Abstract
Previous research reveals that children and parents are not in complete agreement over which traits are most important for the mate of the child. Children tend to prefer traits that suggest genetic quality, whereas parents prefer characteristics that suggest high parental investment and cooperation with the ingroup. Using a sample of parents, mothers (n = 234) and fathers (n =240) the hypothesis was supported; parents perceived characteristics indicating a lack of genetic quality as being more unacceptable to the child, while characteristics indicating a lack of parental investment and cooperation with the ingroup were more unacceptable to themselves. Sex differences between mothers and fathers and sons and daughters were explored.
Introduction
When a new baby is born, parents often imagine the possibilities of what their child will be like when he or she grows up. Will the child be a doctor, a football player, or perhaps the next president of the United States? Then suddenly, parents are bombarded with a thought that may send a shiver down their spine: the realization that their child will one day want to date someone they don't like. Indeed, meeting children's potential partners can be a tumultuous time for parents and children as parents may not agree with their children's choice of partner. This type of conflict has permeated our popular culture through the medium of literature (Romeo and Juliet), film (Meet the Parents and Shrek II), and reality television (MTV's Parental Control). The conflict has also engaged the minds of researchers, in particular those in the behavioral and social sciences, and most recently, it has caught the attention of evolutionary psychologists who wish to explain the ultimate origins of this behavior and to expand our knowledge of human mating strategies.
The potential of parent-offspring conflict over mate choice was demonstrated in a series of studies by Buunk, Park, and Dubbs (2008), who showed that parents and children are not in complete agreement over what type of person would be an ideal partner. Although there is considerable overlap between parents and children in preferences, for example, both parents and children may ideally prefer a partner for their child who is intelligent and kind, the key is that the preferences are not identical. The research conducted by Buunk et al. (2008) was aimed at exposing these potential conflicts and found that children perceive that their parents would prefer themselves (the child) to have a mate with traits indicating high parental investment and cooperation with the ingroup, whereas children tend to prefer mates with traits indicating heritable (genetic) quality. For example, participants (children) indicated that traits connoting a lack of genetic quality, such as being physically unattractive, lacking a sense of humor, and lacking creativity, were significantly more unacceptable to themselves compared to their parents. Conversely, traits connoting a lack of parental investment or cooperation with the ingroup, such as different ethnicity, being divorced, and coming from a lower social class than self, were indicated by participants as being significantly more unacceptable to their parents. These results have been replicated in culturally diverse samples from the United States, Kurdistan, Iraq, the Netherlands, and a sample of international students studying in the Netherlands. Since then, samples from Uruguay (Park, Dubbs, and Buunk, 2009) and Argentina (Buunk and Castro Solano, 2010) have also confirmed these results. In accordance with the above research, Apostolou (2007b, 2008a,b) found that people have differing preferences for mates and in-laws. Physical attractiveness was a more highly desired quality in a mate than it was for an in-law and having a good family background was highly desired for an in-law than it was for a mate.
The limitation of the previous studies conducted by Buunk et al. (2008), Park et al. (2009), and Buunk and Castro Solano (2010) was that the samples were composed of young adults who indicated whether given negative characteristics would be more unacceptable to themselves or to their parents. In order to establish that parents also perceive a potential conflict with their children over the mate choice of the latter, it is necessary to replicate the previous findings with a sample of parents who have children of mating age. Confirmation of the previous results will further support the existence of a parent-offspring conflict over mate choice. It was predicted that parents will find characteristics connoting a lack of genetic quality as being more unacceptable to their child. Conversely, parents will find characteristics connoting a lack of parental investment and cooperation with the ingroup as being more unacceptable to themselves.
The reason for why parents and children have differential mate preferences can be explained with the help of two theories: parent-offspring conflict (Trivers, 1974) and evolutionary trade-offs (Gangestad and Simpson, 2000). While parents and children share approximately half of their genes, they do not always share the same genetic interests. Under the right conditions, it pays individuals to follow a selfish strategy, even if this is to the detriment of their parents and siblings. Hence, parents and children should develop different behavioral strategies that maximize benefits and minimize costs to their own advantage. A common example of parent-offspring conflict in mammals occurs during weaning. Since lactation is costly, and because a mother might have other offspring to care for or because she cannot reproduce again until their existing offspring are weaned, a mother ideally would prefer to wean her offspring relatively early. Even though this strategy may benefit the mother, it represents a cost to the offspring since it will no longer be able to take advantage of an easily obtained source of nutrients. Ideally, the offspring would prefer to prolong nursing even though it is a cost to the mother and other siblings. As noted by Trivers (1974), this phenomenon of parent-offspring conflict may be extended further to include the mate choice of children. In humans, one type of mate, with a particular set of characteristics, might benefit the offspring more than the parents, whereas another type of mate, with a different set of characteristics, might benefit the parents more than the offspring.
The key to which type of a mate choice would benefit a parent or a child more rests in the theory of evolutionary trade-offs (Gangestad and Simpson, 2000). Effectively, a child who mates with an individual high in genetic quality benefits by delivering genetic quality to their offspring. However, because genetic quality and parental effort tend to trade-off with each other (e.g. individuals higher in genetic quality tend to invest less in parental effort), the child runs the risk of having a low investing partner (Buss and Schmitt, 1993; Gangestad and Thornhill, 1997). This can lead to the child needing extra support from its parents in order to sustain themselves and the child's offspring. Relying on the parents for support is, of course, much more costly to the parents than it is for the child. Parents, who would ideally like to distribute their resources evenly to all offspring, would see this as detrimental to their other children and grandchildren. If the child would instead have opted for a partner with traits indicating high parental investment (lower genetic quality), then the parents would not need to invest extra resources into their child and grandchildren. This strategy is more beneficial to the parents, but can be costly to the child.
It is also important to note that the difference between the mate preferences (for the child's mate) of parents and children is not simply due to a generational effect or an effect of mate preferences changing as one grows older. Buunk and Castro Solano (2010) found that the degree of parent-offspring conflict was larger for older children. This is the opposite pattern one would expect if parent-offspring conflict was due to a shift in mate preferences. Further more, research from Apostolou (2008a) showed that a sample of parents preferred different traits for an ideal mate compared to an ideal in-law. Participants rated beauty as being significantly more important for a spouse. Thus, differences found between children and parents when indicating their preferences for a child's mate likely reflect the differential costs and benefits for children and parents.
Sex differences between mothers and fathers will also be explored. Previous research on parent-offspring conflict over mate choice has failed to find sex differences in the preferences that parents have for their in-laws (Apostolou, 2007b). That is, fathers and mothers rank which characteristics they preferred for an in-law similarly. However, due to the different life history strategies of men and women, it is worth re-examining. For instance, mothers are more highly invested (Trivers, 1972) in as well as more certain of their genetic relatedness to their children than are fathers (Buss and Schmitt, 1993; DeKay, 1995; Euler and Weitzal, 1996). Unlike men, women also under go menopause and lose their ability to reproduce directly (Timiras and Valcana, 1972). Because of these differences, mothers tend to play a larger role in the lives of their children, including their children's romantic lives (Bates, 1942; Faulkner and Schaller, 2007) as well grandparenting (Laham, Gonsalkorale, and von Hippel, 2005; Michalski and Shackelford, 2005). Fathers on the other hand, appear to use more direct methods of control their children's mate choice (Bates, 1942) and there are several examples in which fathers simply marry off their daughters to the wealthiest men in order to build alliances or gain social status (Chagnon, 1992; Shadle, 2003). Based on these sex differences, it is possible that sex differences may exist between mothers and fathers in the unacceptability of the negative traits however, no prediction is made in either direction.
The sex of the child may also dictate how unacceptable parents find undesirable characteristics and the potential for parent-offspring conflict. It is well established that parents are more concerned about their daughter's mating decisions. For example, parents tend to set more restrictions concerning dating and give daughters less freedom of choice to choose their own partner than sons (Faulkner and Schaller, 2007; Perilloux, Fleischman and Buss, 2008; Wight and Henderson, 2006). The reason is that making a poor quality mate choice is more detrimental for women, and because parents can be certain that grandchildren from a daughter are actually genetically related. Thus, it is predicted that if differences based on the sex of the parent's child exist, then parents will find undesirable characteristics in a child's mate choice as being more unacceptable if they have a daughter rather than a son. Additionally, since the trade-off between parental quality and genetic quality is more pronounced for women than men, it is predicted that our measure, potential for parent-offspring conflict score, will be greater for parents of daughters.
Lastly, we will examine if there are differences between people who are affiliated with a religious group (i.e., Protestant, Catholic, Muslim, etc.) and those to who are not (i.e., atheist, agnostic, spiritual, etc.). Since religion is a marker of group identity, it is possible that people who identify with a religious group might be more concerned about traits which indicate cooperation with the ingroup, for example such parents may find having different religious beliefs or a different ethnic background, as more unacceptable compared to those who do not.
Materials and Methods
Participants
Four hundred seventy-four parents (fathers n = 240; mean age = 51.89, SD = 5.67; mothers n = 234; mean age = 48.20 SD = 5.54) participated in an online questionnaire. The religious background of the participants was 42% Catholic, 34.8% atheist/agnostic/non-religious, 17.3% Protestant, .6% Muslim, .2 % Jewish, and other 5.1%. The questionnaire was conducted in Dutch and was administered by Flycatcher, an independent research agency that conducts online survey collection in samples representative of the Dutch population. Participants are rewarded for completing a questionnaire, including our own, by receiving a credit for an online store. Flycatcher has a data-base of potential participants and we were able to specify the exact type of participant that we wanted to include in our study. We requested an equal ratio of men and women, none of the participants to be married to each other, and for each participant to have an eldest child between the ages of 15 to 25. We chose this age range, 15 to 25, for a couple reasons. The average age of marriage in the Netherlands is 31.7 and 29.2 for men and women respectively (Eurosat, 2008), however, the frequency of cohabitation is high and a high percentage of women, 71.3%, have a steady partner by the age of 24 (Kalmijn, 2007). By age 15, many children begin to think about or have already begun to date. Thus, 15 to 25 seemed liked a reasonable age range to use for the study.
Procedure
A shortened version of the parent-offspring conflict over mate choice questionnaire used in this study was based upon that used in previous studies (Buunk et al., 2008; Buunk and Castro Solano, 2010; Park et al., 2009). From the original studies, six traits that were consistently shown to be unacceptable to the child (physically unattractive, physically unfit, overweight, bad smell, lacks creativity, and lacks a sense of humor) and the six traits that were consistently shown to be unacceptable the parents (bad family background, lower social class than self, different ethnic background, different religious beliefs, divorced, and poor) across the five samples were retained. Thus, the questionnaire contained 12-items, six traits indicating poor genetic quality and six traits indicating a lack of parental investment and cooperation with the ingroup. All traits were formulated to represent the undesirable variant of trait variables (i.e., physically unattractive, different religious beliefs). Each question was posed in the following manner: “If my child's potential partner was [physically unattractive], this would be…” followed by a 7-point scale on which the possible answers ranged from (1) “Much more unacceptable to Myself [parent]” through (4) “Equally unacceptable to both Myself and my Child” to (7) “Much more unacceptable to my Child”. Participants were instructed to answer the following questions keeping in mind their eldest child. The children of the participants included 239 sons (mean age = 19.80, SD = 3.12) and 235 daughters (mean age = 20.12, SD = 3.26).
Analysis
All 12 traits were compared to the total mean using a non-directional one-sample t-test to see if they differed significantly from the mean in the direction of the child or parent. Next, mean scores for the six traits indicating poor genetic quality and for the six traits indicating a lack of parental investment/cooperation were calculated. A paired t-test was conducted to see if the two means differed significantly from each other in the predicted direction of the child or parent respectively. A non-directional one-sample t-test was also conducted on these two means to see if they significantly differed from the total mean of all 12 characteristics in the predicted direction.
Sex differences between mothers and father as well as sons and daughters were examined next. A MANOVA was conducted on the 12 individual traits to see if there was a main effect of the participant's sex, the sex of their child, and if their were any interactions. A new measure was also calculated—the potential for parent-offspring conflict score—based on the measurement in Buunk and Castro Solano (2010). The potential for parent-offspring conflict score was calculated by subtracting the score of each individual trait from the scale midpoint of 4 and then taking the absolute value. By doing this, we could look at how each individual trait varied from the scale midpoint. A higher value indicates that a given trait deviates more from the middle of the scale, and thus, the trait may represent a greater source of conflict between parents and offspring. A mean score was calculated for all 12 traits combined. The potential for parent-offspring conflict score was also included in the MANOVA to see if there was a main effect of the participant's sex, the sex of their child, and if there were any interactions.
Lastly, a one-way ANOVA between people who belong to an organized religion (Protestant, Catholic, Muslim, etc.) versus people who do not (non-religious, agonistic, atheist, spiritual, etc.) was conducted on the mean of the six traits indicating poor genetic quality and the mean of the six traits indicating a lack of parental investment/cooperation.
Results
Parent vs. child
The mean score for the 12 items was 3.94, which we used as the comparison point. This allowed our results to be comparable with previous studies (see Buunk et al., 2008). We conducted a non-directional one sample t-test on all 12 of the traits to see if they varied from the total mean in the predicted direction, either towards the parent or the child. Table 1 lists the results of the t-tests. Values higher than 3.94 indicate that a trait has a relatively higher degree of unacceptability to the child, and values lower than 3.94 indicate that it has a relatively higher degree of unacceptability to the parents. As displayed in Table 1, four out of the six traits indicating poor genetic quality were found to be significantly different from the mean in the direction of the child. For the traits indicating a lack of parental investment/cooperation, five out of the six of the items were found to be significantly different from the mean in the direction of the parent. When examining all 12 traits, the two that were found to be the most unacceptable to the child were physically unattractive and overweight. The traits that parents found to be most unacceptable were different ethnic background and divorced.
Mean Levels of Unacceptability of Characteristics to Self versus Child
Note: Lower values indicate greater unacceptability to the parents and higher values indicate greater unacceptability the child; asterisks indicate significant differences from the mean score for all 12 items (3.94) in the predicted direction:
p ⩽ .05
p ⩽ .01
p ⩽ .001 (two-tailed). Approached significance (p = .085).
Next, a mean score was computed for traits indicating poor genetic quality (M = 4.07, Cronbach's alpha = .73) and for traits indicating a lack of parental investment/cooperation (M = 3.82, Cronbach's alpha = .82). These two means differed significantly from each other in the predicted direction, t(473) = 6.38, p < .001, and they also differed significantly from the sample mean of 3.94. That is, the mean of the traits indicating poor genetic quality varied significantly from the sample mean, t(473)= 3.28, p = .001, in the direction of the child; the mean of the six traits indicating a lack of parental investment/cooperation significantly varied from the sample mean, t(473) = −3.27, p = .001, in the direction of the parent. Thus, overall, the hypothesis received strong support; parents regarded the six traits indicating poor genetic quality as being more unacceptable to their child and six traits indicating a lack of parental investment/cooperation as being more unacceptable to themselves.
Sex differences
The MANOVA on each of the 12 individual traits and the potential for parent-offspring conflict score using the participant's sex and sex of the child as fixed factors, found minor sex differences. To correct for multiple testing, the Bonferroni correction for alpha inflation reduced the level of significance to .004. Overall the sex of the parent had little to no effect on the unacceptability of the negative traits and the potential for parent-offspring conflict score. The only exception was for fathers who significantly rated the trait lacks a sense of humor as being more unacceptable to themselves than did mothers, F(1, 470) = 9.44, p = .002 (father M = 3.75, SD =1.26; mother M = 4.10, SD = 1.26). The sex of the child did have an effect on the unacceptability of a few traits; however the p-values were higher than the Bonferroni correction of .004. The traits that were found by parents to be more unacceptable for a daughter were different religious beliefs F(1,470) = 6.70, p =.008 (male M = 3.99 SD = .99; female M = 3.72, SD = 1.20), divorced F(1,470) = 3.87, p = .05 (male M = 3.87, SD =1.10; female M = 3.66, SD =1.22), and different ethnic background F(1,470) = 4.16, p =.04 (male M = 3.80 SD = 1.11; female M = 3.58, SD =1.31). The sex of the child also produced a significant difference in the potential for conflict score (M =.70, SD =.57), with parents of females having a higher potential parent-offspring conflict score than parents of males F(1,470) = 9.58, p = .002 (male M = .62, SD = .53; female M = .78, SD = .61). There were no interactions between the sex of the parent and the sex of the child.
Religious affiliation
The one-way ANOVA conducted on parents who identify themselves as belonging to an organized religion (Catholic, Christian, Muslim, etc.) versus those who do not (non-religious, agonistic, atheist, spiritual, etc.) produced a significant main effect. People who identify themselves as belonging to an organized religion, compared to those who do not, have a significantly lower mean for the six traits indicating parental investment/cooperation, F(1,473) = 6.719, p = .01, (organized religion M= 3.75, does not M= 3.95). This indicates that parents who belong to an organized religion view negative traits reflecting a lack of parental investment and cooperation with the ingroup as being more unacceptable in a child's romantic partner than parents who do not belong to an organized religion.
Discussion
Kin, in particular the parents, play a large role in mating decisions of humans. In order to gain a more complete understanding of how human mating decisions are made, it is important to explore the preferences of the individuals who influence our choices, namely the parents. Using the same methodology as past studies, which was specifically designed to uncover the potential sources of conflict between parents and children (Buunk et al., 2008; Buunk and Castro Solano 2010; Park et al., 2009), we found support for the hypothesis that children may especially prefer partners with traits connoting genetic quality whereas parents may especially prefer a child's partner to have traits connoting parental investment and cooperation with the ingroup. Overall, the results from this study, which used a sample of parents, confirmed those from our previous studies. This was a particularly important finding as our previous studies only contained samples of children who were of mating age. Taken together, the existence of a parent-offspring conflict over mate choice is more strongly supported.
The mean of the six traits indicating a lack of genetic quality (physically unattractive, overweight, physically unfit, lacks creativity, bad smell, and lacks a sense of humor) and the mean of the six traits connoting a lack of parental investment and cooperation with the ingroup (lacks a good family background, different ethnic background, divorced, lower social class than self, different religious beliefs, and poor), differed significantly from each other and also from the mean of all 12 items in the predicted directions. That is, when combined, the traits indicating a lack of genetic quality were found to be significantly more unacceptable to the child, while the traits indicating a lack of parental investment and cooperation with the ingroup were found to be more unacceptable to the parent (the participant). Individually, four traits out of the six indicating a lack of genetic quality were significantly different from the mean in the direction of the child. These traits include physically unattractive, overweight, physically unfit, and lacks creativity. Lacks a sense of humor and bad smell did not differ from the mean in either direction. The reason might be that fathers tend to value a sense of humor. Indeed, when examining the individual traits, fathers in this study significantly rated lacks a sense of humor as being more unacceptable to themselves while mothers rated it as being more unacceptable to their child. In past studies, lacks a sense of humor was rated as being one of the most unacceptable traits across nearly all of the samples (Buunk et al., 2008; Buunk and Castro Solano, 2010; Park et al., 2009). Thus having a sense of humor might not only indicate a trait that has been sexually selected for, but also one that men in particular use to bond with kin and perhaps in-laws.
Of the six traits indicating a lack of parental investment and cooperation with the ingroup, four produced a statistically significant difference from the mean in the direction of the parents. These traits included lacks a good family background, different ethnic background, divorced, and lower social class than self. Having different religious beliefs was marginally significant in the direction of the parents. In the child samples, this trait was found to be significantly more unacceptable to the parents. One reason for why it was only marginally significant in this study may be due to the high degree of secularism in the Dutch population. In our sample 34.8% of the participants reported being agnostic, atheist, or non-religious. Additionally, the mean religiosity of the sample was 2.82 (SD = 1.72) on a scale ranging from 1 (not at all religious) to 7 (very religious). Other studies reveal that having same religious beliefs is often an important basis for marriage (Dugsin, 2001; Robinson, 1980; Talbani and Hasanali, 2000). However, in spite of the fact that the Netherlands is a highly secular country, we were able to still find an effect in the predicted direction. The trait poor, on the other hand, did not produce a significant result in either direction.
On a related note, parents that reported belonging to an organized religion (Protestant, Catholic, Muslim, etc.) found the six traits indicating a lack of parental investment and cooperation with the ingroup as being more unacceptable in a child's potential romantic partner than did parents who do not belong to an organized religion. This makes sense, as people who identify with a group are probably more likely to want their child to choose partners who are from the ingroup. Research, in particular on South Asians, is consistent with these results in that parents want their children to marry someone who is of the same religion, language, ethnicity, and caste (Das Gupta, 1997; Netting, 2006; Robinson, 1980; Vaidyanathan and Naidoo, 1990/1991; Zaidi and Shuraydi, 2002).
In addition to complimenting the results from the child samples, the data was also used to explore sex differences between mothers and fathers as well as sons and daughters. The potential for parent-offspring conflict score, which measures the degree to which the individual traits deviate from the mid-point, was also included in the examination of sex differences. There was almost no effect of the sex of the parent on the unacceptability of traits in a child's potential mate choice, which is accordance with results from Apostolou (2007b). However, the sex of the parent's child did have a small effect. Parents tended to rate the traits different ethnic background, different religious beliefs, and divorced in a child's potential partner as more unacceptable to themselves when their child was female rather than male. Additionally, the potential for parent-offspring conflict score was higher for parents of females than for parents of males. It should be noted that caution needs to be taken in interpreting the potential for parent-offspring conflict score, as deviation from the mid-point may or may not indicate that a trait is a source of conflict between parents and children. However, we feel in this case that it may represent a potential for more conflict, as there was a greater tendency for parents to rate certain traits as being more unacceptable for females than males. Overall, these results fit the idea that parents more highly guard daughters and set more restricts concerning dating compared to sons (Dugsin, 2001; Perilloux et. al, 2008; Talbani and Hasanali, 2000; Wight and Henderson, 2006).
Although the results largely support those from previous studies, there are some important limitations of our study. To begin with, parents in our study were instructed to think of their first born child, therefore the results might not be generalizable to later born children. Secondly, the study is self-report and measures parent's perceptions of how their preferences for their child's mate choice may differ from those of their children. Given this, the results found in our study may not actually reflect the differences in the preferences that parents have for their in-laws and children have for their mate choice. However, since the results are consistent with those from the previous child samples (Buunk and Castro Solano 2010; Buunk et al., 2008; Park et al., 2009) and with research on people's preferences for in-laws and mate choice (Apostolou, 2007b, 2008a,b), we feel that our study does tap into the differential preferences parents and children have for the mate choice of the latter. Thirdly, perhaps due to the design of our study, we failed to find sex differences between fathers and mothers. We feel that there are theoretical reasons for differences to exist, and although differences between fathers and mothers may be subtle, we encourage researchers to continue to investigate this issue.
Overall these findings are consistent with our previous research, other research concerning parental influence, and also with the observed sex differences in the treatment of sons and daughters. Although, this study adds to the growing knowledge on parental influence of mate choice, it also generates more questions. How do parents influence their children's mate choice and do children have ways to either counter or manipulate their parents to their own means? Even though our study, as well as previous studies (Apostolou 2007b) failed to find sex differences between mothers and fathers in their preferences for the mate choice of their child, there might be differences in the way that each attempts to influence there child's mate choice. This should be explored. Is there another type of parent-offspring conflict between sons and their parents concerning the right time to “settle down”? Results from Apostolou (2009) using parent-child dyads reveal that there is a parent-offspring conflict over whether a child should engage in a short-term or long-term mating strategy. As one can see, parental influence over mate choice is in a nascent state and more research will help draw a more complete picture of human mating behavior.