Abstract
Despite the prevalence of both chronic and transient loneliness and the detrimental consequences associated with them, as a negatively valenced response to social exclusion, loneliness has received surprisingly little attention in the marketing literature. Drawing on research showing that lonely people often lack meaning in their life, the authors propose that ritualistic behavior that involves consumer products may reduce loneliness by increasing meaning in life. Specifically, a series of studies finds that engaging in even minimal, unfamiliar rituals reduces loneliness among lonely consumers. The results support the important role of meaningfulness. The authors find that the effect of rituals on loneliness is mediated by meaning in life via perceived product meaningfulness. They also find that ritualistic behavior no longer affects loneliness when the experience of meaningfulness can be derived incidentally.
In a 2016 Harris Poll, an astonishing 72% of Americans reported experiencing loneliness—that is, negative affect in response to a lack of connections with others—and of these, over 30% stated that they felt lonely at least once a week. For some consumers, the experience of loneliness is brief and transient; for others, it is a powerful and permanent aspect of daily life (Peplau and Perlman 1982). The COVID-19 pandemic has only further exacerbated the ubiquity of loneliness, with most face-to-face socializing being limited to members of one’s own household—which has meant little or no direct social contact for the 35.7 million Americans who live alone (Ducharme 2020).
Feelings of loneliness are a response to involuntary or forced aloneness—that is, aloneness resulting from social exclusion or isolation despite one’s desire for social connections (Larsen 1990; Peplau and Perlman 1982). Notwithstanding the prevalence of loneliness, research has begun examining the relationship between loneliness and consumer behavior only recently (Gentina, Shrum, and Lowrey 2018; Wang, Zhu, and Shiv 2012). Loneliness, however, deserves attention not just because of its pervasiveness but also because of its association with many negative psychological and physical consequences, including depression, suicide, and early death (Hafner 2016). In the marketplace, loneliness is reflected in significant shifts in preferences (Wang, Zhu, and Shiv 2012) and greater attachment to products (Epley et al. 2008). Loneliness also has implications in the workplace: because of its impact on turnover, productivity, and absences, businesses in the United Kingdom alone lose an estimated £2.5 billion per year.
In light of the documented harmfulness of loneliness, it is important to investigate strategies that may help reduce its experience. Given that loneliness is often associated with a belief that one’s life lacks meaning (Stillman et al. 2009; Tam and Chan 2019; Wheeler, Reis, and Nezlek 1983), an approach may be to give one’s life more purpose. In the current research, we examine consumption rituals as ways in which consumers can enhance meaning in life (King et al. 2006) and reduce loneliness. We also investigate whether greater meaning in life produced by ritualistic activities guides purchase intentions of products included therein.
Although rituals often occur in celebratory, social, or religious settings and are an important part of communities in the marketplace (e.g., Muñiz and O’Guinn 2001), the types of ritual on which we focus in our research do not draw on shared cultural values. Instead, they may be provided by marketers or created by individual consumers. They may even be minimal and seemingly insignificant in nature but nonetheless can add meaning to consumers’ lives. Our psychological orientation toward the study of ritual complements prior research showing, for example, the effect of rituals on self-esteem (Fiese and Kline 1993) or self-control (Tian et al. 2018). Research exploring whether ritualistic behavior may improve emotional well-being—and if so, under which conditions or through which underlying process—has been limited to showing that ritualistic behavior is especially likely to arise in situations exemplified by negative affect, such as high anxiety (Celsi, Rose, and Leigh 1993), and that ritualistic behavior can influence performance by reducing anxiety (Brooks et al. 2016). The current research aims to address this lacuna in the literature.
Our findings add to the literature in several important ways. We are the first to examine the impact of consumption rituals on feelings of loneliness, as well as the conditions under which this effect is likely to occur. In addition, we identify changes in meaning in life as a heretofore unexamined process underlying the influence of engaging in consumption rituals on emotional well-being. Taken together, our results suggest that consumers derive meaning from ritualistic behaviors, such as consumption rituals that are seemingly insignificant, and by providing meaning, these rituals reduce loneliness, improving consumer welfare.
Theoretical Background
Loneliness
The establishment and maintenance of close social relationships is a fundamental human need, and deficits in belongingness engender feelings of loneliness (Stillman et al. 2009). Loneliness, “the unpleasant experience that occurs when a person’s network of social relations is deficient in some important way” (Perlman and Peplau 1982, p. 31), is associated with many adverse and harmful consequences, including low self-esteem (Jones, Freemon, and Goswick 1981), depression (Cacioppo et al. 2006), and—related to our research—the perception that one’s life lacks purpose or meaning (Stillman et al. 2009). Indeed, extant literature provides support for the relationship between loneliness and a lack of meaning in life. For example, Wheeler, Reis, and Nezlek (1983) found a negative relationship between individuals’ loneliness and the meaningfulness of their interactions with others. Moreover, Tam and Chan (2019) showed that both state and trait loneliness were negatively correlated with the meaningfulness of an activity. Similarly, research studying nurses found that lower perceived meaningfulness of their work increases the loneliness they experience in the workplace (Arslan, Yener, and Schermer 2020).
For some consumers, loneliness is a chronic state of aloneness, triggered by persistent social exclusion, while for others, it is a transient state following a temporary lack of social connections. However, feelings of loneliness are but one response to aloneness—that is, of not having social connections—experienced when the aloneness is involuntary or forced (Larsen 1990), resulting from social exclusion or isolation (Peplau and Perlman 1982). In contrast to this low-arousal, negatively valenced emotional response to social exclusion and lack of social connection, aloneness may also engender low-arousal, positively valenced feelings of calmness, restfulness, and peacefulness when the lack of social connection is voluntary (Bakhtiyorovna 2019; Dahlberg 2007). Voluntary aloneness, akin to solitude, is thus often considered a beneficial state (Larsen and Lee 1996). Moreover, whereas involuntary aloneness triggers perceptions that one’s life lacks meaning and purpose (Stillman et al. 2009; Twenge, Catanese, and Baumeister 2003), voluntary aloneness is associated with personal and spiritual growth (Long and Averill 2003).
Given the aversive implications of loneliness, consumers employ various strategies in response to its experience. Those who feel lonely often visit stores yet dislike self-service retailers that do not facilitate social contact (Forman and Sriram 1991). In fact, shopping malls are considered places to reduce loneliness among older consumers (Kim, Kang, and Kim 2005). As suggested by the label “retail therapy,” lonely consumers may also purchase gifts for themselves (Atalay and Meloy 2011). In addition to turning to consumption when feeling lonely, consumers may also resort to social mechanisms, such as trying to engage in relationships (Weiss 1974). Importantly, given that loneliness is often associated with the perception that one’s life is meaningless (Stillman et al. 2009; Tam and Chan 2019), we argue that activities that make life more meaningful might also reduce loneliness.
Meaning in Life
One’s life is perceived to be meaningful when it is understood to matter by the person living it, and when it has impact beyond the trivial or momentary—that is, when it is significant and has a clear sense of purpose (King et al. 2006; Morgan and Farsides 2009; Steger et al. 2006). Finding meaning in life is important given the positive correlation between experiencing one’s life to be meaningful and psychological well-being (Wong 1998). By experiencing greater meaning in life, people achieve greater happiness (King et al. 2006) and report reduced stress and depression (Mascaro and Rosen 2006). Meaning in life has been conceptualized both as a stable trait reflecting a global evaluation of one’s life and as a momentary state that can fluctuate daily (Choi, Capatano, and Choi 2017; King et al. 2006; Machell et al. 2015). Furthermore, momentary experiences of meaningfulness inform trait meaning in life (Kashdan and Steger 2007; Wilt, Bleidorn, and Revelle 2016).
Consumers can find meaning in life in many different ways, from significant ones such as by following a religion or life philosophy (Allport 1961) to seemingly trivial ones such as getting a pet fish (King and Hicks 2009). For instance, Rudow, Iacoviello, and Charney (2014) reported that liver donors experienced greater meaning in life after their donation, and indeed, helping others is often associated with experiencing a greater purpose in life (Schwartz et al. 2009). In addition, personal achievement and self-acceptance can be sources of meaning in life (Wong 1998), as can daily meditation (Steger and Frazier 2005) and even minor chores, such as cleaning or cooking (Sarvimäki and Stenbock-Hult 2000).
Examining factors that contribute to meaning in life also helps differentiate meaningfulness from happiness, despite the large overlap between the two. For example, Baumeister et al. (2013) showed that although feeling connected to others and not being alone contributes to both a happy life and a meaningful life, the former represents a natural quality concerning getting what one wants and needs, whereas the latter represents a cultural quality pertaining to doing things that express and reflect the self.
As we have discussed, experiencing meaning in life also has implications for loneliness, because social relationships greatly contribute to the experience of meaning in life (King et al. 2006; Stavrova and Luhman 2016). That is, one important determinant of meaning in life is the degree to which one maintains social relations. As such, experiencing a sense of belonging or feeling supported by family increases meaning in life (Baumeister et al. 2013), whereas being socially excluded or ignored reduces feelings of meaning in life (Twenge, Catanese, and Baumeister 2003). Yet, if loneliness engenders a perceived lack of meaning in life, then giving life meaning—even if only momentarily—could reduce feelings of loneliness. We thus investigate a novel way for consumers to make their lives more meaningful and less lonely: ritualistic consumption.
Rituals
We define rituals as a type of behavior, provided by marketers or created by consumers, made up of several steps performed in a fixed sequence with formality, rigidity, and repetition and imbued with a sense of meaning. In contrast to Rook (1985) and Hobson et al. (2018), and given our focus on seemingly minimal rituals, we do not expect them to be imbued with symbolic value (unlike, e.g., wedding rituals) or to be characterized by a lack of direct instrumental purpose in order to be meaningful. We reason that even novel and unfamiliar consumption rituals performed for the first time engender meaning, as do rituals with (e.g., preparing a cocktail by adding ingredients in a fixed sequence) and without (e.g., sorting M&M chocolates by color before eating them) direct instrumental purpose. Meaning is also what differentiates rituals from simple habits or routines (Hobson et al. 2018; Kapitany and Nielsen 2015; Rook 1985). Illustrating the difference between rituals and routines, Brooks et al. (2016) compare athletes’ pregame routines, such as stretching exercises to prevent injury, with pregame rituals, such as bouncing a basketball in a particular pattern for an exact number of times. Furthermore, while preparing tea in either a ritualized or routine manner likely achieves the goal of quenching one’s thirst, it is the former that leads Japanese consumers to appraise it as a meaningful experience (Minowa 2005).
Research on the individual psychology of rituals provides evidence that is consistent with the proposition that unfamiliar, minimal activities framed as rituals can influence consumer behavior. Indeed, Hobson et al. (2018, p. 7) recently argued that “framing basic actions as ritual can help regulate negative emotion and anxiety at least in part because of a person’s belief that rituals have some inherent meaning.” Thus, framing even the simplest behavioral steps in tea preparation as a “ritual” might suffice to produce a meaning appraisal that guides downstream consequences (Hobson et al. 2018). For instance, Norton and Gino (2014) showed that asking participants to follow a novel series of steps framed as a ritual reduced the experience of grief because doing so allowed them to regain feelings of control. Rituals composed of steps as seemingly insignificant as breaking a chocolate bar in half and unwrapping and eating the two halves separately can enhance food and drink enjoyment by increasing involvement (Vohs et al. 2013). Furthermore, children in after-school programs who participated in a novel necklace beading activity exhibited greater in-group affiliation when the activity was (vs. was not) labeled a ritual (Wen, Herrmann, and Legare 2016). In addition, Brooks et al. (2016) found that engaging in a minimal ritual—consisting of drawing a picture, sprinkling salt on it, counting to five, crinkling up the paper, and throwing it in the garbage—can improve performance by decreasing anxiety.
There are many types of rituals, ranging from the macro level, comprising holidays that draw on shared cultural values, to the micro level, involving individuals, dyads, or families engaging in idiosyncratic ritualistic behavior (Rook 1985). Rituals have also long been understood as ways to facilitate relationships between consumers and their brands as well as among consumers themselves. In addition to consciousness of kind—the intrinsic connection members of a community feel toward each other—and a sense of moral responsibility, shared rituals and traditions related to consumption experiences with brands are characteristic of brand communities (Muñiz and O’Guinn 2001). For example, Jeep Wrangler drivers give the “Wrangler wave” when passing other Wrangler drivers on the road, which helps build relationships with the vehicle and with members of the brand community (McAlexander, Schouten, and Koenig 2002).
Clearly, rituals serve important functions and are associated with perceived positive consequences, which research has documented (for a review, see Otnes et al. [2018]). Rituals can also foster feelings of collective strength and worth (Carlton-Ford 1992) and enhance self-esteem (Fiese and Kline 1993) or family closeness during the holidays (Sezer et al. 2016). More recently, Tian et al. (2018) found that rituals can enhance self-control, increasing choices of healthy over unhealthy options. In addition, and more important for the current project, rituals also affect product meaningfulness; they may not only sustain or heighten the existing meaning of products or brands but also impart them with new meaning (Fournier 1998; McCracken 1986). For instance, consumers may imbue sneakers or sports drinks with meaning using a running ritual, or soap using skin care rituals (Fournier 1998). In turn, products rendered meaningful via rituals might add much-needed meaning to people’s lives—that is, for consumers experiencing loneliness.
In summary, we argue that lonely individuals may engage in consumption rituals involving consumer products that add meaning to their lives via the meaning imparted to the products, which in turn reduces loneliness. Furthermore, if lonely consumers have an alternative source from which they can derive meaning (i.e., their lives are no longer lacking in meaning), then ritualistic behavior should no longer affect their loneliness. Thus, rituals should only reduce loneliness in the absence of incidentally derived meaningfulness. We began our investigation by examining the association between chronic loneliness and frequency of engaging in consumption rituals. Study 1 also aimed to provide preliminary evidence for the role of meaning in life, and—with a different set of participants—investigated the relationship between the extent to which the consumption behaviors described by respondents in the first part of the study are indeed perceived to be ritualistic and the degree to which they reflect ritualistic characteristics.
Each of our main experiments then consisted of three parts disguised as ostensibly unrelated studies (for our conceptual model, see Figure 1; for study designs, see Table 1). In the first part, we manipulated participants’ affective state to temporarily elicit either loneliness versus a pure control condition (i.e., no affect prime; Studies 2 and 4) or loneliness versus calmness (Studies 3 and 5). Priming loneliness versus calmness enabled us to compare the differential impact of rituals on involuntary aloneness (in which a person lacks desired social connections, engendering feelings of loneliness) with voluntary aloneness (in which a person rejects the companionship of others, engendering feelings of calmness and peacefulness) (Dahlberg 2007; Larson 1990).
Conceptual model.
Outline of Study Designs.
Our affect prime manipulations were based on recall tasks, given their successful use in prior research studying affective consequences of real-world experiences (Van Boven and Gilovich 2003). Those in the ritual-present conditions were then instructed to simulate engaging in behaviors explicitly framed as rituals (Studies 2, 4, and 5) or to actually engage in behaviors not explicitly framed as a ritual (Study 3) involving a consumer product, whereas those in the ritual-absent conditions consumed the product as they usually did, randomly, or in any way they wished (Studies 2, 3, and 4). Alternatively, they engaged in activities identical to those of their ritual-present counterparts, but the activities were framed as random actions (Study 5).
Following the ritual manipulation part of each study, we assessed how lonely participants felt (all studies) as our main dependent variable and how intent they were on purchasing the product included in the ritual (Study 5). Support for the underlying process was provided by mediation analyses using measurements of meaningfulness (Studies 1, 3, and 4) and based on moderation analysis using a manipulation of meaningfulness (Study 5). Subject to time and money constraints, we aimed for at least 100 participants per cell for studies conducted online and 30 per cell for studies conducted in our lab. We did not exclude participants in our analyses unless otherwise noted. In our lab study with Chinese participants, we relied on “仪式” as the common translation of “ritual.” For detailed instructions of all studies, see Web Appendix W1.
Study 1
Method
Participants and procedure
Two hundred fifty-four participants, recruited via Amazon’s Mechanical Turk (MTurk; Mage = 40.63 years, SD = 14.11 years; 43.3% female), completed a series of supposedly unrelated studies. They first reported their chronic loneliness by indicating how often each of 16 items described how they felt (1 = “never,” and 7 = “always”; reverse-coded; adapted from Russell 1996; α = .96; for the complete scale, see Web Appendix W3a). In the introduction to the ostensible next study, participants read that consumers sometimes create rituals when consuming products, and companies may even suggest rituals for consumers to follow (e.g., the three-step twist-lick-dunk ritual when eating an Oreo cookie). They were first asked to describe a specific example of a consumption ritual with which they were familiar and then to indicate how often they engaged in this ritual (1 = “never,” and 7 = “always”), which served as our consumption ritual frequency measure. Finally, they rated their sense of meaning in life (Morgan and Farsides 2009; α = .96) by marking the degree to which they felt their life had a clear sense of purpose, was meaningful, was significant, was worthwhile, and interested and excited them (1 = “strongly disagree,” and 7 = “strongly agree”).
Results
There was a significant negative correlation between participants’ loneliness and the frequency with which they engaged in the consumption rituals they had identified (r = −.35, p < .01). Furthermore, consumption ritual frequency was positively correlated with meaning in life (r = .28, p < .01), indicating that those who engaged in rituals more often experienced greater meaning in life. The correlation between meaning in life and loneliness was also significant (r = −.72, p < .01): greater meaning in life was related to reduced loneliness. Finally, bootstrapping using 5,000 resamples with replacement (Hayes 2013, Model 4) showed that meaning in life mediated the relationship between consumption ritual frequency and loneliness (95% bias-corrected confidence interval [CI]: [−.20, −.07]).
In a follow-up study (preregistered at https://aspredicted.org/vn5vm.pdf), a separate sample of 469 MTurk panelists (Mage = 33.32 years, SD = 11.89 years; 56% female)—divided at random into nine groups to minimize wear-out, with each group receiving a subset of 25 or 26 consumption rituals identified by the prior respondents—was asked to evaluate the prior respondents’ descriptions of consumption rituals, both in terms of the degree to which they perceived them to be ritualistic and the degree to which they possessed characteristics of rituals. In particular, participants were first instructed to indicate (α = .77; 1 = “not at all,” and 7 = “very much”) how much the described consumption behavior sounded like a ritual, sounded ritualized, and sounded like random actions (reverse-scored). Next, they rated (α = .68; 1 = “not at all,” and 7 = “very much”) the degree to which the consumption behaviors described by the prior respondents consisted of predefined sequences characterized by rigidity, formality, and repetition and the degree to which they could be interpreted as having meaning to the prior respondents. To examine whether consumption rituals are imbued with symbolic value and do not serve a direct instrumental purpose, we also asked them to evaluate the degree to which the rituals could be interpreted as having symbolic value to the prior respondents and the degree to which they contained elements that had no direct instrumental purpose.
Overall, participants indicated that the consumption rituals described by the prior respondents were indeed ritualistic (M = 4.65, SD = .75; t(228) = 13.10, p < .001; compared with the scale midpoint). Furthermore, they also interpreted the described consumption rituals as consisting of predefined sequences that are rigid, formal, and repetitive (M = 4.48, SD = .76; t(228) = 9.51, p < .001) and as having meaning to the prior respondents (M = 4.29, SD = .71; t(228) = 6.19, p < .001). The described consumption behaviors did not differ from the scale midpoint in the degree to which they were regarded as having symbolic value to the prior respondents (M = 4.08, SD = .71; t(228) = 1.61, p = .11). Furthermore, the rituals were judged to have some direct instrumental purpose (M = 3.86, SD = .49; t(228) = −4.45, p < .001).
Furthermore, indicating its predominant role, the ritual characteristic of reflecting predefined sequences that are rigid, formal, and repetitive (M = 4.48, SD = .76) was significantly more representative of the identified consumption rituals than the other three qualities (having meaning to the prior respondents: M = 4.29, SD = .71; t(228) = 6.30, p < 001; having symbolic value to the prior respondents: M = 4.08, SD = .71; t(228) = 12.12, p < .001; not having a direct instrumental purpose: M = 3.86, SD = .49; t(228) = 9.73, p < .001). A principal component analysis on the average ratings of the four characteristics further showed an eigenvalue greater than 1 (i.e., 2.721) only for the dimension of reflecting predefined sequences that are rigid, formal, and repetitive, explaining 68.03% of variance, and an eigenvalue close to 1 (i.e., .967) for the dimension of having meaning to the prior respondents, explaining an additional 24.18% of variance. For examples of the consumption rituals that scored highly on each of the dimensions, as well as a list of all consumption rituals and their descriptive statistics, see Web Appendices W2a and W2b.
We next examined the correlations between the degree to which the prior respondents’ consumption rituals were identified as being ritualistic and the extent to which they possessed characteristics of rituals. The correlation coefficients were significant and positive between the ritualistic identification and the degree to which they consisted of predefined sequences characterized by rigidity, formality, and repetition (r = .55, p < .01) and the degree to which they could be interpreted as having meaning for the prior respondents (r = .42, p < .01). The correlations were also significant for having symbolic value for the prior respondents (r = .34, p < .01) and the degree to which they contained elements that had no direct instrumental purpose (r = −.21, p < .01).
Although the first part of this study thus provided initial evidence of the associations between frequency of ritualistic behavior, meaning in life, and chronic loneliness, its correlational design had obvious limitations. The experiments described next made causal inferences possible through manipulations of loneliness and the presence versus absence of ritualistic behavior. Furthermore, Study 1 found that reflecting predefined sequences characterized by rigidity, formality, and repetition explained the largest share of variance among the described rituals. However, and important for our research, we also found evidence in the identified consumption rituals, in addition to those related to religious or secular holidays and celebrations, even more mundane rituals (e.g., eating an Oreo before a sports game, toasting with a glass of wine or beer, sorting M&M’s by color, preparing tea) may have meaning to consumers.
Study 2
Study 2 aimed to provide initial causal evidence for our proposition regarding the relationship between engaging in rituals and loneliness. Specifically, we tested if participants who had been primed with loneliness would feel less lonely following the mental simulation of eating an Oreo cookie using the familiar twist-lick-dunk ritual associated with consuming Oreos, compared with those who simulated eating the Oreo in the way they usually did. We expected that presence versus absence of the ritual simulation would affect participants primed with loneliness, but not their counterparts in a control condition without the affect prime, because participants in the latter condition should be less likely to perceive that their life lacks meaning.
Five hundred fifty-one MTurk panelists (Mage = 38.59 years, SD = 12.23 years; 54.1% female) were assigned at random to conditions in a 2 (affect prime: lonely vs. control) × 2 (ritual frame: present vs. absent) between-subjects design. We eliminated 54 participants assigned to the ritual-absent conditions who indicated that they usually ate Oreos employing the three-step twist-lick-dunk ritual (thus, in effect following the ritual frame instructions), leaving a final sample of 497 (Mage = 38.46 years, SD = 12.41 years; 53.7% female). Participants first completed an affect priming task based on Maner et al. (2007), which instructed those who were primed to be lonely to take several minutes to recall and describe an event that had made them feel lonely, whereas those in the control condition did not carry out this task. Both groups then completed a six-item affect prime manipulation check (α = .91) by indicating how much companionship they were feeling (reverse-coded; 1 = “not at all,” and 7 = “very much”), and how lonely, abandoned, lost, isolated, and lonesome they were feeling (1 = “not at all,” and 7 = “very much”).
Subsequently, using a mental simulation task (Escalas and Luce 2004) participants were asked to mentally visualize and simulate that they had bought themselves some Oreo cookies. Those in the ritual-frame-present conditions were told to imagine that they had decided to follow the Oreo ritual to eat one and to engage in ritualized steps that were described as follows below the picture of an Oreo cookie: “Step 1 of the Oreo ritual: Twist. With each half of the Oreo in your fingertips, you begin to smoothly rotate your hands in opposite directions; Step 2 of the Oreo ritual: Lick. You lick the filling off; Step 3 of the Oreo ritual: Dunk. You dunk the cookie in milk and eat it.” In contrast, participants in the ritual-frame-absent conditions were asked to imagine eating the Oreo in the way they usually do.
Following the ritual frame manipulation, participants indicated how lonely they were feeling, which constituted our dependent variable. In particular, similar to participants in study 1, they marked their level of agreement (1 = “strongly disagree,” and 7 = “strongly agree”; reverse-coded) with each of 16 statements (adapted from Russell [1996]; α = .96; see Web Appendix W3b). As a ritual manipulation check, all participants then rated how much simulating the behaviors involved in eating the Oreo cookie had felt like a ritual and the extent to which they thought that the eating of the Oreo cookie was ritualized (1 = “not at all,” and 7 = “very much”; r = .86, p < .01). To explore the degree to which the Oreo ritual was familiar to participants, those in the ritual-frame-present condition also indicated how well-known, how new (reverse-coded), and how novel (reverse-coded) the ritual they had been asked to visualize was and how familiar they were with the ritual (α = .68). Subsequently, participants indicated if they usually ate Oreos using the twist, lick, and dunk ritual versus randomly as they liked. Finally, they provided demographic information and were debriefed.
Results
Manipulation checks
A 2 (affect prime) × 2 (ritual frame) analysis of variance (ANOVA) on the loneliness prime manipulation check revealed that participants reported higher loneliness in the loneliness prime (M = 3.44, SD = 1.67) versus control (M = 2.69, SD = 1.54) condition (F(1, 493) = 28.36, p < .001). No other effects were significant. Furthermore, the steps involved in consuming the Oreo were perceived to be more like a ritual in the ritual-frame-present condition (M = 4.72, SD = 1.75) than in the ritual-frame-absent condition (M = 3.35, SD = 1.76; F(1, 493) = 69.07, p < .001). No other effects were significant. Finally, those in the ritual-frame-present condition rated their familiarity with the three-step ritual of eating Oreos (M = 5.07, SD = 1.25) significantly above the scale midpoint (t(285) = 14.45, p < .001), demonstrating familiarity with the ritual.
Loneliness
A 2 (affect prime) × 2 (ritual frame) ANOVA on how lonely participants felt after the ritual frame manipulation yielded a main effect of affect prime (F(1, 493) = 6.46, p < .05), such that, unsurprisingly, participants still reported greater loneliness when they initially had been primed to feel lonely (M = 3.39, SD = 1.37) as compared with their counterparts in the control condition (M = 3.12, SD = 1.28). The main effect of ritual frame was also significant (F(1, 493) = 8.39, p < .01), such that participants were less lonely in the ritual-frame-present (M = 3.10, SD = 1.21) versus ritual-frame-absent (M = 3.42, SD = 1.45) condition. More importantly, analyses revealed an interaction effect of affect prime and ritual frame (F(1, 493) = 3.43, p = .065). In particular, among those who initially had been primed to feel lonely, the presence (M = 3.15, SD = 1.24) versus absence (M = 3.72, SD = 1.47) of the ritual frame significantly reduced their loneliness (F(1, 493) = 8.41, p < .01). However, the presence versus absence of the ritual frame did not affect the loneliness of participants in the control condition (M = 3.07, SD = 1.19 vs. M = 3.19, SD = 1.39 for the ritual-frame-present vs. -absent condition, respectively; F(1, 493) = 1.15, p = .28). Furthermore, in the absence of the ritual frame, participants felt lonelier when they had been primed to feel lonely (M = 3.72, SD = 1.47) compared with the control condition (M = 3.19, SD = 1.39; F(1, 493) = 6.93, p < .01). However, participants felt equally lonely in the control (M = 3.07, SD = 1.19) and loneliness prime (M = 3.15, SD = 1.24; F(1, 493) = .68, p = .41) conditions in the presence of the ritual frame.
Discussion
Study 2 provided causal evidence for the effect of ritualistic behavior on loneliness. That is, participants who had been primed with loneliness felt less lonely after mentally simulating the three-step Oreo ritual compared with their counterparts who simulated eating an Oreo cookie in their usual way. Furthermore, the presence versus absence of the ritual frame affected loneliness only among those who initially had been primed to feel lonely; loneliness among those in the control condition did not differ based on the presence versus absence of the ritual frame.
Study 3 aimed to extend these results in important ways. The first limitation of Study 2 was that we assessed participants’ loneliness after their having been instructed to mentally simulate engaging in (rather than actually engaging in) a ritual—even though research has shown that the effect of mentally simulating a behavior often mirrors that of actually engaging in the behavior (Bargh, Chen, and Burrows 1996). Thus, in our next study participants were asked to actually engage in prescribed behavioral steps. Second, even though results of this study were consistent with our expectations, we did not assess meaning in life in this study, and thus we could not examine the underlying process. Study 3 therefore also aimed to address the notion of meaningfulness of the ritual.
Third, our ritual manipulation explicitly framed the consumption behavior “ritual” in the instructions (e.g., “Step 1 of the Oreo ritual”), which gives rise to an alternative explanation based on priming. Therefore, in our next study, we aimed to replicate the effect without explicitly framing the behaviors as a ritual. Study 3 also examined the effect of consumer rituals on the distinct affective consequences associated with the two types of aloneness by priming participants with either loneliness (involuntary aloneness response) or calmness (voluntary aloneness response). Given that only the former is associated with a lack of meaning in life, we expected to replicate our results only among participants primed with loneliness.
Study 3
Method
One hundred twenty-four undergraduate students (Mage = 20.08 years, SD = 1.89 years; 71% female) from a Chinese university participated individually in the study in exchange for compensation and were randomly assigned to conditions in a 2 (affect prime: lonely vs. calm) × 2 (tacit ritual frame: present vs. absent) between-subjects design. Two participants did not complete the study and one participant did not follow instructions, so they were excluded from further analysis, leaving a final sample of 121 participants (Mage = 20.09 years, SD = 1.91 years; 71.1% female). The loneliness prime was identical to that of Study 2, whereas those primed with calmness were asked to recall and describe an event that had made them feel calm. This was followed by the same affect manipulation check items as in Study 2 (α = .96).
Next, participants were asked to prepare an actual cup of milk tea in the lab following instructions we provided, which constituted our ritual-frame manipulation. In the tacit-ritual-frame-present conditions, participants were asked to follow specific steps to prepare the milk tea; namely, to pour half a packet of milk tea into a glass, pour in enough water to fill half the glass, stir the mixture, use their knuckles to knock on the desk three times, take three deep breaths, and close their eyes to meditate for 30 seconds. Then, they were told to pour the remainder of the powder into the glass, pour in enough water to top off the glass, stir, use their knuckles to knock on the desk three times, take three deep breaths, and close their eyes to meditate for another 30 seconds. Lastly, they were asked to drink the milk tea. Conversely, in the tacit-ritual-frame-absent conditions, they were simply instructed to prepare the milk tea in any way they wished and then drink it. To ensure that participants spent as much time preparing the tea in this condition as did those who performed the ritual, they were asked to wait one minute before consuming the milk tea, during which time they could do whatever they liked.
After the ritual frame manipulation and consumption of the tea, participants rated their meaning in life as in Study 1 (Morgan and Farsides 2009; α = .95). Subsequently, they indicated how lonely they were currently feeling as in the previous study (adapted from Russell [1996]; reverse-coded; α = .97). Next, they completed a ritual manipulation check (1 = “not at all,” and 7 = “very much”; α = .95) by rating how much the activities involved in making the milk tea felt like a ritual, the extent to which they thought the activities involved in making the milk tea were ritualized, and the extent to which the activities involved in making the milk tea felt like random actions (reverse-scored).
Next, we assessed several possible alternative explanations. First, research has shown that engaging in rituals increases feelings of control (Norton and Gino 2014), which may affect participants’ loneliness. We therefore asked participants to what extent they felt a sense of control toward what happens in their life (1 = “not at all,” and 7 = “very much”; Norton and Gino 2014). In addition, consuming milk tea using ritualistic steps, rather than randomly, might also simply be more fun; thus, we instructed participants to indicate the degree to which drinking the milk tea was fun (1 = “not at all,” and 7 = “very much”). To ascertain that the one-minute waiting period for participants in the tacit-ritual-frame-absent conditions did not negatively affect the enjoyment of the experience relative to those who had been instructed to engage in the ritual, they also rated the extent to which they enjoyed the milk tea experience, enjoyed every sip of the milk tea, and enjoyed the milk tea itself (1 = “strongly disagree,” and 7 = “strongly agree”; α = .93). Finally, they rated how effortful and difficult they thought making the milk tea had been (1 = “not at all,” and 7 = “very much”; r = .75, p < .01), and provided demographic information. Study materials were translated and back-translated for use in China.
Results
Manipulation checks
A 2 (affect prime) × 2 (tacit ritual frame) ANOVA on the loneliness prime manipulation check and the ritual manipulation check revealed successful manipulations. Specifically, participants reported higher loneliness in the lonely versus calm prime conditions (M = 5.45, SD = .80 vs. M = 2.00, SD = .72, respectively; F(1, 117) = 623.22, p < .001). No other effects were significant. Furthermore, the activities involved in making the milk tea were more likely to be rated to be a ritual in the tacit-ritual-frame-present (M = 5.78, SD = 1.04) versus -absent (M = 2.33, SD = .92; F(1, 117) = 371.69, p < .001) conditions. No other effects were significant.
Loneliness
A 2 (affect prime) × 2 (tacit ritual frame) ANOVA on loneliness yielded a main effect of affect prime (F(1, 117) = 64.53, p < .001), such that participants who initially had been primed with loneliness reported feeling lonelier (M = 3.84, SD = 1.35) than those who had been primed to feel calm (M = 2.42, SD = .79). The main effect of ritual frame was also significant (F(1, 117) = 23.54, p < .001), revealing that those who engaged in the ritual-framed steps when preparing the milk tea reported lower loneliness (M = 2.70, SD = 1.02) than those who prepared the milk tea in any way they wanted (M = 3.56, SD = 1.43). Importantly, as Figure 2, Panel A, shows, the interaction between the two factors was significant (F(1, 117) = 16.82, p < .001). Specifically, when participants initially had been primed with loneliness, the ritualized manner of preparing the milk tea significantly reduced how lonely they felt, compared with those in the tacit-ritual-frame-absent condition (M = 3.04, SD = 1.07 vs. M = 4.60, SD = 1.13, respectively; F(1, 117) = 41.59, p < .001). There was no difference in loneliness due to the presence versus absence of the ritual frame, however, among those in the calm condition (M = 2.35, SD = .84 vs. M = 2.48, SD = .74, respectively; F(1, 117) = .28, p = .60).
Study 3: Interactive effect of affect and tacit ritual frame on loneliness and meaning of life.
Next, we conducted a series of four analyses of covariance (ANCOVAs) that included feelings of control, fun, enjoyment, and effort as covariates. However, in all cases except enjoyment, the covariate was not significant (feelings of control: F(1, 116) = 1.96, p = .16; fun: F(1, 116) = 1.16, p = .28; effort: F(1, 116) = .002, p = .96), whereas the affect prime × tacit ritual frame interaction remained significant (feelings of control: F(1, 116) = 14.50, p < .001; fun: F(1, 116) = 17.58, p < .001; effort: F(1, 116) = 16.59, p < .001). Although enjoyment was a significant predictor of loneliness (F(1, 116) = 11.97, p = .001), the affect prime × tacit ritual frame interaction remained significant (F(1, 116) = 19.40, p < .001). Furthermore, an ANCOVA including all four covariates simultaneously, rather than sequentially, also showed that the affect prime × tacit ritual frame interaction remained significant (F(1, 113) = 17.70, p < .001), demonstrating that the interactive effect on loneliness was independent of potential differences in feelings of control, fun, enjoyment, and effort.
Finally, narrowing down the analysis to the loneliness conditions only and conducting ANCOVAs with ritual manipulation as the independent variable; loneliness as the dependent variable; and feelings of control, fun, or effort as covariates (feelings of control: F(1, 58) = 2.58, p = .11; fun: F(1, 58) = .24, p = .63; effort: F(1, 58) = .20, p = .65), the effect of ritual frame remained significant (feelings of control: F(1, 58) = 24.92, p < .001; fun: F(1, 58) = 29.71, p < .001; effort: F(1, 58) = 28.29, p < .001). Finally, although enjoyment predicted loneliness (F(1, 58) = 9.23, p < .01), the effect of ritual frame remained significant (F(1, 58) = 38.06, p < .001).
Meaning in life
We conducted a similar analysis on meaning in life. An ANOVA revealed a main effect of affect prime, such that participants who initially had been primed with loneliness reported lower meaning in life (M = 4.19, SD = 1.61) than those who had been primed with calmness (M = 5.51, SD = 1.05; F(1, 117) = 39.38, p < .001), which provides evidence that even transient loneliness is characterized by a lack of meaning in life. The main effect of ritual frame was also significant, showing that participants in the tacit-ritual-frame-present condition (M = 5.32, SD = 1.13) reported experiencing greater meaning in life than those in the tacit-ritual-frame-absent condition (M = 4.38, SD = 1.70; F(1, 117) = 19.56, p < .001). Importantly, as Figure 2, Panel B, shows, a significant interaction emerged (F(1, 117) = 31.05, p < .001), such that life was rated as more meaningful among those who had been primed to feel lonely in the tacit-ritual-frame-present (M = 5.25, SD = 1.12) versus -absent (M = 3.16, SD = 1.34) condition (F(1, 117) = 51.39, p < .001). However, meaning in life did not differ according to the presence versus absence of the ritual frame among those primed with calmness (M = 5.39, SD = 1.15 vs. M = 5.63, SD = .95) for the tacit ritual-frame-present vs. -absent conditions, respectively (F(1, 117) = .66, p = .42).
Mediation analysis
To examine whether differences in meaning in life guided the interactive effect of affect prime and ritual frame on loneliness, we followed Hayes’s (2013; Model 8) bootstrapping procedure using 5,000 resamples with replacement. As we expected, meaning in life mediated the affect prime × ritual frame interaction on loneliness (95% bias-corrected CI: [−1.84, −.79]). Specifically, when initially primed to feel lonely, meaning in life mediated the effect of the ritual frame on loneliness (95% bias-corrected CI: [−1.61, −.68]). However, we did not obtain this mediation effect among those in the calmness condition (95% bias-corrected CI: [−.15, .42]). In addition, when meaning in life and feelings of control, fun, effort, and enjoyment were included as parallel mediators in the mediation analysis (Model 8, Hayes 2013), results revealed only meaning in life to be a significant mediator (95% bias-corrected CI: [−1.75, −.70]); however, feelings of control (95% bias-corrected CI: [−.19. .06]), fun (95% bias-corrected CI: [−.03, .15]), effort (95% bias-corrected CI: [−.04, .12]), and enjoyment (95% bias-corrected CI: [−.07, .21]) did not reach significance.
Discussion
Study 3 replicated the results of our previous study, such that actually engaging in the preparation of milk tea using steps not explicitly framed as ritualized reduced loneliness among participants who had been primed to feel lonely, but not among those primed to feel calm, and this effect was independent of potential differences in feelings of control, fun, enjoyment, and effort. Note that in contrast to Study 2, the ritual we asked participants to engage in was one we constructed and thus was unfamiliar to them, showing that even novel rituals can lead to meaning. Replicating the effect without overtly framing rituals in the study instructions also casts doubt on an explanation based on conceptual priming. Moreover, consistent with our theory, we demonstrated that the interaction effect of affect prime and ritual frame on loneliness was mediated by how meaningful participants experienced their life to be following the tacit ritual.
However, a potential confound in this study was that the last step of the activity for participants in the ritual-frame-present condition included meditation. Although meditation is a type of privately conducted ritual (Marshall 2002) and thus may simply have strengthened our ritual manipulation, it may reduce loneliness in and of itself, in ways other than via increased meaning in life. Thus, participants in the ritual-frame-present condition in the next study visualized ritualistic activities that did not include meditation. We also wanted to extend our findings from relatively indulgent products (i.e., cookies and milk tea) to a more virtuous one (i.e., lemon tea), because rituals involving only hedonic or indulgent consumption are unlikely to increase consumer well-being in the long term.
Furthermore, Study 4 aimed to investigate whether consumption rituals may imbue products with meaning (Fournier 1998; McCracken 1986), in turn adding meaning to consumers’ lives and reducing loneliness. In addition, we wanted to cast doubt on involvement as an alternative mechanism driving the effect of ritualistic behavior on loneliness. Because engaging in behavior described as ritualistic may affect the degree to which attention is drawn to the activity (Brooks et al. 2016), which could then provide a distraction from loneliness, we also measured participants’ mindfulness in the next study. Moreover, one might argue that performing activities described as ritualistic engenders greater immersion, leading participants to become “lost” (Green and Brock 2000) in the ritual, thereby reducing loneliness. Thus, we assessed narrative transportation as well.
Study 4
Method
In exchange for compensation, 656 Prolific panelists (Mage = 33.04 years, SD = 12.96 years; 51.5% female) were recruited for a study (preregistered at https://aspredicted.org/q3x63.pdf) and assigned at random to conditions in a 2 (affect prime: lonely vs. control) × 2 (ritual frame: present vs. absent) between-subjects design. We eliminated 43 participants in the ritual-frame-absent conditions who stated that they always drank lemon tea using a ritual and 20 duplicate responses based on their IP address, leaving a final sample of 593 (Mage = 33.17 years, SD = 13.23 years; 52.3% female). Those in the loneliness prime condition first completed the affect priming task as in our previous studies, whereas those in the control condition did not complete this task. Both groups then completed the same six-item affect prime manipulation check as in the prior studies (α = .92). Those in the loneliness prime condition were then asked to imagine vividly that the event that made them feel lonely had just happened that day and that they were buying themselves a cup of lemon tea in a store. In contrast, in the control condition, participants were simply asked to imagine that they were buying themselves a cup of lemon tea in a store.
Next, the scenario for those in the ritual-frame-present condition continued, “After you have bought yourself the tea and sat down, you see a sign that suggests drinking the tea using a ritual, and you decide to follow the suggested ritualistic steps in drinking the tea. The steps of the ritual are listed below.” Then a picture of a cup of tea and two pieces of lemon were shown, followed by a series of ritual steps: “Step 1: Take a sip of the tea; Step 2: Put a piece of lemon into the cup; Step 3: Stir the tea clockwise three times with the spoon to let the lemon mix with the tea; Step 4: Take another sip of the tea; Step 5: Put the other piece of lemon into the cup; Step 6: Stir the tea counter-clockwise three times with the spoon to let the lemon mix with the tea; Step 7: Enjoy the tea with a full mouth!” In the ritual-frame-absent condition, they were just told to imagine that after they had bought themselves the tea, sat down, and had decided to just drink it in a random way, as they always did. In addition, we included two exploratory conditions (included in Web Appendix W1) to shed preliminary light on the potential role of the presence of others in the effect of ritualistic behavior on loneliness, analyses for which are reported in Web Appendix W6.
Following the ritual manipulation, participants reported their loneliness (adapted from Russell 1996; reverse-coded; α = .96) and meaning in life (Morgan and Farsides 2009; α = .94) as before. We then measured product meaningfulness by asking participants to indicate the degree to which the cup of lemon tea had no special meaning to them (reversed-coded; 1 = “strongly disagree,” and 7 = “strongly agree”). Subsequently, as a manipulation check for the ritual manipulation, participants indicated how much the activities involved in drinking the lemon tea felt like a ritual and the extent to which they thought that the tea-drinking experience was ritualized (1 = “not at all,” and 7 = “very much”; r = .82; p < .01). We then asked them to recall how they had been instructed to imagine drinking the lemon tea in the prior task by choosing between two options: following a ritual by themselves versus by drinking it in a random way as they always did.
To examine potential confounds, we also measured mindfulness (adapted from Tanay and Bernstein [2013]; see Web Appendix W4; α = .95), narrative transportation (adapted from Green and Brock [2000]; see Web Appendix W5; α = .66), and involvement. The latter was assessed with the following three items adapted from Vohs et al. (2013): “Imagining drinking the lemon tea bored me” (reverse-coded), “Imagining drinking the lemon tea was fun,” and “Imagining drinking the lemon tea was intrinsically interesting” (1 = “not at all,” and 7 = “very much”; α = .85). To ascertain that lemon tea was indeed considered a relatively virtuous product rather than an indulgence, we had participants rate the degree to which they considered lemon tea to be pleasurable, tempting, and sinful (Verma, Guha, and Biswas 2016; 1 = “strongly disagree,” and 7 = “strongly agree”; α = .65). Finally, they indicated how they usually drank lemon tea by marking whether they always followed a ritual that they designed themselves versus randomly (i.e., any way they felt like) or if they never drank lemon tea. They then provided demographic information.
Results
Ancillary measures
We first examined alternative explanations based on mindfulness, narrative transportation, and involvement. A 2 (affect prime) × 2 (ritual frame) ANOVA on mindfulness did not yield any significant main or interaction effects (all ps > .17), whereas an ANOVA on narrative transportation only found a main effect of affect prime (F(1, 589) = 4.42, p < .05). Thus, in both cases the affect prime × ritual frame was not significant, casting doubt on these two constructs as alternative explanations for the effect, so we do not discuss them further. However, although no main effects reached significance, an ANOVA did find a marginally significant interaction on involvement (F(1, 589) = 2.85, p < .10); thus, we included involvement as covariate in the following analyses.
Manipulation checks
As expected, participants rated lemon tea (M = 3.12, SD = 1.27) significantly below the scale midpoint (t(592) = −16.71, p < .001), confirming that the target product was perceived to be relatively virtuous. Next, a 2 (affect prime) × 2 (ritual frame) ANCOVA with involvement as covariate on the loneliness prime manipulation check showed that participants reported higher loneliness in the loneliness prime versus control condition (M = 3.43, SD = 1.74 vs. M = 3.14, SD = 1.50, respectively; F(1, 588) = 4.77, p < .05). The effect of the involvement covariate was marginally significant (F(1, 588) = 2.93, p < .10). No other effects were significant. Next, the activities involved in drinking the lemon tea were rated to be more ritualized in the ritual-frame-present (M = 4.58, SD = 1.68) versus ritual-frame-absent (M = 3.17, SD = 1.77; F(1, 588) = 102.98, p < .001) condition. The effect of the involvement covariate was also significant (F(1, 588) = 76.52, p < .001). No other effects were significant. Finally, 91.7% of participants in the ritual-frame-absent condition and 94.2% in the ritual-frame-present condition could correctly recall the drinking instructions in their respective conditions (χ2 = 480.63, p < .001).
Loneliness
A 2 (affect prime) × 2 (ritual frame) ANCOVA with involvement as a covariate revealed a significant main effect of ritual (F(1, 588) = 8.27, p < .01), such that those in the ritual-frame-absent condition (M = 3.79, SD = 1.28) felt significantly greater loneliness than those in the ritual-frame-present condition (M = 3.47, SD = 1.40). The main effect for the loneliness manipulation was also significant (F(1, 588) = 5.34, p < .05); those primed with loneliness (M = 3.77, SD = 1.39) still felt lonelier than their counterparts in the control condition (M = 3.51, SD = 1.31). The involvement covariate was also significant (F1, 588) = 25.33, p < .001). In addition, the two-way interaction between affect prime and ritual frame on loneliness was significant (F(1, 588) = 3.87, p = .05). Specifically, as we show in Figure 3, Panel A, among participants who initially had been primed with loneliness, those in the ritual-frame-present condition reported significantly lower loneliness than those in the ritual-frame-absent condition (M = 3.51, SD = 1.44 vs. M = 4.03, SD = 1.29, respectively; F(1, 588) = 10.70, p < .01), replicating our previous results. The difference in loneliness due to the presence versus absence of the ritual frame, however, was not evident among participants in the control condition (M = 3.43, SD = 1.38 vs. M = 3.60, SD = 1.25 respectively; F(1, 588) = .45, p = .50). Furthermore, in the ritual-frame-absent conditions, participants who had been primed to feel lonely still felt lonelier than their counterparts in the control condition (M = 4.03, SD = 1.29 vs. M = 3.60, SD = 1.25, respectively; F(1, 588) = 9.32, p < .01). As we expected, the difference in loneliness was not significant among participants in the ritual-frame-present condition when drinking the tea (M = 3.51, SD = 1.44 vs. M = 3.43, SD = 1.38 for the lonely vs. control conditions, respectively; F(1, 588) = .06, p = .81).
Study 4: Interactive effect of affect and ritual frame on loneliness, meaning in life, and product meaningfulness.
Meaning in life
A 2 (affect prime) × 2 (ritual frame) ANCOVA revealed a significant main effect of the involvement covariate (F(1, 588) = 22.20, p < .001). More importantly, the two-way interaction between affect prime and ritual frame on meaning in life was marginally significant (F(1, 588) = 3.38, p = .066). As we show in Figure 3, Panel B, in the ritual-frame-present conditions, participants who had been primed to feel lonely reported marginally greater meaning in life than their counterparts in the control condition (M = 4.96, SD = 1.58 vs. M = 4.66, SD = 1.62, respectively; F(1, 588) = 3.69, p = .055). As expected, the difference in meaningfulness was not significant among participants in the ritual-frame-absent condition (M = 4.68, SD = 1.57 vs. M = 4.76, SD = 1.50 for the lonely vs. control conditions, respectively; F(1, 588) = .45, p = .50).
Product meaningfulness
A 2 (affect prime) × 2 (ritual frame) ANCOVA revealed a significant main effect of the involvement covariate (F(1, 588) = 235.86, p < .001). More importantly, the two-way interaction between affect prime and ritual frame on product meaningfulness was significant (F(1, 588) = 8.86, p < .01). As we show in Figure 3, Panel C, in the ritual-frame-present conditions, participants who had been primed to feel lonely reported greater product meaningfulness than their counterparts in the control condition (M = 3.34, SD = 1.97 vs. M = 2.97, SD = 1.95, respectively; F(1, 588) = 7.20, p < .01). As expected, the difference in product meaningfulness was not significant among participants who did not visualize following a ritual when drinking the tea (M = 2.96, SD = 1.87 vs. M = 3.13, SD = 2.03 for the lonely vs. control conditions, respectively; F(1, 588) = 2.31, p = .13). Furthermore, among participants who initially had been primed with loneliness, those in the ritual-frame-present condition reported greater product meaningfulness than those in the ritual-frame-absent condition (M = 3.34, SD = 1.97 vs. M = 2.96, SD = 1.87, respectively; F(1, 588) = 4.02, p < .05). Compared with their counterparts primed with loneliness, the difference in product meaningfulness due to the presence versus absence of the ritual frame was smaller among participants in the control condition, though also significant (M = 2.97, SD = 1.95 vs. M = 3.13, SD = 2.03, respectively; F(1, 588) = 4.94, p < .05).
Mediation analysis
To examine whether differences in product meaningfulness affected meaning in life and thus guided the interactive effect of the affect prime and ritual frame on loneliness, we followed Hayes’s (2013; Model 83) bootstrapping procedure using 5,000 resamples with replacement, including product meaningfulness as mediator 1, meaning in life as mediator 2, and involvement as covariate. Results showed that the index of moderated mediation was significant (95% bias-corrected CI: [−.1026, −.0068]). In particular, in the ritual-frame-present condition, product meaningfulness and meaning in life serially mediated the effect of the affect manipulation on loneliness (95% bias-corrected CI: [−.0675, −.0037]). However, this mediation effect was not obtained among those in the ritual-frame-absent condition (95% bias-corrected CI: [−.0040, .0472]).
Discussion
Study 4 replicated our previous results in a relatively virtuous product category and also demonstrated that our findings were independent of differences in involvement, mindfulness, or narrative transportation engendered by ritualistic versus nonritualistic behavior. In addition, we found support for our hypothesis that the effect of ritualistic behavior on loneliness is driven by meaning in life via perceived product meaningfulness. Our final study aimed to extend our findings in several ways. Given potential confounds associated with differences in simulating consumption in a ritual (vs. nonritual) way, in Study 5 all participants visualized completing the identical behavior, framed either as ritualistic or random steps. Instructing participants to visualize the same steps while changing their frames also enabled us to cast further doubt on an alternative explanation suggesting that the ritual-frame-present versus -absent manipulations of the previous studies simply provide a greater distraction from feeling lonely.
Next, we sought additional support for the role of meaning in life by relying on moderation of the effect (Spencer, Zanna, and Fong 2005). Specifically, if lonely consumers can add meaning to their life from an alternative source and thereby reduce their loneliness, ritualistic behavior should no longer affect their loneliness, because their lives no longer lack meaning. Thus, in addition to the affect prime and ritual frame manipulations, we also manipulated meaning to elucidate its role in the association between rituals and loneliness. In particular, reading a passage in a meaningful way can increase how meaningful one’s life is perceived to be (King et al. 2006); therefore, we manipulated meaningfulness following King et al. (2006) and instructed participants to read a text in a meaningful versus meaningless manner. We would find further support for the proposed role of greater meaning in life after engaging in a ritual if we replicated the affect prime × ritual frame interaction among participants in the meaninglessness prime condition, because here only the lonely participants would experience greater meaning in life following the ritual manipulation. However, the affect prime × ritual frame interaction should no longer be a significant predictor of loneliness among those primed incidentally with meaningfulness, because in this case both lonely and calm participants should experience greater meaning in life. In addition, directly manipulating (rather than measuring) meaningfulness enables us not only to ascertain the presence and absence of meaning in each of the ritual frame conditions but also to cast doubt on the alternative explanation suggesting that the underlying process is based on an unobserved construct confounded with meaningfulness, rather than meaningfulness itself.
We also wanted to illustrate managerial relevance of the effect of ritualistic behavior on downstream consequences by assessing purchase intentions of the product involved in the consumption ritual. Vohs et al. (2013) found that evaluations of products became more favorable, and willingness to pay for them increased, when they were consumed in a ritualistic (vs. random) manner because of greater involvement. Here, we expected that purchase intentions of the product would be greater when the product was included in an activity framed as a ritual, compared with the same activity without such a label, as long as the activity increased meaning in life (i.e., among those primed to feel lonely).
Study 5
Method
In exchange for compensation, 417 MTurk panelists (Mage = 32.31 years, SD = 10.48 years; 43.9% female) were assigned at random to one of eight conditions in a 2 (affect prime: lonely vs. calm) × 2 (meaningfulness prime: present vs. absent) × 2 (ritual frame: present vs. absent) between-subjects design. We first conducted two separate pretests to verify the effectiveness of our meaningfulness and ritual frame manipulations (see Web Appendix W7). In the main study, participants were first asked to complete the affect priming task as in previous studies, followed by the pretested meaningfulness manipulation task, and finally the pretested ritual frame manipulation task. Subsequently, they reported their loneliness as before (adapted from Russell [1996]; reverse-coded; α = .96), which again was our dependent variable. Participants also indicated their intention to buy another cake in the near future on a seven-point scale (1 = “not at all,” and 7 = “very much”). Finally, they reported demographic information and were debriefed.
Results
Loneliness
A 2 (affect prime) × 2 (meaningfulness) × 2 (ritual frame) ANOVA revealed a significant two-way interaction between affect prime and ritual frame on loneliness; F(1, 409) = 14.97, p < .001. Specifically, in the ritual-frame-absent conditions, participants who had been primed to feel lonely still felt lonelier than their counterparts initially primed with calmness (M = 3.29, SD = 1.22 vs. M = 2.62, SD = 1.11, respectively; F(1, 409) = 19.24, p < .001). The difference in loneliness was eliminated among those in the ritual-frame-present conditions (M = 2.77, SD = 1.03 vs. M = 2.88, SD = 1.12 for the lonely vs. calm prime, respectively; F(1, 409) = .53, p = .47).
Furthermore, a marginally significant three-way interaction emerged (F(1, 409) = 2.93, p = .088). As we show in Figure 4, Panel A, in the meaningfulness-absent condition, the interaction between affect prime and ritual frame was significant, replicating our previous results (F(1, 409) = 14.07, p < .001). In particular, participants initially primed with loneliness indicated that they felt significantly less lonely in the ritual-frame-present (M = 2.63, SD = 1.01) versus -absent (M = 3.69, SD = 1.16) condition (F(1, 409) = 21.20, p < .001). Conversely, ritual framing did not affect how lonely those who had been primed with calmness felt (M = 2.96, SD = 1.15 vs. M = 2.80, SD = 1.21 for the ritual-frame-present vs. -absent condition, respectively; F(1, 409) = .48, p = .49).
Study 5: Interactive effect of meaningfulness, affect, and ritual frame on loneliness.
However, in the meaningfulness-present condition, the affect prime × ritual frame interaction was no longer significant (F(1, 409) = 1.97, p = .16), as shown in Figure 4, Panel B. Presence versus absence of the ritual frame did not have an effect on loneliness among participants who had initially been primed with loneliness (M = 2.91, SD = 1.04 vs. M = 3.00, SD = 1.19 for the ritual-frame-present vs. -absent condition, respectively; F(1, 409) = .27, p = .61) or among those primed with calmness (M = 2.81, SD = 1.10 vs. M = 2.44, SD = .97; F(1, 409) = 2.28, p = .13).
Purchase intentions
A 2 (affect prime) × 2 (meaningfulness) × 2 (ritual frame) ANOVA revealed a significant two-way interaction between affect prime and ritual frame on intention to buy another cake (F(1, 409) = 9.10, p < .01). In the ritual-frame-absent conditions, those who initially had been primed to feel lonely expressed greater intentions to buy another cake than their counterparts primed with calmness (M = 5.30, SD = 1.58 vs. M = 4.96, SD = 1.68, respectively; F(1, 409) = 5.51, p < .05). The difference in intention to buy another cake, however, was eliminated in the ritual-frame-present condition (M = 4.94, SD = 1.60 vs. M = 5.45, SD = 1.35 for the lonely vs. calm participants; F(1, 409) = 2.57, p = .11).
More importantly, the three-way interaction among the factors on intention to buy another cake was significant (F(1, 409) = 4.89, p < .05). In the meaningfulness-absent conditions, the interaction between affect prime and ritual frame was significant (F(1, 409) = 12.82, p < .001). That is, participants initially primed with loneliness indicated that they would be more willing to purchase another cake in the ritual-frame-present (M = 5.57, SD = 1.58) versus -absent (M = 4.50, SD = 1.60) condition (F(1, 409) = 11.34, p = .001). Yet mentally simulating behavior framed as a ritual only marginally affected intentions to buy another cake among those who had been primed to feel calm (M = 4.88, SD = 1.69 vs. M = 5.41, SD = 1.31 for the ritual-frame-present vs. -absent condition; F(1, 409) = 2.87, p = .09). However, in the meaningfulness-present condition, the affect prime × ritual frame interaction was insignificant (F(1, 409) = .20, p = .66). Presence versus absence of the ritual frame had no effect on purchase intentions among participants who had initially been primed with loneliness (M = 5.04, SD = 1.55 vs. M = 5.25, SD = 1.53 for the ritual-frame-present vs. -absent condition; F(1, 409) = .51, p = .48) or among those primed with calmness (M = 5.04, SD = 1.68 vs. M = 5.49, SD = 1.41; F(1, 409) = 2.02, p = .16).
Discussion
Findings from the current study replicated and extended our previous results in important ways by providing additional support for the role of meaning in life using a moderation-of-process design (Spencer, Zanna, and Fong 2005). Specifically, we again found that lonely participants who simulated behavior framed as a ritual versus random steps subsequently reported reduced loneliness, even when all participants visualized engaging in the same activities, which differed only in whether they were framed as a ritual. Importantly, and consistent with our proposition that engaging in, or merely simulating engaging in, behavior framed as a ritual may render life more meaningful for lonely individuals, differences in loneliness due to the framing of the steps to consume the product were attenuated when an incidental task created the experience of meaningfulness. Finally, in the absence of the incidental meaningfulness task, lonely participants expressed greater purchase intentions of the product involved in ritual frame condition than those primed with calmness and those in the random steps frame condition.
General Discussion
Experiencing loneliness is becoming an increasingly common occurrence. Given that being lonely is often associated with a perception that one’s life is lacking in meaning (Stillman et al. 2009; Tam and Chan 2019; Wheeler, Reis, and Nezlek 1983), our research aimed to explore ritualistic behavior as a heretofore unexamined, yet effective, approach to adding meaning to consumers’ lives and thereby reducing loneliness. A series of studies supported our prediction regarding the impact of ritualistic behavior on loneliness and regarding enhanced meaning in life as the underlying mechanism. In addition, we examined and cast doubt on alternative explanations based on feelings of control, fun, involvement, effort, mindfulness, and narrative transportation.
By establishing a causal link between ritualistic behavior, meaning in life, and loneliness, our findings contribute insight into an important, real-world consumer issue—loneliness and the ways in which it can be reduced—that has yet to receive much attention from consumer researchers. Our findings complement those of Tian et al. (2018) regarding the beneficial impact of ritualistic behavior on consumer well-being; they found that rituals can nudge consumers toward healthier options by increasing self-control. While prior research has found that individuals may rely on social activities aimed at reestablishing social connections (Fuchs and Rehm 1977) in response to loneliness, we introduce rituals as activities that can be undertaken to reduce this type of negative affect.
In addition, we contribute to findings on ritualistic behaviors and their impact on consumption and on consumers’ affective states. Seminal work within the psychological orientation toward the study of rituals by Vohs et al. (2013) found that rituals can enhance the enjoyment of snacks and drinks via increased involvement. In addition, ritualistic behaviors can reduce anxiety and thereby enhance subsequent performance (Brooks et al. 2016). Our work extends these findings by showing that the benefits of ritualistic behaviors generalize from reducing anxiety to other affective states (i.e., loneliness) and to responses other than liking (i.e., purchase intent). Importantly, our results support our argument that even unfamiliar, minimal rituals can engender meaning. That is, even though such rituals may initially be alien, and thus not intrinsically meaningful, consumers may still appraise participating in even unfamiliar, minimal rituals as a meaningful practice.
Furthermore, we proffer meaning in life as a previously unexamined driver through which rituals exert their impact, given the distinct relationship between loneliness and perceptions that one’s life lacks meaning. Prior research on how rituals take effect has identified self-control (Tian et al. 2018) as well as feelings of control (Norton and Gino 2014), involvement (Vohs et al. 2013), and anxiety (Brooks et al. 2016) as underlying mechanisms, and we add to this literature by showing that rituals may make life more meaningful and reduce loneliness.
Our findings also have substantive implications. Given that the experience of loneliness is becoming more common, and results of Study 1 suggest that greater frequency of ritualistic behavior is associated with lower loneliness, it behooves marketers to encourage and communicate ritualistic behavior involving their brands—either suggesting the ritual themselves or urging consumers to develop their own. Furthermore, simply putting a behavior in a ritualistic frame appears to affect meaning, loneliness, and purchase intentions. The results of Study 1 also show that rituals that have meaning for consumers and often occur as part of a special event (e.g., holidays, important sporting events) and may include loved ones. Thus, to ensure that rituals involving their brands are meaningful to consumers, marketers should consider including these characteristics when suggesting ritualistic steps, and—given that we found that the most important quality that transforms common behaviors into rituals was the presence of rigid, formal, and repetitive steps—marketer-provided rituals should reflect these characteristics.
In addition to facilitating brand communities and enhancing consumer–brand relationships (Muñiz and O’Guinn 2001) or increasing attitudes and willingness to pay (Vohs et al. 2013), ritualistic behavior also has a beneficial impact on consumer welfare, reducing loneliness and its associated negative consequences. Of course, from a public policy perspective, reducing loneliness is a worthy end in and of itself, so governmental actors could stimulate rituals that do not include particular product options to add much-needed meaning to people’s lives.
Limitations in this research offer worthwhile opportunities for future directions. First, the current research and much of the literature on which it builds focus on the positive consequences of engaging in rituals. However, ritualistic behavior is unlikely to be uniformly beneficial. For instance, individuals suffering from obsessive-compulsive disorders engage in repetitive and rigid sequences of behavior that serve no functional purpose, which often characterize rituals. Moreover, we focused on rituals involving food options. Although food preparation and consumption are often associated with ritualistic behavior (Rook 1985), future studies might broaden the types of rituals to document the degree to which the meaning inherent in other types might influence loneliness.
Second, the ritualistic behaviors in our main studies were prescribed to the participants. Although that is consistent with the way many marketers communicate a consumption ritual or ritualistic steps in their advertising, such as Nabisco does for eating an Oreo, consumers clearly also generate their own idiosyncratic rituals. Indeed, some marketers encourage consumers to develop their own rituals involving their brand, as exemplified by Hershey’s “There’s no wrong way to eat a Reese’s” campaign. Thus, future research might include studies in which participants self-generate rituals. Furthermore, although our manipulations checks showed that our ritual manipulations—both to engage in and to visualize—worked as intended, it could be interesting to examine if there are cross-cultural differences in how consumers define rituals and what they perceive comprises rituals.
Third, participants in our main studies were instructed to engage in the ritualistic behavior. Although Study 1 found that consumers report greater meaning in life and lower loneliness as their frequency of engaging in consumption rituals increases, future research might examine under which circumstances lonely consumers actively seek out rituals. In addition, it would be noteworthy to examine if the number of steps involved in the ritual affect not only how meaningful it is perceived to be but also the degree to which loneliness is reduced. Finally, our studies provided evidence of the impact of rituals on loneliness when assessed immediately after the ritualistic behavior. It would therefore be worthwhile to examine for how long rituals enhance meaning in life and thus reduce loneliness.
Supplemental Material
Supplemental Material, sj-pdf-1-mrj-10.1177_0022243721993426 - Ritualistic Consumption Decreases Loneliness by Increasing Meaning
Supplemental Material, sj-pdf-1-mrj-10.1177_0022243721993426 for Ritualistic Consumption Decreases Loneliness by Increasing Meaning by Xuehua Wang, Yixia Sun and Thomas Kramer in Journal of Marketing Research
Footnotes
Acknowledgment
The authors would like to thank the review and editorial teams for the helpful guidance received.
Author Contributions
The authors contributed equally and are listed in reverse-alphabetical order.
Associate Editor
James Bettman
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (Grant Numbers: 71772108 and 71602172) and Zhi Jiang Young Scholar Foundation (Grant Number: 19ZJQN03YB).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
