Lay Beliefs About Romantic Relationships: A Mediator of the Effect of Family Dysfunction on Romantic Relationship Satisfaction

This paper tested why people differ with regard to whether they believe it is possible to find enduring love. Variations were assumed to be due to differences in people’s experiences. Those who experienced dysfunction in their family of origin and who did not have positive relationships role-modelled to them were expected to be less likely to have positive lay beliefs about romantic relationships. Positive lay beliefs, in turn, were hypothesised to impact on dysfunction in own romantic relationships later on in life, which were in turn expected to affect relationship satisfaction. In other words, positive lay beliefs were tested as one potential mechanism through which family dysfunction whilst growing up impacts on relationship dysfunction in later adult life. This paper presents a pilot study (N = 176) which introduces a measure for ‘positive lay beliefs about romantic relationships’, and finds this measure to be associated, as expected, with dysfunction in the family of origin. The main study (N = 435) then tested the full hypothesised model (family-of-origin dysfunction → positive lay beliefs → romantic relationship dysfunction → relationship satisfaction) with structural equation modelling, and found that the model fitted the data very well, confirming the hypotheses. It was concluded that lay beliefs about whether or not it is possible to find enduring love are an important mediator of the effects of family-of-origin dysfunction on later romantic relationship satisfaction.


Introduction
Many people strive to find lasting love; yet this goal often remains elusive. For years, divorce statistics have hovered around the 40% mark in the UK (ONS, 2020). Multiple factors have been linked to marital and romantic relationship satisfaction or dissatisfaction, for example, certain features of couple interactions (Gottman, 1994), attachment styles (DeWall et al., 2011;Reizer et al., 2014), caregiving (Feeney & Collins, 2003), the partners' empathic capacity (Cramer & Jowett, 2010), physical intimacy (Leavitt & Willoughby, 2015), personality traits (Solomon & Jackson, 2014), as well as commitment, love, inclusion of other in the self, and dependence (Le et al., 2010). This contribution investigated whether and how experiences that people have whilst growing up, in their family of origin, impact on the way in which they conduct their own romantic relationships as adults. In particular, it was tested, in a two-step sequential mediation model, whether family dysfunction in the family of origin affects lay beliefs about romantic love, which then affects dysfunction within adult romantic relationships, which then affects relationship satisfaction.

How Past Dysfunction Affects Present Dysfunction
There is, indeed, good evidence that experiences growing up provide important templates that people are prone to reenact in their later life. According to Social Learning Theory, children imitate significant adult role models they are exposed to, and replicate their behaviour (Bandura, 1977;Bandura et al., 1961). This general mechanism has also been demonstrated in the context of how people conduct themselves in their intimate relationships. For example, Busby et al. (2005) found that family-of-origin experiences predict later adult romantic functioning. Hayashi and Strickland (1998) showed 1 3 that witnessing protracted interparental conflict growing up impacts on later own romantic relationships, with those who experience strong conflict being more jealous and fearful of abandonment in their love relationships. Moreover, in a small cross-sectional study, Einav (2014) demonstrated that perceptions of the parents' relationship are associated with expectations about own intimate relationships. In a further study, Yu and Adler-Baeder (2007) also confirmed that the quality of the parental relationship, but also the quality of new post-parental-breakup relationships that are formed by the parents, affect the relationship behaviour of adult children themselves. This pattern also emerged in a longitudinal study conducted by Amato and Booth (2001). Last but not least, Weiser and Weigel's (2017) findings also demonstrate that people tend to repeat patterns they witnessed in their family growing up-in this contribution, infidelity was more likely if parental infidelity had been witnessed. Overall, then, it is fair to conclude that family-of-origin patterns are often replicated in the next generation. This means that dysfunction in the family of origin likely leads to dysfunction in romantic relationships later on.
The root causes for family dysfunction are varied, and include intractable conflicts or rivalries between members, domestic violence, mental illness, alcohol or drug abuse, non-consensual extramarital affairs, or gambling. In dysfunctional families, the ways in which members relate to each other emotionally and practically is ineffective or damaging to members. Epstein et al. (1983) propose that families can be assessed along a continuum ranging from function to dysfunction. They distinguish between six subtypes/subscales of dysfunction, which are called problem solving, communication, roles, affective responsiveness, affective involvement, behaviour control, plus a measure of general functioning which assesses the overall health or pathology of the family. Typically, different subscales of dysfunction correlate very highly with each other, and they are summarised into an overall index of dysfunction.
Although the Family Assessment Device (FAD) has previously been used only in relation to families, the issues it taps into are clearly also relevant in couples, and can easily be measured for couples also (e.g. the question 'We work well together' can be answered with participants having the whole family in mind, or their partner). In this work, we tested whether dysfunction in the family of origin predicts romantic relationship dysfunction.

Lay Beliefs About the Possibility of Finding Enduring Love as Linking Past Dysfunction to Relationship Outcomes
A number of mechanisms have been proposed that might explain why family-of-origin dysfunction affects later romantic relationships. Simons et al. (2012) proposed that past experiences can give rise to cynical, distrusting schemata involving insecure attachment or hostile attribution bias. These schemata might then impact on relationship expectations and behaviours later on, and they might be one mechanism through which dysfunction in the past might be perpetuated and lead to future dysfunction. Mental health problems and negative couple interaction have also been demonstrated to play a role (Johnson et al., 2015;Maleck & Papp, 2015), as have conflict behaviours within the couple (Cui & Fincham, 2010). The present contribution aimed to add to the existing literature, by highlighting the important role that positive lay beliefs about the possibility of finding enduring love might play in explaining the link between family-of-origin dysfunction and dysfunction within later romantic relationships of the offspring of such dysfunctional families.
It has been demonstrated that lay beliefs or lay theories are important because they affect information processing, i.e. they affect the lens through which people perceive the world, and they guide behaviour (Fehr, 1999). A number of lay beliefs about romantic relationships have previously been studied. These romance-relevant beliefs include ideas such as disagreement in relationships is destructive, mindreading is expected, the couple should be sexually perfect, sexes are different, partners cannot change (Eidelson & Epstein, 1982), partners are either meant for each other or not (Knee, 1998), and love finds a way (Sprecher & Metts, 1989; for further lay beliefs, see Carswell & Finkel, 2018;Cramer, 2004;Crocker et al., 2017;Franiuk et al., 2012).
Although many lay beliefs about romantic relationships have been studied, no previous work has specifically focussed on beliefs that finding enduring love is an achievable goal. But, when it comes to parental relationship dysfunction and relationship dissolution, this is the belief that can be assumed to be most strongly and most directly impacted. In a nutshell, if children experience relationship failure, they will come to expect relationship failure. Masarik et al. (2013) agree that people's lay theories about marriage and other romantic relationships are rooted in their experiences when growing up.
In fact, there is reason to believe that lay beliefs about romance will be more strongly impacted by formative experiences growing up than by current romantic experiences. This can be expected on the basis of ideas of many theorists who propose that fundamental schemata that are built in childhood have a long reach into adulthood (e.g. Bowlby, 1979;Freud, 1991). Because of the fundamental, formative effects of early experiences, one would expect lay theories to be driven more strongly by family-of-origin factors than by current relationship experiences. Therefore, it was hypothesised that family-of-origin dysfunction would reduce lay beliefs about the possibility of finding enduring love.
Further, there is reason to believe that lay beliefs about the possibility of finding enduring love have an impact on romantic relationship dysfunction. Supporting evidence for this prediction comes from the literature on self-fulfilling prophecies and, relatedly, the work on growth versus entity theories (Dweck, 2006). Put simply, if someone does not believe they can succeed at a task, they are likely to not even try, which will then lead to failure (Claro et al., 2016;Dweck, 2006). It stands to reason that the same process should be at play when it comes to the successful conduct of intimate, enduring relationships. If someone does not believe that finding enduring love is possible, this person will be less likely to aim for this goal, and less likely to achieve it. One would therefore expect a lack of positive beliefs about the possibility of finding enduring love to be associated with more romantic dysfunction. This is because if someone believes the relationship is doomed anyway, they will be less motivated to interact with the partner in sensitive and careful ways. Indeed, growth versus entity beliefs have previously been applied to romantic relationships, and have been found to be predictive of romantic outcomes (Cobb et al., 2013;Knee et al., 2001, see also Canevello & Crocker, 2011). This is why we expected that lay beliefs that finding enduring love is possible (or not) have a strong effect on romantic relationship dysfunction and, in a second step, relationship satisfaction.

Relationship Dysfunction and Relationship Satisfaction
As outlined previously, what was predicted was an indirect effect of family-of-origin dysfunction on romantic relationship dysfunction, mediated by lay beliefs about romantic relationships. In addition, it was also expected that romantic relationship dysfunction would decrease relationship satisfaction. This is to be expected on the basis of several previous studies. It is in line with, for example, Sullivan and Schwebel (1995), who found, in a sample of US undergraduates, that irrational relationship beliefs negatively affect relationship satisfaction. This was also found in a study by Zagefka et al. (2020) which charted the effects of dysfunctional relationship beliefs on relationship satisfaction. The dimensions included in Epstein et al.s' (1983) Family Assessment Device have also been linked to relationship satisfaction. For example, McNulty and Russell (2010) demonstrated that problem-solving behaviours are linked to relationship satisfaction; Eğeci and Gençöz (2006) showed that communication skills impact relationship satisfaction; and Kimmes et al. (2014) demonstrated that empathy, a concept related to affective responsiveness and affective involvement, is related to relationship satisfaction. In sum, just as family-of-origin dysfunction is linked to dissatisfaction within families, so too would we expect romantic relationship dysfunction to be linked to romantic relationship dissatisfaction.
Taken together, it was hypothesised that positive lay beliefs about the possibility of finding enduring love is affected by dysfunction in the family of origin, and that it affects, in turn, dysfunction in own romantic relationships and through this satisfaction with those relationships. In other words, we tested whether positive lay beliefs about enduring love would mediate and explain the effects of past experiences on future romantic relationships.

Effects of Divorce
Children of divorced parents struggle comparatively more to find lasting love. There is mounting evidence for the intergenerational transmission of divorce in different cultural settings (Amato, 1996;Diekmann & Schmidheiny, 2013, although see Li & Wu, 2008). Scholars suggest that children of divorce might often be confused about what healthy, enduring relationships look like, and they reenact unhealthy relationship patterns modelled by their parents (Piorkowski, 2008). Cui and Fincham (2010) also found that parental divorce was linked to low romantic relationship quality among young adults. Others agree that witnessing parental divorce can influence own romantic relationships later in life (e.g. Lee, 2018Lee, , 2019. However, not all studies have found that parental divorce leads to the development of beliefs and patterns that hinder the development of own romantic ties. Some studies (e.g. Cui et al., 2011;Mahl, 2001;Sprecher et al., 1998) found no evidence for this, or they found a more nuanced picture. One potential explanation for these divergent findings is, of course, that positive parental role modelling cannot easily be proxied by a simple 'divorce versus no divorce' dichotomy. Levels of parental conflict in non-divorcing families might be high, and non-divorce might hence sometimes be more damaging for the offspring of the ailing union than divorce (Maleck & Papp, 2015). This is why, in this present study, we focussed on family dysfunction in the family of origin, rather than simply looking at the effects of divorce. Nonetheless, because conflict and dysfunction can often be assumed to precede divorce, and because an experience of parental relationship breakdown can be assumed to call into question beliefs that enduring romantic love is achievable, it was also expected that an experience of parental divorce would lead to more negative lay beliefs about the possibility of finding enduring love.

Hypotheses and Overview of Current Studies
To briefly sum up the predictions, it was hypothesised that family-of-origin dysfunction would reduce the extent to which people believe that finding enduring love is possible. In other words, we expected a negative effect of dysfunction whilst growing up on positive lay beliefs. Positive lay beliefs, in turn, were expected to impact on dysfunction in the present romantic relationship. Finally, romantic relationship dysfunction was expected to negatively impact on relationship satisfaction (for a summary of the processes, see Fig. 1). Importantly, as outlined previously, we expected that lay beliefs are more strongly impacted by formative experiences in the family of origin than by current romantic experiences. Finally, we expected an effect of experience of parental divorce, so that those whose parents had divorced would have less positive lay beliefs about the possibility of finding enduring love.
Two studies were conducted to test these ideas. A pilot study introduced a measure of 'positive lay beliefs', and ascertained that this concept would be associated with the key predictors of dysfunction in the family of origin and perceived parental relationship quality in the predicted way. A main study then followed up these initial findings, and tested a full model of the hypothesised processes, as well as the effects of parental divorce.

Participants
The sample consisted of one hundred and seventy six people. The mean age was 18.65 years (SD age = 2.01; 142 females, 33 males, 1 did not declare their gender). One hundred nineteen reported that their parents were not divorced, and 54 reported that the parents were divorced (3 missing values). Participants were recruited at a university 'taster' day, and further participants were recruited in public spaces (e.g. the library) by research assistants. The 'taster' day was targeted at secondary school students in their penultimate or final year of schooling; hence those participants were still living with their parents. Because we wanted participants to respond in relation to dysfunction in their family of origin (not a family they themselves might have founded), only participants aged 22 or younger, i.e. those who were still dependants living with their parents, were included in the final sample (meaning that some teachers who were present and who also completed the questionnaire were excluded from the analyses). This cross-sectional correlational study included the following measures:

Measures
Family-of-origin dysfunction was measured with a shortened 31 item version of the Family Assessment Device (Epstein et al., 1983) which taps into the family's ability to constructively solve problems, communicate efficiently, have clear rules, and other aspects of functioning. Only 24 items were included in the final scale because that significantly improved the scale reliability. Participants were asked to think of everyone living in their household, and to indicate to which extent the following statements are true for them (1 = "not at all" to 7 = "very much"). Items to measure 'problem solving' were: 'We usually act on our decisions regarding problems', 'We resolve most emotional upsets that come up', 'We confront problems involving feelings'; items to measure 'communication' were: 'We are frank with each other', 'We are good at talking to each other'; items to measure 'roles' were: 'We make sure members meet their family responsibilities', 'We discuss who has to do household jobs'; items to measure 'affective responsiveness' were: 'We are reluctant to show our affection for each other', 'Some of us just don't respond emotionally', 'We do not show our love for each other'; items to measure 'affective involvement' were: 'You only get the interest of others when something is important to them', 'We are too selfcentred', 'Our family shows interest in each other only when they can get something out of it', 'We intrude too much into each other's lives'; items to measure 'behaviour control' were: 'You can easily get away with breaking the rules', 'Anything goes in our family'; and items to measure general functioning were: 'We misunderstand each other', 'In times of crisis we can turn to each other for support', 'We can't talk to each other about the sadness we feel', 'We can express feelings to each other', 'There are lots of bad feelings in the family', 'We feel accepted for what we are', 'We don't get along well together', and 'We confide in each other'. In line with common practice by Epstein and others, all subdimensions were combined into an overall index of dysfunction. Positively worded items were reversed so that higher scores reflect higher dysfunction, α = .91.
Perceived parental relationship quality was measured with five items measured on a scale from 1 to 7 and inspired by similar items used previously by Feeney and Collins (2003): 'My parents are proud of each other', 'My parents love each other', 'My parents accept each other as they are', 'My parents support each other', and 'My parents care about each other', α = .94. Positive lay beliefs about romantic relationships were measured with a scale specifically constructed for this investigation to tap into beliefs that finding enduring love is possible (or not). 10 items were presented under the heading 'To your mind, how much are these statements true?' (1 = "not at all" to 7 = "very much"). Items were: '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 sceptical 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 = .92.
The questionnaires for both the pilot and the main study also included a number of other questions which are not relevant in the present context. These entailed scales tapping into general well-being indices such as self-esteem; but because the variables were not of theoretical interest here, they will not be discussed further. 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.

Results and Discussion
The mean scores for family dysfunction (M = 4.72, SD = 1.02) and for positive lay beliefs (M = 5.53, SD = 1.29) were well above the scale midpoint (4). As expected, the bivariate correlation between family dysfunction and positive lay beliefs was negative and significant, r = − .41, p < .001. Moreover, and confirming the validity of the lay beliefs measure, this measure was positively associated with parental relationship quality, r = .25, p < .001, so that those who perceived their parents to have a good relationship also reported more positive lay beliefs about their own ability to find enduring love. These first findings demonstrate that the measure we devised for 'positive lay beliefs about romantic relationships' is reliable and that it relates in expected ways to another construct. As expected, experiencing dysfunction in the family of origin was negatively associated with positive lay beliefs about the possibility to find enduring love. On the basis of these encouraging results, the main study was designed to test the full theoretical model.

Participants
Four hundred and thirty-five people participated in this online study (mean age = 24.63 years; SD age = 11.02; 109 males, 319 females, and 7 chose the category 'other' to specify their gender). There was no overlap in the participants between the main and pilot studies. Of the total sample, 221 reported currently being in a romantic relationship; 128 reported that their parents were divorced, and 307 reported that their parents were not divorced. This was a convenience sample accessed by research assistants recruiting via social media through private accounts on platforms such as Facebook, Instagram, Reddit, Snapchat and Twitter, and encouraging distribution of the survey link by social media users to others to encourage a snowballing effect. Everyone who was willing to participate was included in the final sample. This cross-sectional correlational study included the following measures:

Measures
To measure family-of-origin dysfunction, participants were asked to think about their family growing up, and to state how much statements based on the Family Assessment Device (Epstein et al., 1983) were true (1 = "not at all" to 7 = "very much"); α = .94. Although in the pilot study participants had been asked to think about 'everyone in their household', this time participants were asked specifically to think about their 'family growing up', to avoid the problem that older participants who have already founded their own families would think of that family rather than the family of origin. Items were also put into the past tense (e.g. rather than 'We confide in each other', 'We confided in each other'). With this new wording, participants were clearly prompted to think about family dynamics in the family of origin in the past, and participants of all ages could be included in the sample.
Whereas in the pilot study only 24 of the 31 items were used because this improved reliability, this time it was possible to include all 31 items in the scale, meaning that these were the additional items that were included: 'When we didn't like what someone had done, we told them' (for 'communication'), 'When you asked someone to do something, you had to check that they did it' (for 'roles'), 'We had rules about hitting people' and 'If the rules were broken, we didn't know what to expect' (both for 'behaviour control'), and 'Individuals were accepted for what they are', 'We avoided discussing our fears and concerns', and 'Making decisions was a problem for our family' (all three for 'general functioning').
Positive lay beliefs about romantic relationships were measured with the same 10 items as in the pilot study (1 = "not at all" to 7 = "very much"); α = .90.
Current romantic relationship dysfunction and current romantic relationship satisfaction items were only answered by those participants who reported currently being involved in a romantic relationship.
To measure romantic relationship dysfunction at present, participants again answered items based on the Family Assessment Device (Epstein et al., 1983), but this time they were asked to state how much the statements are true for their 'current romantic relationship right now' (1 = "not at all" to 7 = "very much"). For example, 'there are lots of bad feelings in our relationship', and 'We confide in each other' (reversed). Used were all 11 items that tapped into 'general functioning', α = .84.
Romantic relationship satisfaction was measured with three items taken from La Guardia et al. (2000). The items were 'I feel loved and cared about by my partner', 'I feel a lot of closeness and intimacy with my partner', and 'I feel a lot of distance in our relationship' (reversed; 1 = "strongly disagree" to 7 = "strongly agree"); α = .79.

Results and Discussion
Descriptive and bivariate correlations for all measures are displayed in Table 1. To test the model whereby dysfunction in the family of origin would affect lay beliefs, which in turn would affect romantic relationship dysfunction, which in turn would affect romantic relationship satisfaction, a structural equation model was built using AMOS. Of course, because the model included measures regarding the current romantic relationship, only the sub-sample of participants who were currently in a romantic relationship and who had answered those questions were included in this analysis. A direct path from family-of-origin dysfunction in the past to dysfunction in the present romantic relationship was also included, because the previous findings show that other factors are also likely to mediate the effects between the two variables (see Johnson et al., 2015;Maleck & Papp, 2015). Inclusion of the direct effect allows for both the estimation of the path of theoretical interest here, whilst acknowledging additional processes that have previously been demonstrated, for example traumainduced emotional detachment (Scholte & van der Ploeg, 2007). Standardised path values are displayed in Fig. 1. All paths that were included in the model are depicted in the figure, and if variables are not linked by a path in the figure that means their direct relationship was assumed to be zero during the analysis.
The model fit the data extremely well, χ 2 (2) = 1.16, ns; CFI = .99; stRMR = .02; RMSEA = .01 (CI RMSEA-low = .00; CI RMSEA-high = .12). For reference, rule of thumb guidelines for acceptable model fit suggested by Hu and Bentler (1999) are a CFI of .95 or above, a RMSEA of .06 or less, and a stRMR of less than .08. All paths were significant and in the expected direction. Bias-corrected bootstrapping was performed to estimate the indirect effects. The indirect effect of family dysfunction in the family of origin on dysfunction in the present romantic relationship via lay beliefs about romantic relationships was significant, b = .13, p < .001, CI lower = .05, CI higher = 16. The indirect effect of family dysfunction in the family of origin on romantic relationship satisfaction, via both lay beliefs about romantic relationships and romantic relationship dysfunction, was also significant, b = − .25, p < .001, CI lower = − .39, CI higher = − 19.
Next, we tested (now based on the full sample, i.e. including both participants who were and who were not in a romantic relationship) whether having experienced parental divorce would be related to positive lay beliefs about romantic relationships. When conducting an ANOVA with parental divorce (yes vs. no) as independent factor and positive lay beliefs as the dependent variable, this yielded a clear significant effect, F (1, 382) = 16.62, p < .001, ηp 2 = .04. Those participants whose parents had not divorced had more positive lay beliefs (M = 5.78) than those who had seen their parents divorce (M = 5.30).
Overall, then, there was strong evidence in these data that experiences of family dysfunction and also parental divorce impact on participants' lay beliefs that finding enduring love is possible, and these beliefs mediate and explain the effect of family-of-origin experiences on own romantic relationships.

General Discussion
Across both the pilot study and the main study, results confirmed the hypotheses. A belief that finding enduring love is possible was influenced by dysfunction in the family of origin. Positive lay beliefs, in turn, negatively predicted dysfunction in the current romantic relationship, and romantic dysfunction negatively predicted romantic relationship satisfaction. The significant indirect effects confirmed that lay beliefs were a significant mediator of the effect of past dysfunction on present dysfunction, and of the effect of past dysfunction on present relationship satisfaction. Moreover, having gone through the experience of parental divorce was also, as expected, associated with diminished positive lay beliefs about the possibility of finding lasting love. There are several theoretical implications of the findings which should be flagged. Although positive lay beliefs about romantic relationships are not the only factor that might explain the impact of family-of-origin dysfunction on later adult intimate relationships, the present results suggest that they are a very important factor, and that these beliefs therefore add an important piece to the puzzle, alongside other previously identified mediators such as mental health problems and negative couple interaction (Johnson et al., 2015;Maleck & Papp, 2015). Although other lay beliefs regarding romantic relationships have previously been investigated (e.g. Eidelson & Epstein, 1982), lay beliefs specifically about whether it is possible to find enduring love have not been studied. However, it is those types of beliefs that can be assumed to be most impacted by experiencing parental relationship dysfunction and dissolution, and the present contribution presents an important innovation by highlighting this. The present findings therefore build on and extend previous research.
The focus on lay beliefs about the possibility of finding enduring love is novel and important. Previous work has not focussed specifically on beliefs about whether or not finding lasting love is an attainable goal. This is surprising, because in some ways beliefs about whether lasting love is possible are more fundamental than beliefs about what relationships should look like (e.g. a belief that mindreading should happen in healthy relationships, Zagefka et al., 2020). The question of whether one can expect to find lasting love is more fundamental than the details about expectations regarding what such a relationship might look like. The question of whether someone expects to be able to have a lasting relationship is primary to the question of what might be expected within such a relationship. Studies that focus, for example, on beliefs about mindreading whilst ignoring beliefs about the possibility of finding enduring love therefore risk ignoring the factors with potentially the greatest predictive power. The implications of the present findings are therefore that future studies on lay beliefs about romantic relationships should certainly consider including beliefs about the possibility of finding enduring love into their focus.

Implications for Practice
The present data show the power of self-fulfilling prophecies in the context of romantic relationships. They show that people who do not believe that it is possible to find lasting love (because they have not had this positively role-modelled whilst growing up) will have less functional romantic relationships as a consequence, and that they will experience more dissatisfaction with their romantic relationships. Therefore, an important vantage point for therapists would be to address unduly negative lay beliefs about relationships, by identifying and facilitating awareness of potentially unconscious assumptions, so that couples can respond consciously rather than be driven by subconscious beliefs.

Limitations
Of course, there are several limitations to the present research that should be acknowledged. First, the present results are based on participant self-reports, which might be inaccurate. Human memory is fallible (Herrmann et al., 2005), which means that participants' reports of dysfunction in the family of origin might be prone to inaccuracies especially for older participants for whom those experiences lie a long time in the past. At the same time, there is also evidence that recall of particularly stressful life events is often quite accurate (Baugerud et al., 2014). Moreover, participants of the pilot study were quite young, and maybe less prone to retrospective biases than older adults. So, there is reason to have a measure of confidence that the present self reports, whilst not perfect, are still a good proxy of actual experiences.
Second, the data are correlational, which means that although the pattern of results obtained was compatible with the hypotheses, there is no evidence that the causality of the observed effects is in the hypothesised direction. To gain greater certainty, experimental designs would need to be employed. However, there are obvious ethical issues that would arise from trying to manipulate family dysfunction. So, cross-sectional correlational data might be the best feasible proxy for studying the effects of interest, short of conducting longitudinal studies spanning at least two decades.
Although studies with such designs exist (see e.g. Seery et al., 2010, who link adverse life events to later resilience), they are clearly extremely costly, and therefore not always practical.

Future Research
There are several avenues that could be explored by further research, and for brevity's sake only three will be highlighted here. First, within relationships research, the dyadic analysis, based on data from both people in a couple, is often seen as the gold standard (e.g. Nisenbaum & Lopez, 2015). The present contribution falls short of that, and future research could usefully improve on this by investigating the effects of positive lay beliefs within dyadic designs. Such an approach could test whether one partner's negative lay beliefs impact on the other partner's relationship satisfaction, or whether lay beliefs of both partners interact, so that one person's more positive outlook presents a protective mechanism against the destructive effects of the other person's more negative beliefs. Questions like these will only be addressable in dyadic designs.
Second, as indicated previously, other mediators of the effects of dysfunction when growing up have been identified, such as mental health problems and negative couple interaction (Johnson et al., 2015;Maleck & Papp, 2015). It remains an open question whether lay beliefs highlighted here present a process that runs in parallel to those other processes, or whether these mechanisms are in fact sequential. It is possible that lay beliefs provide an independent route from past to future, but it might also be the case that past dysfunction might lead to more negative lay beliefs, which in turn impact on mental health and interactions, which in turn lead to certain outcomes. Which of these options applies would be a fascinating question to address in future research.
Third, as already touched upon above, it would be useful to study the lay belief highlighted by the present research in conjunction with other lay beliefs about romance that have previously been studied, to substantiate the claim that beliefs about whether or not enduring romantic love is attainable has primacy over other beliefs in predicting important relationship outcomes. Such a study could take a comprehensive focus and investigate the effects of many different lay beliefs, rather than just a selected few.
To conclude, the main value of this research lies in the introduction of a novel concept, that of lay beliefs about the possibility of finding enduring love. This new concept was found to be important in explaining the effects of family-of-origin experiences on later adult romantic experiences. It is hoped that future work will follow-up by further exploring the specifics of this important, novel concept.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.