Lay Beliefs about the Possibility of Finding Enduring Love: A Mediator of the Effect of Parental Relationship Quality on Own Romantic Relationship Quality

Abstract

This paper tested mechanisms that explain why and how relationship quality between parents might impact on later romantic relationship quality of adult children. In both the pilot (N = 119) and the main (N = 684) study, a model fit the data well whereby perceived “parental relationship quality” predicted “positive lay theories” about romantic love. This, in turn, predicted “secure attachment” to the romantic partner, which in turn predicted “relationship satisfaction” with the romantic relationship. Findings suggest that lay theories about finding enduring love are an important mediator of effects of parental relationships on later adult relationships.

Keywords:

• Lay beliefs
• romantic relationships
• romantic relationship satisfaction
• attachment
• love
• enduring love
• parental relationship quality

Understanding what affects romantic relationship satisfaction or dissatisfaction is an important endeavor, given how many romantic relationships end in separation or divorce (see, e.g., BBC, 2021). Of course, many factors have been linked to romantic relationship quality and success. Factors that have been discussed as impacting on romantic relationship quality are, for example, dispositional traits that individuals bring into the partnership, mental health problems, and interaction styles between the partners (see, e.g., Johnson et al., 2015). The quality of the parental relationship witnessed by offspring during childhood is also an important predictor of later relationship outcomes (Amato & Booth, 2001). Work on the intergenerational transmission of divorce clearly shows that relationship problems between parents can be perpetuated in the next generation (Kunz, 2001). What is not well understood, however, are the mechanisms how this intergenerational transmission occurs. This paper will zoom in on the dynamics in the family-of-origin which adult children were exposed to whilst growing up. It will test the mechanisms through which the quality of the relationship between one’s parents might impact on own romantic relationship quality of adult children later on in life. Two concepts in particular are tested as potential mediators: lay theories about the possibility of finding enduring love, and attachment to the romantic partner. It is proposed that lay theories about the possibility of finding enduring love have an important role to play, and that they in fact influence attachment to the romantic partner. While romantic attachments have been studied before (e.g., Reizer et al., 2014), lay theories about finding enduring love have received little attention, although they plausibly have important consequences for romantic relationships (Zagefka et al., 2021b). Both lay theories and attachments will therefore be tested as potential mediators, due to their potential theoretical and practical relevance.

Parental relationship quality impacts on own romantic relationship quality

Marital quality is often transmitted across generations (Yu & Adler-Baeder, 2007). For example, Amato and Booth (2001) found that offspring’s recollection of parents’ marital discord was negatively related to the offspring’s marital harmony, and positively to marital discord. Cui and Fincham (2010) too found that both parental divorce and marital conflict impact on young adults’ romantic relationships. Evidence that family-of-origin dynamics affect later romantic relationships also comes from studies focusing on infidelity (Weiser & Weigel, 2017), and on the more extreme end of family dysfunction (Maleck & Papp, 2015). It stands to reason the offspring’s perception of the general “tone” of the relationship between the parents (e.g., cold or warm, supportive or distant) might also impact on how adult children conduct their own close relationships later on in life. Whether this is the case, and the mechanisms linking perceptions of the relationship quality between one’s parents to later, own romantic relationship quality, is the focus of this investigation.

Romantic attachment as a mediator of the effects of parental relationship quality on own relationship quality

There are several possible mechanisms that might explain effects of the quality of the relationship between parents on the way adult children later conduct their own romantic relationships. According to Social Learning Theory (Bandura, 1977; Bandura et al., 1961), children are prone to repeating behavior that was role modeled to them. Hence, children who witnessed dysfunctional interactions and conflict between their parents are likely to replicate behavioral patterns that lead to conflict. The effect of past relationship problems on future relationship challenges in the next generation might also be contingent on other factors. For example, Lee (2018, 2019) found that the effect of parental relationship resolution on children’s later romantic relationships is contingent on parent–child relationships during adolescence, and indeed the child’s gender.

One variable that is frequently linked to both problems in the family-of-origin on the one hand and to romantic relationship outcomes on the other hand is attachment (Bowlby, 1979). Attachment scholars propose that children learn from a young age to either have secure or insecure attachment bonds with their primary caregivers. Insecure attachment bonds can be avoidant or anxious in nature. Those attachments to primary caregivers can then later on be reenacted in the context of other significant relationships, for example with the romantic partner (Nisenbaum & Lopez, 2015; Reizer et al., 2014; Zagefka et al., 2021a). However, despite frequent concordance in attachment styles toward significant others, there is also sometimes variation in the way the same person is attached to different significant others (La Guardia et al., 2000).

Attachment to the romantic partner has been found to be affected by the success or failure of the parental relationship (Sprecher et al., 1998), with problems in the parental relationship being linked to insecure rather than secure types of romantic attachment. Further, several studies have found that attachment to the romantic partner has important implications for romantic relationship functioning. One example is the study by Dewall et al. (2011), who found that romantic attachment styles are predictive of fidelity. What is clear then, is that romantic attachment is influenced, on the one hand, by the quality of the relationship between one’s parents, and that it influences, on the other hand, own romantic relationship quality and success. It can therefore be proposed that romantic attachment might mediate the effects of disruption and problems in the parental relationship on adult children’s subsequent own love relationships. This is in line with findings by Hayashi and Strickland (1998), who demonstrated that participants who reported growing up with protracted interparental conflict were more likely to report jealousy and fears of abandonment in their love relationships. This suggests that indeed exposure to low-quality parental relationships and parental relationship discord can negatively affect attachment styles enacted vis-à-vis romantic partners. Insecure romantic attachments then hamper romantic relationship quality and satisfaction.

Lay theories about the possibility of finding enduring love

Many scholars have proposed that lay theories about what is normal and desirable within romantic relationships can have a strong impact on conflict in, and satisfaction with, those relationships (e.g., Eidelson & Epstein, 1982; Fehr, 1999; Knee, 1998; Knee et al., 2001; Masarik et al., 2013; Simons et al., 2012; Sprecher & Metts, 1989; Sullivan & Schwebel, 1995). For example, people might differ in the extent to which they believe that relationships can grow and change (Canevello & Crocker, 2011; Cobb et al., 2013), that passion decay is inevitable in relationships (Carswell & Finkel, 2018), that soulmates exist (Franiuk et al., 2012), that disagreement between partners is an indication that there is something wrong in the relationship (Cramer, 2004; Zagefka & Bahul, 2021), and people might also differ in their marriage ideals (Lam et al., 2016).

One type of lay belief that has not received much attention, but that can be assumed to be very important, are beliefs about the possibility of finding enduring love (Zagefka et al., 2021b). Perceptions of the parents’ relationship quality can clearly be linked to expectations about intimacy and, more generally, beliefs about what is normal and achievable in romantic relationships (Einav, 2014; Mahl, 2001; Sprecher et al., 1998). Having witnessed a more or less successful relationship between one’s parents might well affect beliefs about whether or not the goal of finding a stable, happy relationship is attainable.

Because lay beliefs about romantic relationships can be expected to be impacted by what children witness between their parents, and because lay beliefs are likely to impact on conflict in and satisfaction with romantic relationships, just like attachment they are a potential mechanism linking family-of-origin dynamics to later romantic relationships. The present research will test this mechanism, and highlight positive lay theories about the possibility to find enduring love (or the lack of positive lay beliefs) as a potential mediator between family dynamics in the past on the one hand and present romantic relationships on the other.

Proposed order of effects: lay theories as antecedents of attachment

As seen above, both secure attachment to the romantic partner and positive lay theories about the possibility of finding enduring love and are affected by parental relationship success (Sprecher et al., 1998). Both constructs potentially mediate the effects of parental relationship quality on later romantic relationship quality of adult children. This raises the question of how these two potential mediators are related to each other. It could be that “lay theories” affect “attachment,” or that “attachment” affects “lay theories,” or that both operate as parallel, rather than sequential, mediators of the effect of parental relationship quality on own romantic relationship satisfaction.

It is proposed here that the most plausible causal sequence is that positive lay theories affect romantic attachment, which in turn then affects relationship satisfaction. At a most basic level, lay theories are generic beliefs about love that are thought to be generally true, and from these generalized lay beliefs should stem more specific beliefs about, and orientations toward, one’s concrete romantic partner. It seems more plausible that a general belief will inform orientation toward a specific individual (the romantic partner) than vice versa.

Moreover, the idea that lay beliefs should be an antecedent of romantic attachment can be proposed on the basis of the hypothesis on self-fulfilling prophecies. In numerous domains, for example academic attainment (Madon et al., 1997) and juries within the legal system (Hill et al., 2008), it has been demonstrated that humans are prone to paying particular attention to stimuli they are expecting. Confirmation bias and selective evidence gathering tends to influence conclusions and outcomes. The same can be assumed to be true in terms of expectations regarding one’s ability to find lasting love. If someone has not been role modeled enduring, happy relationships of high quality by their parents, that person is not likely to develop a belief that finding enduring love is attainable. Whether or not someone has positive expectations about finding enduring love will then affect how they behave within their romantic relationships, e.g., whether they behave in ways indicative of secure romantic attachment and whether they are satisfied with their romantic relationships.

Those with negative expectations will be more likely to look for—and therefore find—signals that disconfirm enduring, positive qualities within their romantic relationships. Expecting negativity or instability in romantic relationships will be connected to being less trusting and less open, and less securely attached to romantic partners. In contrast, those with positive expectations who anticipate that their relationships will be enduring and successful will be more willing to rely and depend on their partner, to be open with their partner, and to expect the partner to continue to be emotionally available to them. In other words, those with positive lay expectations about romantic love will be more likely to focus on signals that suggest positive qualities of the romantic relationship, and positive expectations will lead them to be more securely attached to their romantic partners. For these reasons, one can expect that positive lay theories, which result from what adult children observed between their parents, will affect attachment to later romantic partners, which in turn will then impact on satisfaction with the romantic relationship.

The a priori hypothesis was clearly that lay theories preceed romantic attachments in a two-step mediation process that channels the effects of parental relationship dynamics on own romantic relationship satisfaction. However, it is not entirely inconceivable that romantic attachments might instead preceed lay theories within the mediation chain, or that in fact both variables operate in parallel rather than sequential manner. Hence, as will be seen below, the hypothesized model was evaluated against those two theoretical alternatives.

Sum of predictions

In sum, then, it was proposed that a positive perception of the relationship between one’s parents (i.e., the perceived “parental relationship quality”) would be positively associated with “positive lay theories” about the possibility of finding enduring love, which would in turn be positively associated with “secure attachment” to the romantic partner, which would in turn be positively associated with “relationship satisfaction” with the romantic relationship. This process was tested in two online questionnaire studies: one smaller scale pilot study (N = 119) to gauge the strength of effects and conduct a power analysis, and one main study (N = 684) to test and confirm the patterns found in the pilot study.

Pilot study

Method

Participants

One hundred and nineteen people participated in this online study (mean age = 31.84 years; SD_age = 12.69; 32 males, 87 females, 0 identified as neither male nor female). This sample was a sub-sample of a larger study that was conducted in two parts: part one included items that all participants responded to irrespective of their relationship status, and part 2 was only displayed to those participants who reported currently being in a romantic relationship. Because this present paper is aiming to predict romantic relationship satisfaction, only those latter participants were included here. This was a convenience sample recruited via social media by distributing the survey link and encouraging both participation and also further distribution, to encourage a snowballing effect.

Measures

The study included the following measures (all on scales from 1 = “not at all/strongly disagree” to 5 = “very much/strongly agree”): To measure “parental relationship quality,” participants were asked to think about when they were growing up, and to answer the question “how much do you think your mum and dad were acting in these ways toward each other?” Items were “warm,” “loving,” “supportive,” “cold,” “angry,” “distant” (with the last three items reversed); α = .87.

Of course, not all people grow up in traditional families with a mother and a father (the parents might be divorced, there might be two parents of the same sex, children might be adopted, etc.). Several measures were included to screen for the potential impact of this. First, there was an open ended response question where participants were invited to mention anything else relevant about their family setup (divorce and death of a parent were given as examples of things that might be relevant). Second, participants were explicitly asked whether their parents were divorced (and if so what age the child was at the time), whether there had been a new live-in partner, and several questions related to these issues. Third, participants were asked to state how much contact they had with their mum and dad growing up (to potentially differentiate between parents who move out and lose all contact and those who leave the family home but retain good relations with the offspring). These screening efforts clearly revealed variations on family constellations during childhood. However, including or excluding participants on the basis of several criteria (e.g., divorce) did not have any substantial effects on the hypothesized process of interest here, and therefore results are based on the full sample. Indeed, the questions are applicable to different situations (e.g., it is possible to report on the parental relationship as being warm or not irrespective of whether parents are divorced or not).

“Positive lay beliefs” about romantic relationships was measured with 10 items: “I think I myself can have a stable, loving romantic relationship”; “I think it is possible for two people to be happily together in the long run”; “I think I can find a partner whom I can rely on permanently”; “I expect that I will have a great romantic relationship in the future”; “I am hopeful to find lasting love”; “I don’t really believe long-term romantic relationships can work” (reversed); “I am skeptical about whether it is possible to be happily together with one person for the long run” (reversed); “I don’t expect to have a successful long-term relationship myself” (reversed); “I have a clear idea of what skills are needed in order to have a successful long-term relationship”; and “I think I know what is important in order to make a long-term relationship work”; α = .88.

To measure “secure attachment” to the romantic partner, four items were adapted from Fraley et al. (2000). Participants were asked to state their agreement with the following items, thinking about their romantic partner: “I will turn to my partner in times of need”; “I usually discuss my problems and concerns with my partner”; “I talk things over with my partner”; “I find it easy to depend on my partner”; α = .92.

“Romantic relationship satisfaction” was measured with three items loosely based on Solomon and Jackson (2014). The items were “I am satisfied with my romantic relationship”; “I feel loved and cared about by my partner,” “I feel a lot of closeness and intimacy with my partner”; α = .89.

For both the pilot and the main studies, the general part of the questionnaire which was also answered by those who were not currently romantically involved included a number of other scales that are not relevant to the present hypotheses and that were not analyzed here, such as question on general psychological well-being. All aspects of the research were in line with American Psychological Association (APA) and British Psychological Society (BPS) ethics guidelines. Participation was voluntary and no monetary compensation was offered, although to encourage participation participants were offered an incentive of being entered in a raffle to win some online vouchers. The data for the pilot and main studies can be found here: https://osf.io/6gyvw/?view_only=d1df7dd6986a451887e8dc95f4bce5e7.

Results and discussion

Descriptives and bivariate correlations for all measures are displayed in Table 1. To test the model whereby “parental relationship quality” would affect “positive lay beliefs” about romantic relationships, which in turn would affect “secure attachment” to the romantic partner, which in turn would affect “romantic relationship satisfaction”, a structural equation model was built using AMOS. A direct path from “lay theories” to “relationship satisfaction” was also included, given that those two concepts—as seen above—can also be expected to be directly linked. Standardized path values are displayed in Figure 1. All paths and covariances between variables and errors terms that were included in the model are depicted in the figure, and if variables are not linked by a path this indicates that their direct relationship was assumed to be zero.

Figure 1. Structural equation model of positive lay beliefs about romantic relationships mediating effects of parental relationship quality on own romantic relationships. Pilot study (N = 119) standardized path coefficients in parentheses, main study (N = 684) coefficients free-standing.

Table 1. Bivariate correlations and means, for both the pilot study (N = 119) and the main study (N = 684).

Variable	Pilot study M	Pilot study SD	Main study M	Main study SD
Parental relationship quality	3.46	1.19	3.44	1.23	–	.30 ***	.22 *	.25 **
Positive lay beliefs about romantic relationships	4.24	0.68	4.37	0.62	.27 ***	–	.40 ***	.72 ***
Secure attachment to romantic partner	4.32	0.82	4.27	0.68	.12 ***	.51 ***	–	.62 ***
Romantic relationship satisfaction	4.29	0.83	4.43	0.73	.05	.49 ***	.63 ***	–

Note. * p < .05, ** p < .01, *** p < .001. M = mean, SD = standard deviation. Pilot study correlations above the diagonal, main study correlations below the diagonal.

The model fitted the data well, χ²(2) = 0.89, ns, CFI = .99, RMSEA = .001. All hypothesized paths were significant and in the expected direction. Bias corrected bootstrapping (2000 samples) was performed to estimate the indirect effects. There are two potential indirect effects in the model illustrated in Figure 1. The indirect effect of parental relationship quality on attachment to the romantic partner via lay beliefs about romantic relationships was significant, b = .15, p < .001, CI [.06; .27], sd. error = .06. The indirect effect of parental relationship quality on romantic relationship satisfaction, via both lay beliefs about romantic relationships and attachment, was also significant, b = .21, p < .001, CI [.11; .35], sd. error = .07.

As outlined above, when testing a process involving two mediators, there are three potential mechanisms that could be at play. The “black box” linking the predictor and the ultimate outcome variable might contain: 1) mediator A affecting mediator B, 2) mediator B affecting mediator A, or 3) both mediators operating in parallel, rather than in sequential order. To obtain stronger evidence for the hypothesized causal order of effects, the two alternatives to the predicted model were also tested. The goal was to demonstrate that these models would fit less well than the preferred model described above.

First, a model was calculated where the order of secure attachment and lay theories were swapped compared to the previous model, so that secure attachment now featured as a predictor of lay theories. To allow for comparability with the hypothesized model, again a direct path from attachment to relationship satisfaction was also included. As expected, the alternative model did not have acceptable fit with the data, χ²(2) = 6.36, p < .05, CFI = .97, RMSEA = .14. Second, a model was calculated where secure attachment and lay theories both were parallel mediators of the effect of parental relationship quality on relationship satisfaction, rather than them being ordered sequentially. As expected, this alternative model also did not have acceptable fit, χ²(2) = 28.82, p < .001, CFI = .82, RMSEA = .34.

Overall, then, there was evidence in this data that parental relationship quality observed whilst growing up impacts on lay beliefs that are formed about romantic relationships about whether or not finding enduring love is possible. These lay beliefs then affect how adult children relate to their romantic partner, and specifically whether they are securely attached to that partner. This, in turn, impacts on relationship satisfaction, with secure attachment being positively associated with relationship satisfaction.

Main study

One of the main aims of the pilot study was to determine the approximate size of the effects relevant to the hypotheses, to ensure sufficient statistical power for the main study. Based on the results of the pilot study, the smallest path coefficient was .30 (c.f. Figure 1). An a priori power analysis was conducted with G*Power, requesting a test for size of slope with a Slope H1 = 0.30, an alpha level of .05, and a power of .95. This recommended a total sample size of N = 111. Hence, this was the minimum acceptable sample size for the main study. As detailed below, the data collection method for the main study was identical to the one for the pilot study: the survey remained life online and available to potential participants for a period of approximately 4 weeks, and anyone willing to participate during this time was included in the final sample. The sample for the main study was independent of the pilot sample, i.e., there was no overlap in participants.

Method

Participants

Six hundred and eighty four people participated in this online study during the four-week period (mean age = 27.80 years; SD_age = 12.36; 169 males, 496 females, and 19 chose the category “other” to specify their gender or had missing data on the gender variable). Again, the sample was a sub-sample of a larger study that also had a section for all participants, including those not currently romantically involved. Again, this was a convenience sample accessed via social media by distributing the survey link and encouraging participation and further distribution.

Measures

The study included the same measures as the pilot study (all on scales from 1 = “not at all/strongly disagree” to 5 = “very much/strongly agree”), with one exception: a revised scale was used to measure romantic attachment, to ensure items would tap also into the absence of both avoidance and anxious attachment orientations. Again, attachment items were based on Fraley et al. (2000), but geared toward the current romantic partner. There were 9 items. Four items were the same as the ones used in the pilot study: “I will turn to my partner in times of need”; “I usually discuss my problems and concerns with my partner”; “I talk things over with my partner”; and “I find it easy to depend on my partner.” Five additional items, all of them reverse scored, were also used: “I don’t feel comfortable opening up to my partner,” “I prefer not to show my partner how I feel deep down”; “I often worry that my partner does not really care for me”; “I’m afraid that my partner may abandon me”; “I worry that my partner won’t care about me as much as I care about her/him.”

The alphas were: α = .94 for “parental relationship quality”, α = .90 for “positive lay beliefs” about romantic relationships, α = .83 for “secure attachment” to the romantic partner, and α = .86 for “romantic relationship satisfaction.”

Results and discussion

Descriptives and bivariate correlations for all measures are displayed in Table 1. To test the model whereby “parental relationship quality” would affect “positive lay beliefs” about romantic relationships, which in turn would affect “secure attachment” to the romantic partner, which in turn would affect “romantic relationship satisfaction,” a structural equation model was built using AMOS. A direct path from “lay theories” to “relationship satisfaction” was again also included. Standardized path values are displayed in Figure 1. All paths and covariances that were included in the model are depicted in the figure, and if variables are not linked by a path this indicates that their direct relationship was assumed to be zero.

The model fitted the data well. Although the chi square was significant which is unsurprising given the large sample, χ²(2) = 7.97, p < .02, the fit indices suggested good fit, CFI = .99, RMSEA = .07. All the paths were significant and in the expected direction. Bias corrected bootstrapping (2000 samples) was performed to estimate the indirect effects. The indirect effect of parental relationship quality on attachment to the romantic partner via lay beliefs about romantic relationships was significant, b = .14, p < .001, CI [.11; .18], sd. error = .02. The indirect effect of parental relationship quality on romantic relationship satisfaction, via both lay beliefs about romantic relationships and attachment, was also significant, b = .13, p < .001, CI [.10; .17], sd. error = .02.

Two alternative models were again estimated, to get further support for the prediction that lay theories causally precede secure attachment to the romantic partner rather than vice versa, and rather than them operating in parallel. First, a model was calculated where the order of secure attachment and lay theories were swapped compared to the previous model, so that secure attachment now featured as a predictor of lay theories (again, also including a direct path from attachment to satisfaction). As expected, this alternative model did not have acceptable fit with the data, χ²(2) = 50.86, p < .001, CFI = .93, RMSEA = .19. Second, a model was calculated where secure attachment and lay theories both were parallel mediators of the effect of parental relationship quality on relationship satisfaction, rather than them being ordered sequentially. As expected, this alternative model also did not have acceptable fit, χ²(2) = 206.30, p < .001, CFI = .69, RMSEA = .39.

Overall, then, again there was strong evidence in this data that parental relationship quality observed whilst growing up impacts on lay beliefs that are formed about romantic relationships about whether or not finding enduring love is possible. These lay beliefs then affect how adult children relate to their romantic partner, specifically whether they are securely attached to that partner. This, in turn, impacts on relationship satisfaction, with secure attachment being positively associated with relationship satisfaction.

General discussion

Across both the pilot and the main study, there was robust evidence that positive lay theories about the ability to find enduring love can explain why and how relationship patterns are transmitted across generations from parents to their children. Perceived parental relationship quality was directly associated with lay beliefs that were formed about the possibility of finding enduring love. These lay beliefs then acted as a self-fulfilling hypothesis: those who expected positive, enduring love to be a real possibility were more securely attached to their romantic partners. Secure attachment, in turn, was positively associated with relationship satisfaction.

These findings go beyond the existing literature in some important ways. First, previous work has attended to the effects of several lay theories, beliefs, and expectations people might have about romantic relationships. These have included the belief that disagreement is destructive, that mindreading is expected, that one has to be sexually perfect, that sexes are different, and that partners cannot change (Eidelson & Epstein, 1982). They have also included a belief in destiny or growth (Knee, 1998), or a believe in love at first sight (Sprecher & Metts, 1989), and a belief in soulmates (Franiuk et al., 2012). However, none of these previous efforts have focused on a belief that finding enduring love is an attainable goal. This is surprising because in some ways, this belief is more fundamental than some of those beliefs that have previously been studied. Whether enduring love is possible at all is a more fundamental question than questions about what this love should look like in detail, e.g., whether disagreement within relationships is problematic or not. The present work therefore puts a very important lay belief on the map, and it is hoped that this belief will attract future research attention.

Second, previous work has highlighted that attachment orientations might be an important mediator of transgenerational effects, and a few scholars have also suggested a potential mediating role for lay beliefs, but no work to date has considered these two factors jointly. The present findings suggest that lay beliefs might be an antecedent of romantic attachment, rather than vice versa. This is not only interesting in theoretical terms, but it also has important practical implications.

The findings suggest that in order to improve relationship satisfaction, it might be useful to address negative lay beliefs about the possibility of finding enduring love, by replacing negative parental role models with alternative models for couples to emulate. This is an important aspect that could potentially be considered within couples therapy.

Of course, the present research has several limitations which should be acknowledged. Most importantly, although SEM and the testing of alternative models can provide some preliminary insights into the likely causal direction of effects, ultimately the data used here is correlational in nature. In order to make strong causal claims, experimental data would be needed (however, implementing such a design would raise some obvious ethical questions), or longitudinal data would be needed (and running such a design would be time and resource intensive and might not easily be feasible for those reasons).

Relatedly, correlational designs cannot rule out the possibility of confounding factors which might impact results. For example, witnessing unfulfilling relationships of others (not only the parents) within one’s social network could also promote skepticism, as could the participant’s own romantic relationship history. Ideally, such factors would need to be measured in order to chart their exact effect, or control for it.

Another limitation is that this study relies on self-reports of remembered marital discord amongst one’s parents, and of course memory is subject to distortion and biases (Herrmann et al., 2005). Although one can assume a certain degree of correspondence between the actual levels of marital discord and those remembered at a later date by the offspring, especially when it comes to events on the more extreme end of the negativity continuum (Baugerud et al., 2014), again expensive prospective designs would be needed to rule out potential distortion due to memory effects. Moreover, the data used here were from convenience samples, and to achieve better confidence about the replicability of the effects data from larger scale, representative samples would help. Last but not least, we had to use a rather pragmatic approach for measuring the constructs of interest, because participation was voluntary, and this restricted the length of the questionnaire that was feasible. This limitation could be addressed by future studies that could employ longer, validated scales to tap into the constructs of interest.

There are several exciting avenues for future research. It would be interesting to study the lay belief introduced here, about the feasibility of finding enduring love, in conjunction with other lay beliefs about romantic relationships that have previously been studied. It would also be worthwhile to study the cross-cultural invariability or variability of the processes found here. Moreover, one could study whether the processes are invariant across different gender identities. Last but not least, it would be intriguing to study the processes highlighted here from a dyadic perspective, within a dyadic design. For example, at present it is not known whether one partner’s pessimism about the prospect of enduring love could be compensated for by the other partner’s optimism. These are all issues worthy of further research attention. For sure, future research would benefit from triangulation of methodological approaches. Qualitative data could generate much richer data and insights into the meaning and lived experience concerning the processes described here, as well as insights into the role of the potentially confounding variables mentioned above.

In sum, lay beliefs about the possibility of finding enduring love play an important role in shaping relationship satisfaction, and they are a key mechanism through which relationship patterns get transmitted across generations. It is therefore hoped that future research will further explore the nuances of this exciting new concept.

Clinical implications and applications

There are a number of implications of the present findings in the form of concrete applied suggestions that might be useful for everyday clinical practice in the context of couple therapy. First, the data suggest that contentment within romantic relationships is contingent on lay beliefs, and fundamentally on whether partners believe that enduring love is feasible. Skepticism about the possibility of finding enduring love will turn into a self-fulfilling prophecy, and will significantly reduce relationship satisfaction, even if objectively that relationship is functioning well. Couple therapists should consider this, and work with their clients to identify and—if necessary—redress negative beliefs about love. Without such reflective work, the impact of lay beliefs on relationship success might remain unconscious, and lay beliefs will continue to manipulate and jeopardize relationship success unnoticed and unchecked.

Second, the findings suggest that lay beliefs themselves are fundamentally affected by positive or negative role models. Those who did not have loving, enduring relationships modeled to them by their parents are in danger of being condemned to having less successful romantic relationships themselves. Clearly, clinical interventions are not able to turn back the clock and improve parental role modeling. However, a promising avenue might be to provide couples with alternative, new and better role models. There are two ways in which this could be achieved. First, couple therapists could do justice to the importance of role models by working with their clients in sessions to help them identify positive role models that the client can relate to and emulate. Second, role modeling can happen within group sessions, and this could be one reason why couple therapy benefits from group sessions shared between different couples: different clients can observe each others’ couple dynamics, and learn from each other by finding behaviors worth emulating (e.g., https://harvilleandhelen.com/couples-workshops/; https://www.kalanitbenari.com/workshop-for-couples).

Therapists clearly have an intuitive understanding of the importance of role models: many couple therapists publicize their role modeling credentials by publicizing their own happy relationship status, and indeed many clients might be reluctant to accept guidance from a therapist who is not seen as a good role model. This suggests that there is an intuitive understanding of the importance of role models, however despite this the issue is not typically attended to in couples therapy. It is time, then, that the importance of both potentially sabbotaging lay beliefs about love and of positive role models is given more attention by couple therapists.

Declarations

• All participants gave informed consent, in line with APA and BPS guidelines.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

There is no funding source to be acknowledged.

Data availability statement

The data for the pilot and main studies can be found here: https://osf.io/6gyvw/?view_only=d1df7dd6986a451887e8dc95f4bce5e7.