship between social distance and mental abstraction along with our construal level–based mechanism could also corroborate the findings of Preston and Ritter (2013): Prosocial attitudes toward outgroups (vs. ingroups) who are construed relatively more abstractly (vs. concretely) might have intensified after reminders of God (vs. religion) due to the construal level fit. Stated differently, God and religion might have distinct motivational and cognitive impacts, leading to similar outcomes for prosocial behavior toward close versus distant others. The difference between the two models, however, is that the cognitive model that we developed in the current work allows for further predictions based on the level of abstraction in the decision context.

Second, the present work contributes to the broad literature on religion—prosociality link by introducing a novel moderator: the specific religious concept that is activated. As past research has documented both prosocial (Ahmed & Salas, 2011; Pichon et al., 2007; Randolph-Seng & Nielsen, 2007 ; Shariff & Norenzayan, 2007) and anti-social (M. K. Johnson et al., 2010; McKay et al., 2011) effects of religion, this moderator has been proposed to reconcile these contradictory findings by asserting that prosociality is limited to conditions related to the context or the type of religiosity (Saroglou, 2006). This suggests that the question of whether religion promotes prosociality does not have a straightforward answer; rather, several factors moderate the impact of religion on prosocial responses. This possibility has recently motivated several studies in this domain. In one investigation, for example, Stavrova and Siegers (2014) identified social enforcement of religion as a factor that influences the scope of the impact of religion on prosociality. We advance the extant knowledge by suggesting that the specific religious concept being activated—as well as its impact on human cognition—is another moderator that interacts with the cognitive characteristics of the context in determining prosociality.

By so doing, we help explain the general pattern of past findings on the impact of religious primes on prosociality. Although these studies did not distinguish between specific religious concepts and used qualitatively distinct concepts in their priming procedures, a closer examination reveals a similar pattern: In studies, where religious priming involved predominantly God-related (vs. religion-related) concepts, religious priming has been found to result in higher prosocial intentions when other factors in the decision context implied higher (vs. lower) abstraction. For example, in their well-known paper, Shariff and Norenzayan (2007) used predominantly God-related words to prime religion. Their priming procedure involved predominantly God-related words (i.e., God, divine, spirit, and sacred) with the exception of a single religion-specific word (i.e., religion), and this priming led to higher cooperation with an anonymous other. Anonymity causes a lack of detail and information; hence, it implies a big-picture orientation as opposed to a detailed and direct experience, which implies a low-level construal. The abstraction implied by the form of the target (i.e., anonymity) might have contributed to higher prosocial intentions after this prime on top of other motivational effects of the prime. In another study, Pichon and Saroglou (2009) investigated the causal role of religion on helping close and distant targets. Either a close (i.e., local, homeless) or distant (i.e., foreigner) target was depicted in front of a religious or neutral building. The close target was helped more when depicted in front of the religious building. Note again that the manipulation is a religion-specific manipulation (i.e., a church), and it leads to a greater help behavior for the relatively more concretely construed target (i.e., a local). Finally, Sasaki et al. (2013) used the same God-dominated priming procedure as Shariff and Norenzayan (2007) and found a greater tendency to volunteer in prosocial causes. In this study, the authors used abstract and more general descriptions for causes such as “promoting energy efficiency” to “tap into prosocial behavior toward society in general rather than a specific person or group” (Sasaki et al., 2013, p. 211). Our framework would suggest that this effect might also be partly driven by the fit between abstract construal of the target behavior and the abstraction implied by the God construct.

The construal level account that we developed in the current research opens up several opportunities for future research. One of the basic tenets of CLT is that different types of psychological distance exhibit a correspondence (Trope & Liberman, 2010). In other words, a relationship holds true for a type of psychological distance if it exists for any other type. In the present research, we tested the role of spatial distance as a correlate of mindset abstraction to test our fit prediction. However, it is likely that the same fit effect also exists for other types of distances. To illustrate, a God prime might result in higher prosocial intentions for a cause that takes place in the distant future, whereas activating thoughts about religion might increase prosociality toward a cause that will take place in the near future. Future research could investigate similar novel predictions that derive from our theorizing.

Substantively, the model developed in this research has significant implications for fundraising. Note that the effect that we show relates to human cognition in general, as God and religion influence the abstraction level of human thinking in general rather than influencing only how abstractly people think of religion-related issues. An expected consequence of this is that the fit effect is not limited to religious messages but applies to all messages that involve differential levels of abstraction. Note also that we demonstrated the effect for nonreligious targets such as a secular student club or UNICEF. Hence, both religious and secular charity organizations could utilize religion and God in their fundraising campaigns to increase donations by simply matching the abstraction level of the donation request with the specific religious concept that they remind potential donors of.

One limitation of the present research is that we replicated these effects among populations who follow the two largest Abrahamic religions. However, there are several differences between Abrahamic teachings and Eastern teachings. Our conceptualization refers to several characteristics of Abrahamic religions, such as religion involving concrete rites and rituals and God as a moralizing agent with supernatural qualities. Further studies could test the proposed framework among non-Abrahamic samples to help develop a complete understanding of how God and religion influence cognition cross-culturally.

Supplemental Material

Karatas_Online_Appendix – Supplemental material for A Construal Level Account of the Impact of Religion and God on Prosociality

Supplemental material, Karatas_Online_Appendix for A Construal Level Account of the Impact of Religion and God on Prosociality by Mustafa Karataş and Zeynep Gürhan-Canli in Personality and Social Psychology Bulletin

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Migros Chair Funds.

ORCID iD

Mustafa Karataş

Supplemental Material

Supplemental material is available online with this article.

References

Ahmed

Hammarstedt

(2011). The effect of subtle religious representations on cooperation. International Journal of Social-Economics, 38, 900–910.

Ahmed

Salas

(2013). Religious context and prosociality: An experimental study from Valparaíso, Chile. Journal for the Scientific Study of Religion, 52, 627–637.

Ahmed

A. M.

Salas

(2011). Implicit influences of Christian religious representations on dictator and prisoner’s dilemma game decisions. The Journal of Socio-Economics, 40, 242–246.

Bar-Anan

Liberman

Trope

(2006). The association between psychological distance and construal level: Evidence from an implicit association test. Journal of Experimental Psychology: General, 135, 609–622.

Baron

(1994). Nonconsequentialist decisions. Behavioral and Brain Sciences, 17, 1–10.

Batara

J. B. L.

Franco

P. S.

Quiachon

M. A. M.

Sembrero

D. R. M.

(2016). Effects of religious priming concepts on prosocial behavior towards ingroup and outgroup. Europe’s Journal of Psychology, 12, 635–644.

Batson

C. D.

Oleson

K. C.

Weeks

J. L.

Healy

S. P.

Reeves

P. J.

Jennings

Brown

(1989). Religious prosocial motivation: Is it altruistic or egoistic? Journal of Personality and Social Psychology, 57, 873–884.

Batson

C. D.

Schoenrade

Ventis

W. L.

(1993). Religion and the individual: A social-psychological perspective. Oxford University Press.

Bekkers

Schuyt

(2008). And who is your neighbor? Explaining denominational differences in charitable giving and volunteering in the Netherlands. Review of Religious Research, 50, 74–96.

10.

Bellah

R. N.

(1964). Religious evolution. American Sociological Review, 29, 358–374.

11.

Boyer

(2001). Religion explained: The evolutionary origins of religious thought. Basic Books.

12.

Brooks

A. C.

(2006). Who really cares: The surprising truth about compassionate conservatism. Basic Books.

13.

Burkert

(1996). Creation of the sacred: Tracks of biology in early religions. Harvard University Press.

14.

Carroll

Dudley

McKinney

(1986). Handbook for congregational studies. Abingdon Press.

15.

Cesario

Higgins

E. T.

(2008). Making message recipients “feel right”: How nonverbal cues can increase persuasion. Psychological Science, 19, 415–420.

16.

Douglas

K. M.

Sutton

R. M.

(2003). Effects of communication goals and expectancies on language abstraction. Journal of Personality and Social Psychology, 84, 682–696.

17.

Durkheim

(1995). The elementary forms of the religious life. Free Press. (Original work published 1912)

18.

Eyal

Liberman

Trope

(2008). Judging near and distant virtue and vice. Journal of Experimental Social Psychology, 44, 1204–1209.

19.

Freitas

A. L.

Gollwitzer

Trope

(2004). The influence of abstract and concrete mindsets on anticipating and guiding others’ self-regulatory efforts. Journal of Experimental Social Psychology, 40, 739–752.

20.

Fujita

Henderson

M. D.

Eng

Trope

Liberman

(2006). Spatial distance and mental construal of social events. Psychological Science, 17, 278–282.

21.

Galen

L. W.

(2012). Does religious belief promote prosociality? A critical examination. Psychological Bulletin, 138, 876–906.

22.

Gervais

W. M.

Norenzayan

(2012). Like a camera in the sky? Thinking about God increases public self-awareness and socially desirable responding. Journal of Experimental Social Psychology, 48, 298–302.

23.

Gray

Wegner

D. M.

(2010). Blaming God for our pain: Human suffering and the divine mind. Personality and Social Psychology Review, 14, 7–16.

24.

Guttmann

(1984). Cognitive morality and cheating behavior in religious and secular school children. The Journal of Educational Research, 77, 249–254.

25.

Haidt

(2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108, 814–834.

26.

Hayes

A. F.

(2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (2nd ed.). Guilford Press.

27.

Higgins

E. T.

(2006). Value from hedonic experience and engagement. Psychological Review, 113, 439–460.

28.

Johnson

Kruger

(2004). The good of wrath: Supernatural punishment and the evolution of cooperation. Political Theology, 5, 159–176.

29.

Johnson

M. K.

Rowatt

W. C.

LaBouff

(2010). Priming Christian religious concepts increases racial prejudice. Social Psychological and Personality Science, 1, 119–126.

30.

Kim

Rao

A. R.

Lee

A. Y.

(2008). It’s time to vote: The effect of matching message orientation and temporal frame on political persuasion. Journal of Consumer Research, 35, 877–889.

31.

Laurin

Kay

A. C.

Fitzsimons

G. M.

(2012). Divergent effects of activating thoughts of God on self-regulation. Journal of Personality and Social Psychology, 102, 4–18.

32.

Lee

A. Y.

Keller

P. A.

Sternthal

(2010). Value from regulatory construal fit: The persuasive impact of fit between consumer goals and message concreteness. Journal of Consumer Research, 36, 735–747.

33.

Liberman

Sagristano

M. D.

Trope

(2002). The effect of temporal distance on level of mental construal. Journal of Experimental Social Psychology, 38, 523–534.

34.

Liberman

Trope

(2008). The psychology of transcending the here and now. Science, 322, 1201–1205.

35.

Liberman

Trope

McCrea

S. M.

Sherman

S. J.

(2007). The effect of level of construal on the temporal distance of activity enactment. Journal of Experimental Social Psychology, 43, 143–149.

36.

McCullough

M. E.

Worthington

E. L.

(1999). Religion and the forgiving personality. Journal of Personality, 67, 1141–1164.

37.

McKay

Efferson

Whitehouse

Fehr

(2011). Wrath of God: Religious primes and punishment. Proceedings of the Royal Society of London B: Biological Sciences, 278, 1858–1863.

38.

Medin

Ortony

(1989). Psychological essentialism. In Vosniadou

Ortony

(Eds.), Similarity and analogical reasoning (pp. 179–195). Cambridge University Press.

39.

Menegatti

Rubini

(2013). Convincing similar and dissimilar others: The power of language abstraction in political communication. Personality and Social Psychology Bulletin, 39, 596–607.

40.

Morgan

S. P.

(1983). A research note on religion and morality: Are religious people nice people? Social Forces, 61, 683–692.

41.

Nowell

Laufer

(1997). Undergraduate student cheating in the fields of business and economics. The Journal of Economic Education, 28, 3–12.

42.

Pergament

K. I.

(1997). The psychology of religion and coping: Theory, research, practice. Guilford Press.

43.

Pichon

Boccato

Saroglou

(2007). Nonconscious influences of religion on prosociality: A priming study. European Journal of Social Psychology, 37, 1032–1045.

44.

Pichon

Saroglou

(2009). Religion and helping: Impact of target thinking styles and just-world beliefs. Archive for the Psychology of Religion, 31, 215–236.

45.

Preston

J. L.

Ritter

R. S.

(2013). Different effects of religion and God on prosociality with the ingroup and outgroup. Personality and Social Psychology Bulletin, 39, 1471–1483.

46.

Preston

J. L.

Ritter

R. S.

Hernandez

J. I.

(2010). Principles of religious prosociality: A review and reformulation. Social and Personality Psychology Compass, 4, 574–590.

47.

Ramsay

J. E.

Tong

E. M.

Pang

J. S.

Chowdhury

(2016). A puzzle unsolved: Failure to observe different effects of God and religion primes on intergroup attitudes. PLOS ONE, 11, e0147178.

48.

Randolph-Seng

Nielsen

M. E.

(2007). Honesty: One effect of primed religious representations. The International Journal for the Psychology of Religion, 17, 303–315.

49.

Ritter

R. S.

Preston

J. L.

(2013). Representations of religious words: Insights for religious priming research. Journal for the Scientific Study of Religion, 52, 494–507.

50.

Saroglou

(2002). Religion and the five factors of personality: A meta-analytic review. Personality and Individual Differences, 32, 15–25.

51.

Saroglou

(2006). Religion’s role in prosocial behavior: Myth or reality? Religion, 31, 1–66.

52.

Sasaki

J. Y.

Kim

H. S.

Mojaverian

Kelley

L. D.

Park

I. Y.

Janušonis

(2013). Religion priming differentially increases prosocial behavior among variants of the dopamine D4 receptor (DRD4) gene. Social Cognitive and Affective Neuroscience, 8, 209–215.

53.

Schweiker

W. F.

(1969). Religion as a superordinate meaning system and socio-psychological integration. Journal for the Scientific Study of Religion, 8, 300–307.

54.

Semin

G. R.

Fiedler

(1988). The cognitive functions of linguistic categories in describing persons: Social cognition and language. Journal of Personality and Social Psychology, 54, 558–568.

55.

Shariff

A. F.

Norenzayan

(2007). God is watching you: Priming God concepts increases prosocial behavior in an anonymous economic game. Psychological Science, 18, 803–809.

56.

Shariff

A. F.

Willard

A. K.

Andersen

Norenzayan

(2016). Religious priming: A meta-analysis with a focus on prosociality. Personality and Social Psychology Review, 20, 27–48.

57.

Smith

R. E.

Wheeler

Diener

(1975). Faith without works: Jesus people, resistance to temptation, and altruism. Journal of Applied Social Psychology, 5, 320–330.

58.

Sosis

Ruffle

B. J.

(2004). Ideology, religion, and the evolution of cooperation. Socioeconomic Aspects of Human Behavioral Ecology, 23, 89–117.

59.

Stavrova

Siegers

(2014). Religious prosociality and morality across cultures: How social enforcement of religion shapes the effects of personal religiosity on prosocial and moral attitudes and behaviors. Personality and Social Psychology Bulletin, 40, 315–333.

60.

Stephan

Liberman

Trope

(2010). Politeness and psychological distance: A construal level perspective. Journal of Personality and Social Psychology, 98, 268–280.

61.

Stephan

Liberman

Trope

(2011). The effects of time perspective and level of construal on social distance. Journal of Experimental Social Psychology, 47, 397–402.

62.

Sunstein

C. R.

(2005). Moral heuristics. Behavioral and Brain Sciences, 28, 531–541.

63.

Trope

Liberman

(2003). Temporal construal. Psychological Review, 110, 403–421.

64.

Trope

Liberman

(2010). Construal-level theory of psychological distance. Psychological Review, 117, 440–463.

65.

Vallacher

R. R.

Wegner

D. M.

(1987). What do people think they’re doing? Action identification and human behavior. Psychological Review, 94, 3–15.

66.

Vallacher

R. R.

Wegner

D. M.

(1989). Levels of personal agency: Individual variation in action identification. Journal of Personality and Social Psychology, 57, 660–671.

67.

Vallacher

R. R.

Wegner

D. M.

McMahan

S. C.

Cotter

Larsen

K. A.

(1992). On winning friends and influencing people: Action identification and self-presentation success. Social Cognition, 10, 335–355.

68.

Yan

Sengupta

Hong

(2016). Why does psychological distance influence construal level? The role of processing mode. Journal of Consumer Research, 43, 598–613.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB