Acts that appear altruistic – where one individual incurs an immediate cost in order to confer a benefit on another individual – can potentially lead to an increase in the actor’s inclusive fitness that outweighs its short-term costs . Such behaviours can earn direct fitness benefits as a result of mutual benefit, reciprocity, evasion of punishment or enhancement of social status, as well as indirect (kin-selected) fitness benefits (reviewed in [2–4]). An evolved preference for other-regarding behaviour in potential mates has been proposed as an additional mechanism by which helping behaviours can yield direct fitness benefits in humans [5–8]. This hypothesis has not yet received the level of empirical attention paid to other mechanisms by which helping behaviours earn direct fitness benefits.
An evolved preference for helping behaviour in mates is predicted to arise when parents need to cooperate to raise offspring successfully. The incidence of biparental care across the tree of life is sparse, but it is common in birds  and canids  and is practised by a significant minority of cichlid fishes  and primates , including humans. (See also ). Human babies are born at a very early stage of development and so depend on parental care for survival [14, 15]. A correlation between offspring survival and paternal care, especially under subsistence conditions , has been cited as one explanation for the prevalence of social monogamy and biparental care in humans ([17, 18], see also ). While women pay the larger minimum cost of reproduction due to pregnancy, lactation and a lower reproductive rate, typical paternal investment in our species can be significant [16, 20, 21]. Thus, both sexes may be predicted to exhibit a significant degree of mate choice based on the likely genetic and/or non-genetic benefits of mating with different individuals [22–24], especially when considering long-term relationships with a higher perceived chance of reproduction .
Animals may choose mates based on signals of genetic quality (; see also . Thus individuals may employ high-cost signals to communicate to potential mates their ability to supply ‘good genes’ and so high-fitness offspring. Therefore it has been hypothesised that helping behaviours may act as costly signals for genetic benefits such as high intelligence or good physical health [6, 8, 28]. Further, some helping behaviours involve risking one’s own life or physical wellbeing (e.g. jumping into a river to save a drowning person) and there is evidence that risk-taking or ‘heroic’ acts, whether they are apparently helping behaviours or not, function as costly signals of genetic quality in human males . Perhaps more pertinent to understanding the likely links between sexual selection, parental care and mating systems  is the hypothesis that one or both sexes may choose mates based on their likely investment in parental care. Indeed, it has been shown empirically that ability and willingness to contribute to childcare may make a man more attractive as a long-term sexual partner [31, 32]. Further, a classic study by Buss  showed that both men and women value traits such as kindness, sympathy and helpfulness in potential mates, and it might logically be argued that these are facets of what we might call a cooperative personality. Thus, if we think of biparental care as a form of cooperation (offspring being a public good shared by the parents), we might expect that people who display a cooperative or helpful phenotype in non-mating contexts may be perceived as more likely to cooperate in a care context and thus more desirable as sexual partners. According to this hypothesis, non-heroic helping behaviours could function as a signal of ability and/or willingness to supply non-genetic benefits to future mates .
Four experimental papers have recently tested the hypothesis that cooperative or helping behaviour is attractive. Farrelly et al.  report that in dyadic economic games, partners who act more cooperatively are perceived as more attractive; conversely, people preferentially directed cooperative behaviour towards more attractive members of the opposite sex. However, this purely monetary approach makes it difficult to disentangle the effect on attractiveness of cooperativeness from a simpler effect of economic resources, which are known to be valued by women [35–37]. In the second study, Phillips et al.  used a survey method to report that people find helping behaviours attractive in potential mates and that this preference is more pronounced in females (interestingly, these authors later used a twin study to discover significant genetic effects on this preference: ). However, because this study did not compare helping behaviours with a neutral control, it does not allow inferences to be drawn about whether helping behaviour has an absolute positive effect on attractiveness. Third, Barclay  presented men and women with vignettes and photographs describing opposite-sex individuals; each participant saw four vignettes, two of which included information on helping activities and two of which presented information on neutral (i.e. not helpful but not selfish) activities. Both men and women rated individuals described as taking part in helping behaviours as more attractive as potential long-term romantic partners; women also preferred helpful males for one-night stands, while this preference was not found in men. Most recently. Farrelly  provided evidence that fertility (stage of the menstrual cycle) has little effect on female preferences for helpful males and interpreted this as being consistent with women perceiving helping behaviours as a signal of likely non-genetic benefits [19, 32, 41]. However, this seems at odds with Barclay's finding that women also found helpful behaviour attractive in a partner for a one-night stand and in these latter two studies each participant rated only eight or four individuals respectively, making for a rather small sample size.
The results of these four studies are intriguing, but it is interesting to note the lack of a simple study that i) tests the hypothesis that a report of helping behaviour makes a given individual more attractive to individuals of the opposite sex, as compared with information that is neutral with regard to attractiveness, and ii) asks participants to rate a large number of opposite-sex individuals in order to reduce potential impact of the ‘baseline’ attractiveness of the individuals being rated. Our objective was to conduct such a study. We employed a within-subjects design to address three specific hypotheses. First, we predicted that heterosexual people would find members of the opposite sex more attractive if they were reported to take part in helping behaviours, as opposed to having a neutral activity (one that contained no information on helping behaviour) reported. Because we focussed on non-heroic, low-risk helping behaviours, our second hypothesis was that the effect of helping behaviour on attractiveness would be stronger when participants rated attractiveness for a long-term relationship as opposed to a short-term fling. This is because considering a long-term relationship is generally taken to lead participants to consider both likely genetic and non-genetic benefits, whereas a partner for a short-term fling is less likely to supply any non-genetic benefits (but see Discussion, below).