Thanks Be to God: Divine Gratitude and Its Relationship to Well-Being

Abstract

Gratitude is a pivotal concept in the psychology of religion because it is one of the most frequently experienced emotions toward God. The authors developed a trait measure of Gratitude to God (GTG-T) and investigated the association between Gratitude to God and psychological and spiritual well-being in four studies. Study 1 provided evidence for the validity of the GTG-T as Gratitude to God was associated with overall well-being and predicted increases in spiritual well-being over time. Study 2 used a prospective design to show that Gratitude to God predicted increases in general gratitude. In a third study, results showed that Gratitude to God predicted increased grateful emotion and increased confidence in the existence of God over time. In Studies 2 and 3, mediation analyses supported the model that Gratitude to God enhances well-being by promoting generalized gratitude. Study 4 provided further valid evidence for the GTG-T and also showed that gratitude to God is much more strongly related to one’s heartfelt conception of God as loving than to their doctrinal conception of a loving God. Gratitude to God deserves additional scientific scrutiny in the psychology of religion and spirituality than it has received to date.

1. Introduction

Gratitude and spirituality seem to go hand-in-hand. In a recent survey, of those students who believe in God, fully 100% said that God was a major source of benefits in their lives (Scheibe et al. 2017). Because most people believe in God (Pew Research Center 2018), investigations into God as a benefactor and gratitude to God as an emotional response to God’s benefaction are needed; research on gratitude in human social interactions has accelerated in the past decade (Algoe 2012), but until recently very little research has been reported on gratitude to God, despite the central role that recognizing God’s provision of benefits plays in all major world religions (Carman and Streng 1989). Across theistic traditions, God is viewed as the source of all good, and this realization highlights the priority of divine goodness over every other created good. Although there has been a recent push for increased work on gratitude to God (as indicated by two special issues on the subject), the current scientific literature on gratitude toward God is limited.

The purpose of the studies described in this paper is twofold: (1) to develop a trait measure of gratitude to God, and (2) to investigate the connections between gratitude to God, general gratitude, and psychological and spiritual well-being. For the purposes of this paper, general gratitude is defined as gratitude that is experienced where God is not necessarily the benefactor. Trait gratitude to God is defined as one’s disposition for experiencing the state of gratitude to God; i.e., people high in trait gratitude to God should experience state gratitude to God frequently across a wide variety of situations.

1.1. The Importance of a Divine Benefactor

A number of studies have found that religious people tend to be more grateful (Emmons and Kneezel 2005; McCullough et al. 2002). Indeed, McCullough et al. (2002) found that people who are more religiously involved tend to be more grateful. Religious persons are more likely to attribute positive events that are not easily attributed to humans to supernatural agency (e.g., Hovemyr 1998; Lupfer et al. 1996). Thus, religious people have a divine benefactor to thank in situations where non-believers would not. For example, religious people may be more likely to attribute beautiful sunsets, sublime mountain views, or their flight landing safely in the midst of a storm “as occasions for gratitude to God” (McCullough et al. 2002, p. 114). Because positive events can be interpreted as gifted actions of a supernatural agent, belief in a divine benefactor may promote gratitude. Thus, gratitude to God may cultivate general gratitude. Because gratitude is a strong predictor of well-being (Watkins 2014), and a large majority of the population believes that God is a significant source of benefits for them (Pew Research Center 2018), it follows that it is essential to investigate the psychological causes and consequences of gratitude to God.

1.2. Prior Research on Gratitude to God

Preliminary evidence suggests that gratitude to God is a common spiritual emotion. Samuels and Lester (1985) investigated the frequency with which nuns and priests experienced 50 different emotions one might experience toward God. Gratitude to God was the second most experienced emotion, followed only by love toward God. Perhaps surprisingly, Chun and Lee (2013) found that gratitude to God was a distinct emotional theme experienced by all of their spinal cord injury victims who were Christians. Similarly, when people see difficult events as part of a divine plan to help them grow, they are more likely to experience gratitude to God (Pargament et al. 2000). Thus, early evidence suggested that gratitude to God is a common emotional experience for religious individuals, even in the wake of stressful events.

In the majority of research conducted to date on gratitude to God, it is most often an independent or predictor variable with various indicators of flourishing, well-being, gratitude, religiousness, and spirituality serving as dependent or criterion variables (e.g., Rosmarin et al. 2011). The underlying rationale is that because gratitude is encouraged in every major religious tradition, participation in activities such as prayer, meditation or worship should increase gratitude to God, and then gratitude to God would drive various beneficial outcomes in psychological, physical, relational, and spiritual domains of functioning. The perception of receiving positive benefits from God should consistently lead to more general gratitude. Because gratitude is the quintessential positive trait, gratitude to God becomes a distal causal variable in the immediate benefits of grateful living. For instance, (Krause and Ellison 2009, see also Krause et al. 2015) have shown several religious variables are linked to greater gratitude to God, including religious attendance, perception of social support at church, and congregational emotional support. Rosmarin et al. (2011) found gratitude to God predicted psychological functioning above and beyond general gratitude. Other research has examined the experience of gratitude across various religious traditions both in terms of actual and desired levels of the trait (Kim-Prieto and Diener 2009). More recent research has supported these conclusions and has also shown that gratitude to God is unique from gratitude to humans and domain-general gratitude (Hodge et al. 2024; Park et al. 2022; Tsang et al. 2023; Watkins et al. 2022; Zuniga et al. 2024).

A related construct, indebtedness to God or transcendent indebtedness, has recently emerged as an important variable related to gratitude to God (Nelson et al. 2023). Although strongly related to gratitude to God, indebtedness to God appears to be a distinct construct. What seems to be unique to transcendent indebtedness is that it is more positively associated with salubrious aspects than is trait indebtedness to humans. Trait indebtedness to humans shows very little relationship to generalized gratitude and is even negatively related to subjective well-being, whereas this does not seem to be the case with indebtedness toward God.

1.3. Gratitude to God and Coping

Viewing gratitude as an expression culture, often reinforced by a particular religious and spiritual meaning or practices that promote coping, meets the criteria for the three moral functions of gratitude described by McCullough et al. (2001). First, the gratitude expressed in religious and spiritual practices signals the importance and value of one’s relationship with God as a moral barometer. Second, the moral reinforcing function serves to increase the subjective likelihood of receiving God’s gifts in the future. Third, the moral motive function of gratitude serves to ignite compassionate and kind behavior toward others. Each of these moral functions is associated with psychological well-being and may be the driving force behind grateful coping practices that confer protection from psychological distress (Wood et al. 2007). In addition to invoking God or a Higher Power for assistance or direction, a common aspect of these coping behaviors includes the expression of Gratitude to God for blessings despite or even in the face of significant hardship. We define gratitude as the cognitive and affective response to the recognition that one has received gifts and benefits from another. It is an affirmation of goodness in conjunction with recognizing that the good is sourced outside of oneself. When grateful, a person responds to a benefit by avowing its goodness, apprehending it as flowing from a source beyond his or her own agency, and desiring to reciprocate the goodness received. This definition fits the descriptive nature of participants who describe gifts of grace, life, or even seemingly mundane experiences such as the opportunity for moments of peace and connection. A qualitative study on feelings of gratitude to God (Krause et al. 2014) found that the relationship between feelings of gratitude to God and health may depend upon whether people view God as all-powerful and benevolent or all-powerful and punishing. Study participants who felt that God was benevolent reported that when they felt grateful, they developed closer relationships with God. This suggests a model in which feelings of gratitude to God influence feelings of closeness to God, and developing a close relationship with God is, in turn, associated with better health. In contrast, participants who feel that God is punishing may be more likely to have a troubled relationship with God. As research on spiritual struggles reveals (Exline 2013), people who have a troubled relationship with God are more likely to experience psychological distress, and this, in turn, impacts their feelings of gratitude to God. Thus, one goal of the studies described here is to investigate how a benevolent view of God impacts gratitude to God.

Krause and colleagues have shown that gratitude to God buffers the impact of stress in elders (Krause 2006), predicts greater hope (Krause et al. 2015), predicts better health and lower depression (Krause et al. 2014), and better physical health in women (Krause et al. 2017). Although these findings support the theory that gratitude to God promotes well-being, the cross-sectional correlational nature of these findings is inconclusive. However, Al-Seheel and Noor (2016) found that the intervention of Islamic-based gratitude to God was more effective at enhancing happiness in Muslim participants than a secular gratitude treatment. In sum, there is sufficient reason to suspect that gratitude to God contributes to the well-being of religious individuals.

Although gratitude to God shows small but consistent relationships with general well-being (Rosmarin et al. 2011), gratitude to God might be even more central to spiritual well-being. How might gratitude to God be important to spiritual well-being? A secure, close relationship with God is likely to amplify the spiritual well-being of believers, and gratitude to God may be part of a secure relationship with God. A number of studies have shown that gratitude is central to the development and maintenance of human relationships (Algoe 2012). Thus, gratitude to God may promote one’s relationship with God (Fredrickson 2004). Gratitude has been shown to encourage approach behavior toward one’s benefactor (Algoe 2012; Watkins et al. 2006). Thus, it follows that gratitude to God should promote approach behavior to God. Furthermore, gratitude has been shown to enhance meaning in life (Watkins et al. 2019). Because gratitude to God is more likely to be related to transcendent meaning, it seems likely that when one experiences gratitude to God, a sense of meaning and coherence should result. Because of an increase in transcendent meaning, gratitude to God might even contribute to an increase in one’s confidence in the existence of God and help resolve religious doubt, particularly doubt about the existence of God and God’s loving character. Whereas many argue that religiousness promotes gratitude, we propose that the reverse explanation is also possible: gratitude to God promotes religiousness. We hope that our data can help adjudicate the nature of this relationship. In sum, our theoretical analysis suggests that gratitude to God might have a particularly strong relationship to spiritual well-being because it promotes a secure relationship with God by promoting approach to God (or conversely, a lack of avoidance), it promotes transcendent meaning, and it promotes confidence in the existence of a loving God.

As the above studies by Krause suggest, gratitude to God should also support general well-being. We propose that gratitude to God may enhance subjective well-being by increasing general gratitude. As pointed out by McCullough et al. (2002), belief in God as a good benefactor provides an advantage for those who are predisposed to gratitude to God because they have a benefactor to thank for all benefits, regardless of whether the benefit was provided by a human benefactor. Thus, we propose that one mechanism for the relationship between gratitude to God and general well-being is that it enhances general gratitude, which in turn supports subjective well-being. We tested this theory in mediational analyses in Studies 2 and 3.

1.4. Gratitude to God as a Moral Motivator

What motivates religious behavior? What motivates activities that are key to the spiritual lives of religious people, such as prayer, worship, giving, and forgiving? This would seem to be a significant question for religious believers and for the psychology of religion. Various theologians have argued that gratitude to God is a spiritual motivator. For example, Karl Barth—perhaps the most influential theologian of the 20th Century—argued that “If the essence of God as the God of man is His grace, then the essence of men as his people, that which is proper to and demanded of them in covenant with God, is simply their thanks” (Barth [1956] 1961, p. 42). Similarly, contemporary popular theologian Timothy Keller wrote, “This is a summary of the whole of the Christian life. The believer’s life is to be lived out of gratitude” (Keller 2015, p. 8). In their seminal review of gratitude, McCullough et al. (2001) argued that gratitude was essentially a “moral affect”. As a “moral motivator”, gratitude inspires the person to engage in moral behavior. Since the time of that review, considerable evidence has supported the theory that gratitude motivates prosocial behavior (e.g., Bartlett and DeSteno 2006; Bartlett et al. 2012; DeSteno et al. 2010; Watkins et al. 2006). If general gratitude motivates moral behavior, it would seem to follow that gratitude to God motivates religious behavior. When one feels grateful for divine benefits, this should motivate behavior that is seen to be central to following God—i.e., religious behavior. In this sense, gratitude to God might motivate religious behavior more out of desire than a sense of obligation. Thus, gratitude to God might motivate religious behavior more because one wants to follow religious directives rather than feeling that one is compelled to follow these instructions, and this, in turn, is likely to support spiritual well-being. In sum, research into gratitude to God is vital because it is a frequent spiritual emotion, it might be central to the well-being of believers, it may be a significant spiritual motivator, and it is likely to foster the spiritual well-being of religious individuals.

1.5. The Need for Gratitude to God Measures

Although self-report measures of gratitude to God exist (e.g., Rosmarin et al. 2011), in these studies, we sought to develop a unidimensional, theory-driven measure of trait gratitude to God that relies on what we know about secular gratitude from the literature. First and foremost, we wanted to develop a unidimensional measure of gratitude to God because we believe that the construct is indeed unidimensional in reality. Of course, in the end, this is an empirical question, but to anticipate, we believe that the data in this paper bear out this proposal. Second, unidimensional constructs are easier to understand than multidimensional constructs. For example, if gratitude to God is actually multidimensional when gratitude to God correlates with another construct, we cannot be sure what aspects of gratitude to God are actually contributing to the construct. Finally, we felt it important to develop a unidimensional measure of gratitude to God because unidimensional measures are better at explaining more complex multidimensional constructs. Clearly, religiousness and spirituality are complex multidimensional constructs. Using multidimensional measures in an attempt to understand religiousness and spirituality may not help clarify our understanding in the way that a unidimensional measure can. In the studies that follow, we describe the development of unidimensional gratitude to God measure that is derived from the theoretical foundations of the Gratitude Questionnaire (GQ-6) (McCullough et al. 2002) and the Gratitude, Resentment, and Appreciation Test (GRAT) (Watkins et al. 2003). Both the GQ-6 and the GRAT are measures of dispositional gratitude and thus attempt to measure gratitude as an affective trait. In brief, a person high in dispositional gratitude should have a low threshold for experiencing grateful emotion. The GQ-6 is a measure of trait gratitude and is based on the theory that grateful persons should experience gratitude more intensely, more frequently, with greater span (a greater variety of life circumstances should provoke gratitude in grateful people), and greater density (grateful people should feel grateful to a greater number of individuals for a single benefit). We attempted to develop items that assessed these facets, except for density, because our intent was to focus on one benefactor (God). We propose that individuals who are dispositionally grateful to God should experience gratitude to God more intensely and more frequently, and a greater variety of life events should provoke gratitude to God in these individuals.

The GRAT (Watkins et al. 2003) was developed with the idea that grateful people should be characterized by three facets that should be foundational to trait gratitude: appreciation of others, a sense of abundance, and an appreciation for simple pleasures. Thus, we propose that individuals high in trait gratitude to God should show an appreciation for the other people God has brought into their lives, they should feel that God’s blessings have been abundant, and they should see even simple pleasures as divine gifts. Both the GQ-6 and the GRAT have shown good psychometric characteristics, and thus, we reasoned that the affective trait of gratitude to God should be much the same as the general gratitude trait, with the exception that it is focused on one transcendent benefactor: God. Thus, we modified items from these measures to be directed toward a divine benefactor. For example, the item from the GQ-6, “I have so much in life to be thankful for,” was rewritten as “God has given me an overwhelming number of blessings in my life,” and the item from the GRAT, “I think that it’s important to pause often to “count my blessings” was changed to “I think that it’s important to regularly remember the good things God has done for me.” We also added items we thought would be related to constructs closely tied to gratitude to God, such as one’s experience of divine grace (e.g., “My life is filled with God’s grace”). Finally, we thought that it was important to include several reverse-scored items (e.g., “Because my life has been so bad, I have very little to thank God for”).

To our knowledge, the only extant measure of dispositional gratitude to God was developed by Rosmarin et al. (2011). Although this measure served the purposes of the authors for answering their specific research questions, there are several reasons why we felt another gratitude to God measure was called for. First and foremost, we wanted to develop a theory-driven measure, as the Rosmarin et al. scale of “religious gratitude” was created by simply rewording items from the GQ-6. Second, although their measure of “religious gratitude” showed good psychometrics, it was not unambiguously unidimensional (no confirmatory factor analysis or factor item loadings were reported). Finally, all data gathered using their gratitude to God scale was cross-sectional. We felt that the temporal stability of a dispositional measure of gratitude to God would be important, and it would also be important to examine how gratitude to God predicted important variables over time. For these reasons, we developed the Gratitude to God-Trait scale (GTG-T).

1.6. Overview of the Current Studies

To summarize, God is a deep source of benefits for those who believe in God (Scheibe et al. 2017). Because most people believe in a benevolent God (Pew Research Center 2018), it is important to study God as a perceived benefactor in people’s lives and the gratitude to God that people experience in response to these benefits. As a spiritual emotion, gratitude to God is likely to be central to the well-being of religious individuals—particularly to their spiritual well-being. Finally, gratitude to God could serve as a spiritual motivator. In these studies, we develop a trait measure of gratitude to God and investigate its association with spiritual well-being and belief in God. In Study 1 we developed a gratitude to God trait measure (GTG-T) in which we administered the items with a number of religious and well-being measures across several different populations. A sub-sample also completed measures one month after the first administration, allowing us to investigate how gratitude to God might predict various variables. Using the GTG-T developed in Study 1, we conducted a six-week prospective study to investigate the impact of gratitude to God on the cultivation of gratitude in Study 2. In Study 3, we conducted a longer and more rigorous prospective study (about two months) where we administered the GTG-T with several measures that we thought might be relevant to the well-being of religious individuals. Finally, In Study 4, participants first wrote about either a positive experience or a positive spiritual experience. We then assessed their emotions, various R/S measures, and the GTG-T, with the goal of providing further construct validity for this measure.

2. Study 1

In this study, we describe the initial development of the GTG-T. We began with 32 items based on the theories behind the GQ-6 (McCullough et al. 2002) and the GRAT (Watkins et al. 2003), as described earlier. Because our purpose was to develop a unidimensional measure of gratitude to God, we expected that many of these items would load on several factors and thus would not be retained in the final measure. In this study, our purpose was to develop the GTG-T and provide initial tests of its reliability and validity. Thus, we administered our initial version of the GTG-T along with several other self-report questionnaires. A subset of our population also completed measures one month later, which allowed us the opportunity to investigate how gratitude to God impacts variables over time. We were particularly interested in how gratitude to God impacts spiritual well-being and thus predicted that gratitude to God would predict increases in spiritual well-being over time. Construct validity of the GTG-T was tested using several measures. Because gratitude to God may be important to general gratitude, we expected trait gratitude measures would correlate moderately with gratitude to God. Research has shown that general gratitude is strongly associated with subjective well-being (Watkins 2014). Thus, we administered several subjective well-being measures and expected small to moderate positive correlations with the GTG-T. Obviously, gratitude to God should be a construct important to religiousness and spirituality and should be strongly related to these variables. Thus, we predicted that the GTG-T should be positively associated with one’s experience with God and with spiritual transcendence. Intrinsic and extrinsic religious orientation is a time-honored construct in the psychology of religion. Thus, we also investigated how gratitude to God predicts these variables. Because recent work has shown that both intrinsic and extrinsic orientations positively correlate and that extrinsic scales do not provide a pure assessment of extrinsic motivations for religious behavior in an unequivocal manner (Nielsen et al. 2013), we predicted that the GTG-T would be positively associated with both intrinsic and extrinsic orientations, but that it would be more strongly correlated with intrinsic religiosity. This is because if an individual was genuinely grateful to God, it follows that they would be pursuing religious activities because of their pursuit of God, not because of instrumental gains that religion might have to offer. Our final measure of religiousness assessed one’s attachment to God. Because general gratitude is important to relationships and secure attachment (Algoe 2012), we predicted that gratitude to God would also be associated with a secure attachment with God and would particularly be negatively associated with avoidance of God.

2.1. Participants and Procedure

Participants were undergraduates from a mid-level regional public university in the Pacific Northwest who received partial course credit for their participation. Although we did not assess denominational status, historically, attendance at religious services in the Pacific Northwest is comparatively low (e.g., Gallup 2015), and because of the secular nature of this university, we can assume that our population was somewhat less religious than the general American public. The 32 items from the initial GTG-T were administered with various measures at two points in time, one month apart. Because the initial validation study used participants from several populations, the number of participants completing the measures varied, but 308 participants completed the GTG-T at T1 and T2. The mean age of participants was 20.96, with a range of 18 to 65; 71.2% reported gender as female, 23.5% male, and 5.3% “other”. No significant differences were found in gender, age, T1 GTG-T scores, and T1 spiritual well-being scores between those who dropped out and those who completed both waves of data. This study was approved by the Institutional Review Board of Eastern Washington University and complied with the ethical standards of the American Psychological Association.

2.2. Measures

GTG-T. We began with 32 items deemed appropriate to assessing one’s disposition for Gratitude to God. Participants responded to statements such as “God has given me an overwhelming number of blessings in my life” on a 9-point scale ranging from “I strongly disagree” to “I strongly agree with the statement.”

Trait Gratitude Measures. To assess general gratitude, we used two of the most widely used measures for this purpose: the GQ-6 (McCullough et al. 2002) and the GRAT (Watkins et al. 2003). The GQ-6 consists of 6 questions that participants rate on a 7-point scale ranging from strongly disagree to strongly agree. Similarly, the response scale on the GRAT ranges from “strongly disagree” to “I strongly agree with the statement” but uses a 9-point scale.

Well-being Measures. We administered several scales to assess positive well-being and one to assess ill-being. The Satisfaction With Life Scale (SWLS) (Diener et al. 1985) is one of the most commonly used measures of SWB. This scale purportedly taps the more cognitive aspect of happiness (judgments of satisfaction, α = 0.855 for this sample), but we also administered the Subjective Happiness Scale (SHS; α = 0.872) (Lyubomirsky and Lepper 1999) which assesses the subjective aspect of global happiness. The SWLS contains 5 questions and uses a 7-point response scale ranging from “strongly disagree” to “strongly agree.” The response scale for the SHS also uses a 7-point scale with 4 items, but the scale is anchored differently according to the item. We also included one measure of ill-being: the Beck Depression Inventory (BDI; α = 0.905) (Beck 1976). This is a well-validated measure of depression symptoms and thus provides us with some indication of an individual’s psychological suffering. Participants respond to 21 symptoms of depression on a 4-point scale. Because we expected that gratitude to God would be more strongly associated with spiritual well-being, we administered the Spiritual Well-being Scale (Paloutzian and Ellison 1982) (α = 0.912). This is a frequently used measure in the psychology of religion and contains two facet subscales: religious and existential well-being. Because these are facets that theoretically contribute to the more general construct of spiritual well-being, we used the total score of this scale. This measure uses a 6-point scale ranging from “strongly agree” to “strongly disagree.”

Religiousness and Spirituality Measures. These included the Age Universal I/E Scale (I/E-R) (Gorsuch and Venable 1983), which assesses intrinsic (α = 0.850) and extrinsic (α = 0.769) religious orientations. On this measure, participants respond to 14 items on a 5-point scale ranging from “strongly disagree” to “strongly agree.” The Religious Experience Questionnaire (REQ; α = 0.895) (Edwards 1976) is an 11-item scale tapping one’s personal experience with God with items such as ‘I feel very close to God in prayer, during public worship, or at important moments of my life.’ This measure uses a 7-point scale ranging from “never” to “always”. The Spiritual Transcendence Scale (STS short version) (Piedmont 1999) is a non-sectarian measure of spirituality that attempts to assess an individual’s ability to see broader, transcendent issues beyond the self (α = 0.731). Participants responded to items with a 5-point scale ranging from “strongly disagree” to “strongly agree.” Finally, we administered the Attachment to God Inventory (AGI) (Beck and McDonald 2004). This questionnaire follows human attachment theory in an attempt to assess one’s attachment to God. This measure has two subscales: avoidance of God (α = 0.916) and anxiety toward God (α = 0.888). The AGI has 28 items and uses a 7-point response scale ranging from “disagree strongly” to “agree strongly.”

Personality Measures. To measure basic personality traits, we used the HEXACO Personality Inventory (Lee and Ashton 2004). The HEXACO taps the Big-5 personality traits (Emotional Stability, Extraversion, Agreeableness, Conscientiousness, and Openness) but also posits a sixth major trait: Honesty/Humility. We utilized this personality measure because it is one of the best instruments for assessing general/basic personality traits. Moreover, the Honesty/Humility trait is not included in standard Big-5 measures, and this trait should have a notable relationship with gratitude to God. Because of the length of time required to complete the measures described above, the HEXACO was administered one day after the primary measures at T1. Thus, only 130 participants completed both the GTG-T and the HEXACO. This measure contains 200 items that participants respond to on a 7-point scale ranging from “strongly disagree” to “strongly agree.”

We also administered two measures for more exploratory purposes. The Indebtedness Scale-Revised (IS-R, α = 0.908) (Van Gelder et al. 2007) assesses a person’s disposition to experience indebtedness in response to benefaction. This questionnaire includes 22 items that participants respond to on a 6-point scale ranging from “strongly disagree” to “strongly agree.” Because research has shown that general trait gratitude is negatively associated with indebtedness (Van Gelder et al. 2007), we predicted that gratitude to God would be negatively correlated with indebtedness. We also administered the Rosenberg Self-esteem Scale (RSE; α = 0.881) (Rosenberg 1965). The RSE is one of the most used measures of self-esteem and utilizes 10 items that participants respond to on a 5-point scale ranging from “strongly disagree” to “strongly agree.” The RSE was administered for exploratory purposes, and thus, we had no specific predictions regarding this measure.

Following is the order of the measures as presented to participants at T1: GTG-T, report of belief in personal God, IS-R, SWLS, SHS, GRAT-S, Spiritual Well-being Scale, GQ-6, BDI, and HEXACO1. At T2, participants took these measures in the following order: GTG-T, GRAT-S, Spiritual Well-being Scale, IE/R, REQ, STS, AGI, GQ-6, and RSE.

2.3. Results and Discussion

Development of the GTG-T and Psychometric Analyses

As referred to above, in creating a measure of one’s disposition for gratitude to God, our goal was to develop a unidimensional measure. Internal consistency of the original 32 items, as assessed by Cronbach’s Alpha, was quite good (T1 a = 0.97; T2 a = 0.98). Exploratory analysis using PCA showed that four factors had Eigenvalues greater than 1.0. One primary factor emerged that explained 54.80% of the variance of the GTG-T. A second factor explained 11.13% of the variance, suggesting that a number of items were loading on a separate factor. We retained items that loaded strongly on the primary factor and also appeared to offer a unique contribution to the GTG-T. Ten items remained after eliminating statistically and conceptually redundant items, and these items form our final version of the GTG-T (see

Table 1. Final Items and Statistics for the Trait Gratitude to God scale (GTG-T).

Item	Factor Loading	Corrected Item-Total
1. God has given me an overwhelming number of blessings in my life.	0.898	0.862
2. Simple pleasures like a beautiful view or a delicious meal often remind me of how good God has been to me.	0.854	0.835
3. When I think of how good others have been to me, it makes me feel very thankful to God.	0.883	0.864
4. Because my life has been so bad, I have very little to thank God for. *	0.600	0.574
5. When I see the beauty of new life in the spring, I feel very thankful toward God.	0.870	0.808
6. When I think of what others have done for me in my life, it makes me appreciate God even more.	0.897	0.928
7. I feel very grateful to God when I think of my family.	0.882	0.870
8. I think that it’s important to regularly remember the good things God has done for me.	0.907	0.914
9. My life is filled with God’s grace.	0.869	0.868
10. Life is a wonderful gift from God.	0.893	0.878

* Reverse scored item.

). Internal consistency for the final version of the GTG-T was quite good (T1 a = 0.96, T2 a = 0.96), all corrected item-total correlations were above 0.5, and this shorter version correlated well with the original 32-item scale (T1 r = 0.97; T2 r = 0.98). PCA exploratory factor analysis of the final 10 items showed that only one factor emerged with an Eigenvalue over 1.0, and this factor explained 73.91% of the variance (no other factor explained more than 7% of the variance). This factor structure was replicated at T2. To test the unidimensional nature of these items, we conducted a CFA using AMOS for one factor. This analysis showed an acceptable fit (CFI = 0.921). Thus, it is apparent that one factor best fits the pattern of data, and the final 10-item GTG-T represents a unidimensional measure of one’s disposition for GTG. Table 1 shows the final items of the GTG-T, along with relevant item statistics.

We also evaluated the temporal stability of the GTG-T with test-retest correlations. If the GTG-T is a dispositional measure of gratitude to God, it should show good temporal stability. Indeed, the test–retest correlation over one month was quite strong (r = 0.90, n = 308), thus providing further support for the reliability of the GTG-T.

We first evaluated the construct validity of the GTG-T by comparing scores between people with varying beliefs in the existence of God. Individuals reported whether they believed in a personal God (“yes,” “no,” or “maybe”). As predicted, GLM showed a strong main effect for belief in God, F(2, 284) = 91.52, p < 0.001, and post hoc LSD analyses showed that those who said they believed in God (M = 7.32, SD = 1.25; n = 182) were significantly higher in GTG than those who thought that “maybe” God existed (M = 5.19, SD = 1.46; n = 65), and these individuals were significantly higher in gratitude to God than those who said they did not believe in a personal God (M = 3.86, SD = 2.42; n = 38; all ps < 0.001). Although gratitude to God should be related to general gratitude, if gratitude to God is a unique construct, these associations should not be strong. Indeed, the GTG-T was moderately related to both the GQ-6 and GRAT-S (see

Table 2. Correlations of GTG-T and Personality Scales.

Scale	GTG-T	GQ-6	GRAT-S
GQ-6	0.368 ***	---	0.724 ***
GRAT-S	0.383 ***	0.724 ***	---
Emotionality (HEXACO)	0.024	−0.125	−0.163
Extraversion (HEXACO)	0.212 *	0.559 ***	0.561 ***
Agreeableness (HEXACO)	0.159	0.245 **	0.270 **
Conscientiousness (HEXACO)	0.161	0.100	0.092
Openness (HEXACO)	0.167	0.241 **	0.284 **
Honesty/Humility (HEXACO)	0.244 **	0.192 *	0.275 **
Self-esteem (RSE)	0.177 **	0.484 ***	0.544 ***
Indebtedness (IS-R)	−0.232 **	−0.335 ***	−0.216 **

Note. * p < 0.05; ** p < 0.01; *** p < 0.001.

), supporting the theory that gratitude to God is related to but distinct from general gratitude.

The more critical evaluation of the construct validity of the GTG-T would be found in relationships with religiousness and spirituality. If gratitude to God is not simply gratitude but is actually a special case of gratitude called “religious gratitude” (Rosmarin et al. 2011), then relationships of gratitude to God with religiousness and spirituality should be higher than relationships with general gratitude. These correlations are shown in

Table 3. Correlations of GTG-T and General Gratitude Measures with Religiousness and Spirituality.

Scale	GTG-T	GQ-6	GRAT-S
Spiritual Well-being	0.757 ***	0.523 ***	0.547 ***
-Religion subscale	0.840 ***	0.276 **	0.309 ***
-Existential subscale	0.215 **	0.662 ***	0.652 ***
Extrinsic Religiosity	0.547 ***	−0.218	0.012
Intrinsic Religiosity	0.712 ***	0.164	0.306 ***
Religious Experiences Questionnaire	0.834 ***	0.401 **	0.272 *
Spiritual Transcendence Scale	0.511 ***	0.392 ***	0.425 ***
-Prayer subscale	0.703 ***	0.357 ***	0.289 *
-Universal subscale	0.379 ***	0.419 ***	0.401 **
-Connection subscale	−0.061	0.054	0.213
Attachment to God Anxiety	0.085	−0.385 **	−0.350 **
Attachment to God Avoidance	−0.705 ***	−0.297 *	−0.164

Note. * p < 0.05; ** p < 0.01; *** p < 0.001.

. In general, these correlations were as predicted. Discriminant validity was also shown in that religiousness and spirituality measures correlate much more strongly with the GTG-T than with general gratitude. Also, as predicted, the GTG-T correlated more strongly with intrinsic than extrinsic religiosity. It should be admitted, however, that gratitude to God correlated strongly with both indices, but given that more recent research on religious orientation has shown that intrinsic and extrinsic religiosity are positively correlated (Nielsen et al. 2013), this should be expected. An interesting pattern of relationships emerged with attachment to God; whereas moderate negative relationships were found between anxiety about God and general gratitude, there was no relationship observed between the GTG-T and anxiety about God. This is likely because committed religious individuals have a strong respect for the God they worship; thus, their anxiety about God is somewhat ambivalent. On the other hand, a strong negative association was found between gratitude to God and avoidance of God, but small and inconsistent relationships were found between this variable and general gratitude. Thus, gratitude to God strongly predicts an approach to God. Additional relationships to various variables can be found in correlation matrices the Supplementary Materials.

In

Table 4. Correlations of GTG-T and Well-being Measures.

Scale	GTG-T	GQ-6	GRAT-S
Satisfaction with Life Scale	0.192 **	0.495 ***	0.582 ***
Subjective Happiness Scale	0.215 ***	0.524 ***	0.590 ***
Beck Depression Inventory	−0.071	−0.314 ***	−0.310 ***

Note. ** p < 0.01; *** p < 0.001.

, we present the relationship between gratitude to God and well-being measures. These correlations show that gratitude to God is positively associated with well-being, but note that the general gratitude measures predict SWB much more strongly than gratitude to God. As seen in Table 3, the GTG-T was very strongly associated with spiritual well-being. Thus, the pattern that seems to be emerging is that gratitude to God is more central to spiritual well-being, whereas general gratitude is more related to general well-being. One possibility is that gratitude to God promotes the cultivation of general gratitude, which in turn promotes SWB. We explore this possibility in Studies 2 and 3.

Personality measures were administered to explore the personality characteristics of people high in gratitude to God. Table 4 shows these relationships. These relationships paint a picture of one high in gratitude to God as being more humble, not prone to indebtedness, slightly more extroverted, and having slightly higher self-esteem2.

The strong relationships of gratitude to God with spiritual well-being invite the possibility that gratitude to God fosters spiritual well-being. Because some participants took the GTG-T and the Spiritual Well-being Scale at two points in time, a month apart, this allowed us to investigate prospective relationships between gratitude to God and spiritual well-being. We conducted a hierarchical regression predicting T2 spiritual well-being. In step 1, we entered T1 spiritual well-being and the six personality characteristics from the HEXACO. In step 2, we entered T1 GTG-T scores to see if gratitude to God predicted increases in spiritual well-being after controlling for relevant T1 variables. Both models were significant (p < 0.001), but adding T1 gratitude to God in step 2 significantly improved the model, ΔR² = 0.050, F_change (1, 47) = 11.875, p = 0.001, partial correlation = 0.391. Thus, gratitude to God predicted increased levels of spiritual well-being over time.

In sum, evidence from Study 1 supports the validity of the GTG-T to measure one’s disposition for gratitude to God. Second, the GTG-T was found to be strongly related to several measures of spirituality, intrinsic religiosity, and one’s approach tendencies to God. Although gratitude to God showed small relationships with secular well-being measures, gratitude to God was more strongly related to spiritual well-being, and indeed, gratitude to God predicted significant increases in spiritual well-being over one month. In Studies 2 and 3, we investigate the possibility that gratitude to God helps one cultivate general gratitude.

3. Study 2

Study 1 showed that although gratitude to God was associated with subjective well-being, general dispositional gratitude measures showed much stronger relationships with SWB. This leads to the possibility that gratitude to God helps one cultivate gratitude more generally and thus impacts well-being indirectly. We investigated this possibility in Study 2. In this prospective design, we administered the GTG-T with state gratitude and other measures at two time points, six weeks apart. This study also served the purpose of cross-validating the GTG-T.

3.1. Method

3.1.1. Participants and Procedure

In this study, 108 participants completed several self-report measures at two time points, six weeks apart, in exchange for partial course credit (T1 n = 131, T2 n = 108). Of these participants, 28 were males and 80 were females. Unfortunately, data regarding age and ethnicity was not collected. In terms of gender, no significant differences were found between those who completed both waves of data and those who only completed T1 variables. Those who dropped out were significantly lower in GTG-T scores at T1 (p = 0.03), but there were no group differences in state gratitude scores (p = 0.74), which was our critical dependent variable in this study. This study was approved by the Institutional Review Board of Eastern Washington University and complied with the ethical standards of the American Psychological Association.

3.1.2. Measures

As in Study 1, we administered the GTG-T, trait gratitude measures (GQ-6 and GRAT-S), the SWLS, the AGI, and the I/E-R. The critical measure used for assessing state gratitude was a modified version of the Positive and Negative Affect Schedule-Expanded form (PANAS-X) (Watson and Clark 1994). The PANAS-X is a 60-item scale that taps various forms of positive and negative emotional states. Because the PANAS-X does not include any adjectives for grateful emotion, we included the three adjectives from the Gratitude Adjectives Scale (GAS) (McCullough et al. 2002): ‘thankful,’ ‘grateful,’ and ‘appreciative.’ The PANAS-X is a well-validated measure of affect (Watson and Clark 1994), and the GAS has also been shown to be a sound measure of grateful emotion. In this study, α of the GAS was 0.91 (T1) and 0.93 (T2). The items of the GAS were randomly interspersed with the items of the PANAS-X, and we used the time frame for reporting “the past week.” Participants respond to the adjectives on this measure using a 5-point scale ranging from “very slightly or not at all” to “extremely.” Also, differing from Study 1, we included the Center for Epidemiological Studies Scale for Depression (CES-D, Radloff 1977, a = 0.911) as an additional measure of well-being. Participants respond to 20 common symptoms of depression on a 4-point scale ranging from “rarely or none of the time” to “most or all of the time.” We used the CES-D as an alternative measure of depression because this measure is often used in depression research and offers better psychometrics than the BDI (Radloff 1977). After providing informed consent, participants completed the measures at both time points in the following order: PANAS-X, SWLS, CES-D, AGI, GRAT-S, GQ-6, I/E-R, GTG-T3.

3.2. Results and Discussion

The first goal of this study was to provide a test of cross-validation of the GTG-T. As in Study 1, the reliability of the GTG-T appears to be quite good. Coefficient a was 0.96 (at T1 and T2). Test-rest reliability also showed that the GTG-T was temporally stable across six weeks: r = 0.86. Structural validation was seen in this study in that the GTG-T again showed a unifactorial structure. As in Study 1, we conducted a CFA to evaluate the model of a single-factor structure of the GTG-T. The model fit was good (CFI = 0.968). This factor structure was replicated at T2. Taken with results from Study 1, we can conclude that the GTG-T represents a reliable unidimensional measure of gratitude to God.

We also tested the construct validity of the GTG-T. As in our first study, we found a strong main effect for belief in a personal God (p < 0.001), with people who said that they believed in a personal God reporting significantly higher gratitude to God than those who said that “maybe” God existed (p < 0.001), and those individuals reported more gratitude to God than atheists (p < 0.001). Correlation analyses also supported the construct validity of the GTG-T. At T14, the GTG-T was significantly correlated with state gratitude as measured by the GAS (r = 0.26, p = 0.003), trait gratitude as assessed with the GRAT-S (r = 0.41, p < 0.009) and the GQ-6 (r = 0.42, p < 0.001).

Our positive well-being variables were also correlated with gratitude to God. The GTG-T was significantly associated with the SWLS (r = 0.22, p = 0.01) and the positive affect scale on the PANAS-X (r = 0.21, p = 0.02). Interestingly, the GTG-T was not reliably correlated with our negative well-being measures, as represented by the CES-D (r = −0.13, p = 0.16) and the negative affect scale on the PANAS-X (r = −0.06, p = 0.52).

For the most part, the correlations between the GTG-T and our religiousness and spirituality measures were consistent with Study 1. We found that the GTG-T showed stronger correlations with intrinsic (r = 0.65, p < 0.001) than extrinsic (r = 0.48, p < 0.001) religiosity. Also consistent with Study 1, the GTG-T was strongly and negatively associated with avoidance of God (r = −0.72, p < 0.001) but somewhat surprisingly showed a moderate positive correlation with anxiety about God (r = 0.32, p < 0.03). Although this scale was intended to assess anxious attachment to God, it includes items that would likely be more central to those who feel their relationship with God is important. For example, it is easy to see how one might feel very grateful to God yet still endorse items such as “I worry a lot about my relationship with God” and “I worry a lot about damaging my relationship with God.” Considering that this finding differs somewhat from Study 1 (where no association was found), it may be that this scale assesses several aspects of one’s feelings about their relationship with God, not simply anxiety.

In Study 1, we raised the possibility that gratitude to God might foster gratitude. Thus, we used our prospective design to evaluate if gratitude to God might predict increases in grateful emotion. We conducted a partial correlation of T1 GTG-T with T2 GAS scores while controlling for T1 GAS. Indeed, T1 gratitude to God predicted significant increases in grateful emotion at T2 (partial r = 0.25, p = 0.008). The reverse partial correlation—grateful emotion predicting increases in gratitude to God—was not significant (r = 0.10, p = 0.26). This raises the possibility that gratitude to God enhances well-being by cultivating general gratitude. We evaluated this theory with mediation analysis by investigating the relationship between gratitude to God and well-being, with gratitude (as measured by the GAS) as the mediator. Using a least squares path analysis, gratitude to God indirectly influenced hedonic well-being through its effect on generalized gratitude. As participants’ gratitude to God increased at T1, so too did their GAS scores at T2 (a = 0.578, p = 0.0001), and increasing T2 GAS predicted increased subjective well-being at T2 (b = 2.831, p < 0.0001). Importantly, we found that the indirect effect (ab = 1.635) confidence interval was completely above zero (0.695 to 2.830), which was computed from a bias-corrected bootstrap based on 5000 bootstrap samples, indicating a reliable indirect effect. Sobel’s test of the indirect effect was also significant (z = 3.44, p = 0.0006). There was no significant direct effect (c’ = −0.998, p = 0.15), indicating that there was no evidence that gratitude to God impacted well-being independently from generalized gratitude. Of course, interpreting mediational analyses as actually revealing a causal effect can be problematic (Hayes 2022). This study is still essentially correlational in nature, and thus, the reader should be cautioned against making definitive causal interpretations from this mediation analysis.

In sum, results from Study 2 largely replicated findings from Study 1 and thus provided cross-validation for the GTG-T. The unique contribution from Study 2 was that gratitude to God prospectively predicted increases in grateful emotion over a six-week period, and generalized gratitude mediated the relationship between gratitude to God and well-being. Thus, results from this study support the theory that gratitude to God supports the cultivation of gratitude, and this might be how gratitude to God enhances well-being.

4. Study 3

Studies 1 and 2 provided evidence for the validity of the GTG-T and prospective evidence that gratitude to God predicts enhanced spiritual well-being and the cultivation of gratitude more generally. In Study 3, we provided a more rigorous test of the hypothesis that gratitude to God cultivates gratitude by controlling for self-report bias and the Big-5 personality characteristics. As discussed in the general introduction, research shows that gratitude encourages the development of relationships. Thus, it follows that gratitude to God should encourage one’s relationship with the Divine. Because gratitude to God might encourage one’s relationship with the Divine, we also sought to investigate whether gratitude to God enhances one’s confidence in the existence of God. As Chesterton argued (Chesterton [1908] 1986), when one finds oneself grateful but has no obvious benefactor to thank, one may turn to be thankful to God, and thus gratitude to God might encourage belief in a benevolent God.

4.1. Method

4.1.1. Participants and Procedure

Students were awarded partial course credit for completing the study, and 153 participants completed the measures at both time points; 211 students completed the measures at T1, and 172 completed the questionnaires at T2. No significant differences in terms of gender, age, income, or T1 GTG-T scores were found between those who completed only T1 measures and those who completed both waves of data. For the population completing the measures at T1, 33.6% were males, 65.4% were females, and 0.8% reported “other”. At T2, 31.4% were males, 68.0% were females, and 0.58% reported “other.” The age range at T1 was between 18 and 51, with a mean age of 21.15 (SD = 4.16). At T2, age ranged between 18 and 52, with a mean age of 21.11 (SD = 4.21). At T1, 55.0% reported “white” ethnicity, 2.7% African American, 1.9% Asian, and 9.6% Hispanic, and the remainder reported either other ethnicities, mixed ethnicity or did not report. At T2, 46.2% reported “white” ethnicity, 1.5% African American, 2.7% Asian, 9.6% Hispanic, and the remainder reported either other ethnicities, mixed ethnicity or did not report. Participants completed self-report questionnaires twice, seven to nine weeks apart. This study was approved by the IRB of Eastern Washington University and was conducted according to the ethical principles of the APA.

4.1.2. Measures

We again used our modified form of the PANAS-X as described in Study 2, which included the adjectives from the GAS (α = 0.92; 0.92). Participants completed two forms of the PANAS, one for how they were feeling “now” and one for the frequency of experience during “the past few weeks.” Using the past week’s form, we created a measure of hedonic well-being by subtracting scores from the 10-item negative affect scale from the 10-item positive affect scale (Watson et al. 1988). Following the two forms of the PANAS-X, participants completed two recently developed joy scales5 (Watkins et al. 2018), followed by the two dispositional gratitude scales: the GRAT-S (α = 0.84; 0.90) and the GQ-6 (α = 0.81; 0.87). We created four different orders of these initial measures, with the two forms of the PANAS-X always occurring in succession. In order to control for the impact of major personality traits, participants then completed the Big Five Inventory (BFI) (John et al. 2008; John et al. 1991). This measure of the Big Five personality traits contains 44 items that participants respond to on a 5-point scale ranging from “disagree strongly” to “agree strongly.” Because self-report bias is a major concern for individuals reporting on issues of religiousness, we felt that it was necessary to control for reporting bias and thus used the Balanced Inventory of Desirable Responding (BIDR) (Paulhus 1991). This measure has 41 items and uses a 7-point response scale ranging from “not true” to “very true.” We used both subscales of this instrument: impression management (α = 0.74) and self-deceit (α = 0.67).

Participants then completed the GTG-T, followed by the PSI Grace Scale (Bufford et al. 2015; Sisemore et al. 2006). The PSI Grace Scale contains 35 items and uses a 5-point response scale ranging from “strongly disagree” to “strongly agree.” The PSI Grace Scale was included for exploratory purposes to investigate the relationship between one’s experience of Divine grace and gratitude to God. Unfortunately, this scale did not reveal good psychometrics. This measure contains four subscales, and the scale that was of most interest to us—‘Grace Independent of Works’—was particularly poor in terms of internal consistency (α = 0.44; 0.48), and thus, we did not analyze results relevant to this factor. The other factors that were of interest to us showed adequate internal consistency: “Grace Toward Self,” α = 0.61; 0.67; “Responsible Grace,” α = 82; 0.87; and “Grace Toward Others,” α = 79; 0.80. This latter scale was used to investigate the hypothesis that gratitude to God is a spiritual motivator. If so, then gratitude to God should be correlated with this scale and should predict increases in one’s willingness to forgive and show grace to others over time. Finally, participants completed our measure of one’s confidence in the existence of God, “In terms of how you’re feeling right now, how confident are you that a loving personal God actually exists?” Students responded on an 11-point scale that ranged from “0%; I’m completely confident that God does not exist” to “100%; I’m completely confident that God exists”. Finally, participants completed demographic information6.

4.2. Results and Discussion

As with our first two studies, we conducted cross-validation analyses on the GTG-T. To summarize our findings, results supported the conclusion that the GTG-T is sound psychometrically. The GTG-T demonstrated excellent internal consistency (α = 0.96; 0.97) and temporal stability (test-retest r = 0.89). Furthermore, CFA supported the unidimensional nature of the GTG-T. CFA using AMOS showed an acceptable model fit for one factor (CFI = 0.927). We should state that our reverse-scored item did not load strongly on the factor at either time point; thus, researchers could choose to leave out this item when they use the GTG-T.

Once again, GTG-T was significantly related to trait gratitude (GRAT-S r = 0.38, p < 0.009; GQ-6 r = 0.35, p < 0.001), grateful emotion as measured by the GAS (‘now’ r = 0.38, p < 0.001; ‘last few weeks’ r = 0.37, p < 0.009), and hedonic well-being (r = 0.22, p < 0.001). Correlations of the GTG-T with the BFI showed that gratitude to God was most strongly related to agreeability (r = 0.23, p = 0.001) and conscientiousness (r = 0.14, p = 0.04). Gratitude to God was not significantly related to extraversion (r = 0.10, p = 0.15), openness (r = −0.02, p = 0.83), or neuroticism (r = −0.08, p = 0.24). These small and non-significant correlations show that the gratitude to God trait is distinct from the Big-5 personality characteristics. Our finding that extraversion did not correlate with gratitude to God contrasts somewhat with the results from Study 1. Although the correlation with extraversion in Study 1 was quite small, this inconsistency may be due to the contrasting nature of the Big-5 measures, but future research could further examine this relationship.

The primary purpose of this study was to provide a more rigorous test of the theory that gratitude to God helps people cultivate gratitude more generally. Thus, we conducted a prospective hierarchical regression analysis where we investigated whether gratitude to God at T1 predicted T2 levels of grateful emotion (measured by the GAS) after controlling for initial levels of the GAS, the Big-5 personality characteristics, and self-report response bias (via the BIDR). Thus, we entered our control variables in step one, and indeed, this model was significant (R² = 0.253, F(8,143) = 6.05, p < 0.001). Model 2 was also significant, but importantly, ΔR² = 0.077, F_change (1,142) = 16.413, p < 0.001, partial correlation = 0.322, indicating that the T1 GTG-T predicted T2 gratitude above and beyond T1 gratitude, Big-5 personality characteristics, and response bias. We believe that this was a strong test of the hypothesis that gratitude to God enhances general gratitude. Although the prospective aspect of this design eliminates many confounds inherent in cross-sectional correlations, experimental studies are called for to more conclusively support this hypothesis.

A secondary purpose of this study was to investigate if gratitude to God might enhance one’s confidence in the existence of God over time. Thus, we conducted a similar prospective hierarchical regression, where we entered baseline confidence in the existence of God, Big-5 personality characteristics, and response bias (BIDR) in Step 1 and T1 GTG-T in Step 2 in predicting T2 confidence in the existence of God. Both models were significant (p < 0.001; p = 0.003), but critically, adding T1 GTG-T scores in Step 2 showed a significant improvement in predicting confidence in the existence of God, ΔR² = 0.013, F_change (1, 127) = 9.290, p = 0.003, partial correlation = 0.261. Thus, after controlling for major personality characteristics and response bias, gratitude to God predicted significant increases in confidence in the existence of God over time.

We also conducted an exploratory analysis to investigate if gratitude to God predicted increases in general well-being over time. As mentioned above, in Step 1, we entered T1 levels of hedonic well-being (as measured by our composite PANAS measure), Big-5 personality traits, and response bias in predicting T2 well-being. In Step 2, we entered T1 GTG-T scores. As expected, model 1 was significant (p < 0.001), but adding T1 GTG-T scores in Step 2 was shown to be marginally significant, ΔR² = 0.012, F_change (1, 143) = 3.736, p = 0.055, partial correlation = 0.160. This result supports the idea that in addition to supporting spiritual well-being (see Study 1), gratitude to God may support general SWB as well. Of course, because of the marginally significant result, this finding must be interpreted with caution, and we encourage more researchers to investigate this idea with different measures of SWB.

Because of the evidence that gratitude to God impacts general gratitude and well-being over time, this raises the question of whether gratitude to God increases general well-being by enhancing general gratitude. As in Study 2, we evaluated this possibility with a simple mediation analysis. Using a least squares path analysis, gratitude to God indirectly influenced our hedonic well-being measure (PANAS) through its effect on generalized gratitude. As students’ gratitude to God increased at T1, so too did their GAS scores at T2 (a = 0.180, p < 0.0001), and increasing T2 GAS predicted increased subjective well-being at T2 (b = 0.690, p < 0.0001). Importantly, we found that the indirect effect (ab = 0.124) confidence interval was completely above zero (0.071 to 0.194), which was computed from a bias-corrected bootstrap based on 5000 bootstrap samples, indicating a reliable indirect effect. Sobel’s test of the indirect effect was also significant (z = 4.37, p < 0.0001). There was no significant direct effect (c’ = −0.020, p = 0.62), indicating that there was no evidence that gratitude to God impacted well-being independently from generalized gratitude. Again, however, as with the mediation analysis in Study 2, we caution the reader against making definitive causal interpretations from this analysis.

We also conducted analyses investigating the relationship between gratitude to God and grace. The GTG-T was shown to be independent of “Responsible Grace,” r = 0.007, p = 0.922, showed a small inverse relationship with “Grace Toward Self,” r = −0.16, p = 0.02, but a strong positive association with “Grace Toward Others,” r = 0.77, p < 0.001. We then investigated our proposed prospective relationship between gratitude to God and grace toward others. Indeed, we found that gratitude to God predicted increased grace toward others over time. Using a hierarchical regression approach, in step 1, we entered T1 “Grace Toward Others” scores, scores from the BFI personality scales, and scores from the two subscales of the BIDR in the prediction of T2 “Grace Toward Others” scores. In step 2, we entered T1 GTG-T scores in an attempt to investigate if GTG-T predicted T2 grace toward others above and beyond T1 grace toward others, Big-5 personality traits, and response bias. Both models were significant (ps < 0.001), but importantly, adding T1 GTG-T scores in step 2 showed a significant improvement in predicting grace towards others, ΔR² = 0.052, F_change (1, 139) = 17.81, p < 0.001, partial correlation = 0.337. Thus, as predicted, gratitude to God predicted increases in one’s self-reported willingness to forgive and show grace toward others, supporting the theory that gratitude to God is a spiritual motivator.

Finally, we conducted a simple mediation analysis testing the idea that gratitude to God mediates the relationship between religiousness and well-being. Using a least squares path analysis, confidence in the existence of God indirectly influenced hedonic well-being through its effect on gratitude to God. As student’s confidence in the existence of a loving God increased, so too did their gratitude to God (a = 0.06, p < 0.0001), and increasing gratitude to God predicted increased well-being as indicated by our composite PANAS measure (b = 0.18, p = 0.03). Critically, we found that the indirect effect (ab = 0.011) confidence interval was completely above zero (0.0007 to 0.0205), which was computed from a bias-corrected bootstrap based on 5000 bootstrap samples, indicating a reliable indirect effect. Sobel’s test for the indirect effect was also significant (z = 2.11, p = 0.035). There was no significant direct effect (c’ = −0.005, p = 0.38), indicating that there was no evidence that confidence in God’s existence impacted well-being independently from gratitude to God. This replicates Rosmarin et al. (2011), supporting the theory that religiousness enhances well-being by encouraging gratitude to God.

Summarizing the results from Study 3, we found good evidence for the cross-validation of the GTG-T. Moreover, gratitude to God was found to predict increases in general gratitude and confidence in the existence of God over time. Finally, we found that religiousness may enhance well-being because it encourages gratitude to God. Thus, gratitude to God may be key to the cultivation of general gratitude and one’s relationship with the Divine.

5. Study 4

As demonstrated in the general introduction, one’s conception of God likely has important links with gratitude to God. In the final study, in addition to providing additional evidence for the construct validity of the GTG-T, we also investigated how gratitude to God is related to one’s view of God as authoritarian or loving. To investigate this relationship, we used a new measure that assesses one’s doctrinal and experiential view of God (Van Tongeren et al. 2019; Zahl and Gibson 2012). Moreover, Studies 1, 2, and 3 were all conducted using paper and pencil measures, and the data was collected well before the COVID-19 pandemic. This study was conducted online, and because in the current research environment, most researchers will likely administer the GTG-T online, we felt it important to have at least one study that demonstrated the validity of the GTG-T in an online format. Finally, we felt it important to validate the GTG-T with a more current data set.

5.1. Method

Participants and Procedure

In this preregistered online study, 226 students completed the survey and passed most of the attention check items (3 of 4 items). The age range of participants was 18 to 51 (M_age = 22.48); most participants were female (77.4%; Male = 20.4%; “other” = 0.9%; “prefer not to answer” = 1.3%), and white (52.7%; Hispanic = 16.3%; Native = 2.9%; Black = 2.9%; Asian = 2.0%; a significant proportion of participants reported mixed ethnicities).

Initially, participants were randomly allocated to one of two recall conditions: positive event control or positive spiritual event. In each recall condition, participants were instructed to recall three significant positive events, with the only difference being that they were to recall “important spiritual experiences” in the spiritual recall condition. We defined spiritual in the following way: “By a spiritual experience we mean an experience that you had with the Sacred. That could be God, nature, or anything you feel is transcendent—beyond yourself”. They were then to choose the one event they felt was “most important” and were to write about the event for five minutes.

After the recall writing task, participants reported various emotions in three ways: how they thought they should feel, how they actually felt “right now,” and how they felt during the writing exercise. The emotion descriptors followed the format of the Discrete Emotions Scale, where each emotion is represented by three descriptors. For example, the gratitude to God item stated, “Grateful, thankful, or appreciative to God.” We included three gratitude items: gratitude to God, gratitude to humans, and general gratitude. These items were interspersed with 19 other emotions (9 positive and 10 negative—see

Table 5. Construct Validity Correlations of GTG-T and General Gratitude in Study 4.

Scale	GTG-T	GQ-6
Meaning in Life Questionnaire
-Presence subscale	0.359 ***	0.481 ***
-Search subscale	0.161 *	−0.071
Santa Clara Strength of Religious Commitment	0.841 ***	0.235 ***
Loving God—Doctrinal	0.421 ***	0.203 **
Loving God—Experiential	0.832 ***	0.240 **
Authoritarian God—Doctrinal	−0.136 *	−0.030
Authoritarian God—Experiential	−0.113	−0.032
Nearness to God-Revised	0.856 ***	0.281 ***

Note. * p < 0.05; ** p < 0.01; *** p < 0.001.

).

After reporting on the positivity of the event that they wrote about on a 7-point Likert-type scale, participants reported their confidence in the existence of a loving God as described in previous studies. They then completed measures related to testing the construct validity of the GTG-T in the following order: the Meaning in Life Questionnaire (MLQ) (Steger et al. 2006), an experimental measure of state gratitude to God, the GQ-6 (McCullough et al. 2002), the Santa Clara Strength of Religious Faith scale (SRF) (Plante 2021), the Religious and Spiritual Struggles Scale (RSS) (Exline et al. 2014), and a doctrinal vs. experiential version of the Authoritarian/Benevolent God Scales (Van Tongeren et al. 2019; Zahl and Gibson 2012). This measure includes two columns in which participants report first their beliefs about “what is theologically true about God,” followed by their report on the same attribute as to the participants’ “personal experience of what God is like.” Thus, this questionnaire has the unique ability to assess a person’s “head versus “heart” concept of God. This was followed by the GTG-T, and finally, participants completed a revised version of the Nearness to God scale (NTG-R) (Gorsuch and Smith 1983) before reporting their demographic information.

5.2. Results and Discussion

Our recall manipulation was generally not successful because participants wrote about significantly more positive events in the positive recall control condition than in the spiritual recall condition. This resulted in participants reporting significantly more gratitude in the positive recall control condition on all three of our gratitude items, contrary to our expectations (this pattern was evident in the other positive emotions as well). In future studies, we recommend that researchers do not use an unspecified positive event for comparison with a divine benefit. For example, Watkins et al. (2024) used a more specific recall comparison condition (amusement), which resulted in the predicted emotional effects. However, one interesting and somewhat unexpected finding emerged from the writing samples. Only 15.9% of participants mentioned God in their writing about a positive event. Although participants were more likely to mention God in the spiritual recall condition (Pearson Chi-Square = 8.97, p = 0.003), only 23% mentioned God after recalling an important spiritual event. This finding is remarkably similar to another recent gratitude to God study (Van Cappellen et al. 2024). It seems obvious that thinking about God in the context of a positive event would be an important, if not necessary, antecedent for gratitude to God, but our findings imply that most people are not likely to think about God, even after recalling an important positive spiritual event.

As with the first three studies, the GTG-T showed good psychometric characteristics. Indeed, Crobach’s a was quite high (0.969). We should mention, however, that contrary to the first three studies, the reverse-scored item did not show adequate corrected item-total correlation (0.117). We believe that this was because this study was conducted online and thus probably had more careless responses than the first three studies. Researchers may choose to eliminate this item when administering the scale online, but one improvement of the measure might be to include more reverse-scored items, and if the first item is reverse-scored this might encourage participants to complete the measure in a more thoughtful manner.

As with the previous studies, the construct validity of the GTG-T was good (see Table 5). The GTG-T was more strongly related to religious/spirituality variables than general gratitude, as represented by the GQ-6. Perhaps the most interesting finding that emerges from Table 5 is that gratitude to God was much more strongly related to one’s experiential (i.e., heartfelt) conception of God as loving than their doctrinal or “head” conception of God as loving. This relationship was not clearly evident with general gratitude and replicated Watkins and colleagues’ findings (Watkins et al. 2023). Interestingly, this pattern was not evident in the conceptions of an authoritarian God. This underscores the finding that one’s heartfelt view of God as loving is crucial to gratitude to God (see also Exline and Wilt 2024; Park et al. 2022). Taken together with the relationship to nearness to God, this supports the theory that gratitude to God is a critical component of spiritual well-being. Although this study was cross-sectional, prospective studies have found that gratitude to God predicts enhanced spiritual well-being over time (Watkins et al. 2022).

Table 6. Correlations of GTG-T with Emotional Response to Writing in Study 4.

Emotion	GTG-T	GQ-6
Gratitude to God	0.683 ***	0.261 ***
Gratitude to Humans	0.330 ***	0.215 **
General Gratitude	0.372 ***	0.249 ***
Amused	0.270 ***	0.351 **
Awe	0.211 ***	0.234 **
Hope	0.270 ***	0.315 ***
Inspired/Elevated	0.244 ***	0.246 ***
Interest	0.158 *	0.138 *
Joy	0.254 ***	0.263 ***
Love/Trust	0.196 **	0.317 ***
Proud	0.140 *	0.127
Serene/Content	0.186 **	0.238 ***
Angry/Irritated	−0.100	−0.292 ***
Ashamed	−0.106	−0.284
Contempt	−0.070	−0.195 **
Disgust	−0.169 *	−0.276 ***
Embarrassed	−0.123	−0.255 ***
Guilty	−0.020	−0.065
Hate	−0.049	−0.223 ***
Sad	−0.094	−0.220 ***
Fear	−0.051	−0.266 ***
Stressed/Nervous	−0.058	−0.231 ***

Note. * p < 0.05; ** p < 0.01; *** p < 0.001.

provides the correlations of the GTG-T with one’s emotional response to writing about a significant positive event (both spiritual and non-spiritual7). As seen in this table, relationships would be as expected if the GTG-T is a valid instrument: in general, the GTG-T correlated much more highly with state gratitude to God than did the GQ-6, but general gratitude correlated more strongly and positively with other “secular” positive emotions. There are, however, two noteworthy exceptions to this pattern: dispositional gratitude to God was more strongly related to gratitude to humans and non-specified gratitude than the GQ-6. Interestingly, whereas the GTG-T did not show significant negative correlations with negative emotions, for the most part, the GQ-6 did.

We investigated the relationship of gratitude to God to spiritual struggles, and these correlations are seen in

Table 7. Correlations of Religious and Spiritual Struggles with Gratitude in Study 4.

Scale	GTG-T	GQ-6
Religious and Spiritual Struggles Total	0.43	−0.285 ***
-Divine Spiritual Struggles	−0.024	−0.343 ***
-Demon Spiritual Struggles	0.043	−0.142 *
-Moral Spiritual Struggles	0.273 ***	−0.091
-Interpersonal Spiritual Struggles	−0.108	−0.127
-Doubt Spiritual Struggles	0.025	−0.237 ***
-Ultimate Meaning Spiritual Struggles	−0.129	−0.439 ***

Note. * p < 0.05; *** p < 0.001.

. Although one’s general disposition for gratitude appears to protect one from spiritual struggles, gratitude to God does not and is even positively associated with moral and spiritual struggles. Of course, if one does not believe in God, one is not likely to experience spiritual struggles, and so we suspected that confidence in God’s existence might interact with these findings. Using our confidence in the existence of God scale, we divided our participants into three groups based on this scale: those who thought there was a 0 to 40% that God exists (n = 57), those who think there is a moderate likelihood that God exists (50–70%: n = 50), and those who believe there is a high chance that God exists (80–100%; n = 114). Indeed, the relationship between gratitude to God and spiritual struggles varied considerably based on participants’ confidence in the existence of God. For simplicity, here we report only the correlations of the GTG-T with the total spiritual struggles score, as the subscales generally reflected this pattern. For those who had a low confidence in God’s existence, gratitude to God was positively related to spiritual struggles (r = 0.302, p = 0.02). Whereas those with moderate confidence in God’s existence showed no relationship between gratitude to God and spiritual struggles (r = −0.031, p = 0.83), for those with high confidence in God’s existence, the GTG-T was negatively related to spiritual struggles (r = −0.199, p = 0.03). The GQ-6, on the other hand, was significantly and negatively related to spiritual struggles for all three groups. In other words, for those who have very little confidence in God’s existence, gratitude to God seems to promote spiritual struggles, whereas for those who are confident in the existence of God, gratitude to God appears to protect them from struggling spiritually. Because this was not a preregistered analysis, these conclusions must be held cautiously, but we look forward to future research that investigates how gratitude to God might decrease—or increase—one’s religious/spiritual struggles. In sum, Study 4 provided further evidence for the validity of the GTG-T and the importance of gratitude to God for spiritual well-being.

6. General Discussion and Conclusions

These studies offer four primary contributions to understanding the place of gratitude to God within the psychology of religion. First, results provide solid evidence of construct validity for the GTG-T for use in studying gratitude to God. Recent research has shown that gratitude to God is indeed unique from gratitude to humans and general gratitude (Hodge et al. 2024; Park et al. 2022; Tsang et al. 2023; Watkins et al. 2024; Zuniga et al. 2024), and thus we see the development of the GTG-T as a significant methodological contribution to this research. Second, Study 4 showed that gratitude was strongly related to spiritual well-being, and Study 1 showed that gratitude to God predicted increased spiritual well-being over time. Third, in Study 3, we found that gratitude to God predicted increases in one’s confidence in the existence of a loving God, and in Study 4, we found that gratitude to God was much more strongly related to one’s experiential view of a loving God than one’s doctrinal conception of God as loving. Finally, in Studies 2 and 3, gratitude to God predicted increases in general gratitude over time; thus, gratitude to God may enhance general gratitude. Moreover, mediational analyses from these studies supported the theory that gratitude to God promotes well-being by enhancing general gratitude. We discuss the implications of these findings below.

In four studies, the GTG-T showed good evidence for its reliability and construct validity as a unidimensional measure of one’s disposition for gratitude to God. The GTG-T showed excellent internal and temporal consistency, and factor analysis supported its structural validity across all four studies. Construct validity of this measure was supported in that it correlated, as expected, with religiousness and spirituality measures. Thus, researchers in the psychology of religion may want to incorporate this measure in their research. For example, although studies reliably show that religiousness is positively correlated with happiness, explanations for this association have not been exhaustively explored. Our results suggest that gratitude to God and general gratitude may be linked to the connection between religiousness and well-being. Clearly, however, this idea needs more empirical support, and this is one issue of many where researchers might find the GTG-T useful.

In this context, we emphasize that the GTG-T is a trait measure and thus is only appropriate for certain research questions concerning gratitude to God. For example, if researchers want to investigate how the manipulation of belief in the afterlife impacts gratitude to God (cf. Dechesne et al. 2003), the GTG-T would not be an appropriate dependent variable. Rather than a dispositional measure, a sensitive state measure of gratitude to God would be called for, such as those used in Study 4. In this regard, we are concerned that the temporal stability of the GTG-T might actually be too high. Test-retest correlations across Studies 1–3 were 0.90, 0.86, and 0.89. Although we should expect high temporal stability in trait measures, in our judgment, these associations appear to be too high. Previously, we developed a measure of state gratitude to God based on the three adjectives from the GAS (Scheibe et al. 2017), and this measure also had test-retest correlations above 0.8, even at six to eight weeks. It seems reasonable to expect that gratitude to God should change, so we are concerned about why people’s reports of their gratitude to God are so consistent over time. There are probably several reasons for this, but we believe that this evidence calls for the development of a sensitive state measure of gratitude to God (cf. Dunnington 2024).

The main finding of Study 1 was that gratitude to God was central to spiritual well-being. Not only was gratitude to God strongly associated with spiritual well-being, but it also predicted increases in spiritual well-being over time. This replicates Watkins et al. (2022), who found that gratitude to God predicted various measures of spiritual well-being over a longer time frame and provides further support for the relevance of gratitude to God for religious people. However, these findings beg the question as to how gratitude to God supports spiritual well-being. We propose that one reason that gratitude to God is important is because it promotes one’s relationship with the divine. Gratitude is a relationship-strengthening emotion (Algoe 2012), and this should include divine relationships as well as humans. Indeed, in Study 3, we found that gratitude to God predicted increases in people’s confidence in the existence of a loving God over time. Although gratitude scholars have argued that the relationship between gratitude and religiousness exists because religion promotes gratitude as a virtue, our results suggest that gratitude is conducive to increased religiousness. This would be consistent with the finding of Emmons and McCullough (2003) that a controlled gratitude intervention led to increases in feelings of closeness to God as well as enhancements in the spiritual virtues of love, compassion, and forgiveness. In this regard, we believe that investigations of one’s intimacy or nearness to God would be fruitful.

Gratitude to God might also promote spiritual well-being by enhancing one’s perception of the goodness of God. If gratitude to God enhances one’s ability to be grateful for more benefits in one’s life, as McCullough et al. (2002) suggest, this may enhance the perception of the goodness of God, which should, in turn, support spiritual well-being. We have proposed that gratitude facilitates well-being because it psychologically amplifies the good in one’s life (Watkins 2014). When the good in one’s life is psychologically amplified, one is more easily able to accept the goodness of God, and this ought to enhance one’s sense of spiritual fulfillment. In Study 4, the strong correlations of gratitude to God with one’s experience that God is loving support this theory, but this relationship could also be interpreted as indicating that one needs to experience God as loving for one to experience gratitude to God. We propose that both interpretations may be in play: a concept of a loving God likely contributes to gratitude to God, but gratitude to God also enhances one’s view of God as loving.

We have also suggested that gratitude to God might be a moral/spiritual motivator. Research has shown that gratitude promotes prosocial behavior (Bartlett et al. 2012; Bartlett and DeSteno 2006; DeSteno et al. 2010), and thus, gratitude to God may encourage religious and moral behavior. In support of this hypothesis, Study 3 found that gratitude to God predicted increases in one’s grace toward others over time. Future research should investigate how gratitude to God might encourage more specific religious practices such as prayer, worship, giving, and involvement in one’s community of faith. Theologians have suggested that gratitude to God is a compelling motivation for Christians (e.g., Barth [1956] 1961; Billings 2011; Keller 2012, 2015), and we believe that empirical investigation of this issue should be encouraged.

Evidence from our studies supports the proposition that gratitude to God aids in the cultivation of gratitude. In Studies 2 and 3, we found that the trait of gratitude to God predicted increases in grateful emotion over time. McCullough et al. (2002) argued that belief in God might promote general gratitude because theistic believers have a benevolent benefactor to thank for all positive outcomes, regardless of whether the benefits are the result of human benefaction. Thus, gratitude to God may enhance the “span” of gratitude, which increases the frequency of grateful emotion, which in turn may enhance SWB. Indeed, mediation analyses from Studies 2 and 3 supported this model; gratitude to God was shown to support well-being by enhancing general gratitude. Although evidence from our studies supports this path model, experimental studies investigating this issue are needed. In short, gratitude to God may be particularly conducive to the well-being of believers, and in Study 3, we found that gratitude to God mediated the relationship between belief in God and SWB. This result is consistent with the findings of Rosmarin et al. (2011) and calls for more research investigating how gratitude to God might foster flourishing in religious people.

As with any research, these studies have notable limitations. Of course, there are many issues raised with self-report measures, such as those used in these studies, and self-report bias with religious issues is of particular concern. However, we found that the GTG-T had essentially no relationship with self-report bias, as indicated by the BIDR. Our samples also limit the generalizability of our findings as our studies were primarily of Caucasians in the Pacific Northwest.

Because our results supported the importance of gratitude to God for spiritual well-being, this invites speculation on interventions that might enhance gratitude to God. Two gratitude interventions might easily be adapted for gratitude to God: grateful recounting and grateful reappraisal. Grateful recounting—or counting one’s blessings—is a gratitude exercise where individuals regularly list blessings that they are grateful for (e.g., Emmons and McCullough 2003). We have adapted this procedure to listing three blessings and writing about how one is grateful for each benefit (Watkins et al. 2015), and this could be easily adapted for gratitude to God. Individuals could list three “good things” (either recent benefits or ongoing blessings) and then write about how each of these things makes them feel grateful to God.

Grateful reappraisal could also be adapted to enhance gratitude to God. In this technique, participants write about the positive consequences of a negative event that they can now be grateful for (Watkins et al. 2008). We found that this form of reappraisal brought more closure and less negative affect and intrusiveness associated with the negative event memory. Adapting this technique for gratitude to God would involve asking religious individuals to write about positive things that they believe God has brought into their lives as a consequence of a negative event. We suspect that religious individuals may have been doing this in our previous study, and this could prove to be another exercise useful for enhancing gratitude to God.

Although grateful recounting and grateful reappraisal may be adapted to encourage gratitude to God, it is quite possible that time-tested religious practices may prove to be more effective in encouraging gratitude to God. Thus, researchers could investigate the impact of prayer, sacred reading, contemplative practices, and personal and corporate worship on gratitude to God. Recent research has shown that positive spiritual appraisals are critical antecedents to gratitude to God (Exline and Wilt 2024; Watkins et al. 2023; Watkins et al. 2024). Evidence from Study 4 and from Van Cappellen et al. (2024) suggests that people make these spiritual appraisals about events very infrequently. Thus, the crucial question is, how can people be encouraged to engage in spiritual appraisals? It could be argued that one of the primary functions of the historic spiritual practices listed above is to encourage positive spiritual appraisals. In Study 4 of this paper and in Van Cappellen et al. (2024), spiritual primes enhanced spiritual appraisals, but still, most people were not spontaneously explaining even a spiritual event in terms of the divine. For us, the fact that most people do not spontaneously bring God into their explanations of a positive event is an intriguing finding that deserves more research. It stands to reason that understanding why even religious people do not include God in their explanations of blessings would bring a crucial understanding of how spiritual appraisals and, thus, gratitude to God could be enhanced.

In his classic text Religious Affections, Jonathan Edwards wrote, “There is such a thing as a spiritual gratitude, which is a holy and divine emotion” (Edwards [1746] 1999, p. 132). Our results support Edwards’ idea that gratitude to God is indeed different from general gratitude. In four studies, we provided support for the validity of the GTG-T, a trait measure of gratitude to God that appears to be distinct from the general trait of gratitude. Furthermore, we found that gratitude to God was strongly associated with spiritual well-being and may even encourage greater belief in God. Finally, we showed that gratitude to God may foster the cultivation of general gratitude, which in turn enhances subjective well-being. Although the relationship between religiousness and well-being has been consistent in the literature (Myers 2000), the mechanisms of this relationship are not completely understood. Evidence from our studies suggests that religiousness encourages gratitude to God, which enhances general gratitude, which in turn supports well-being. C. S. Lewis observed, “Gratitude exclaims, very properly, ‘How good of God to give me this’” (Lewis 2003, p. 155). In these studies, we have seen that gratitude to God may encourage individuals to see the goodness of God more clearly, and thus, divine gratitude may be an important component of flourishing for religious people.