Facial attraction: an exploratory study of the judgements made by people with intellectual disabilities

Background Although romantic or sexual attraction is a major research topic in the general population, little is known about people with intellectual disabilities ’ (ID) views of attractiveness. Methods Fifty-eight participants ( 16 – 40 years) took part in this exploratory study, 29 with ID and 29 without ID. Participants were shown 50 images of men or women ’ s faces and asked to rate how attractive they thought the faces were. Results A strong association was found between what men and women with ID and those without ID considered attractive in romantic partners. However, people with ID were more likely to consider themselves desirable to others. Conclusions The ﬁ ndings suggest that people with mild ID make the same subtle judgements about facial attraction as other individuals.


Background
The sexual repression of people with intellectual disabilities (ID) has been well documented (Brown 1994;McCarthy 1999). There have been commonly held contrasting misconceptions that people with ID are asexual and do not have the same sexual desires as others or that they are promiscuous (Brown 1994). Since the 1980s, there has been an increasing emphasis on the human rights of people with ID (Joint Committee on Human Rights 2008). This has helped to foster a growing awareness that people with ID, like anyone else, want and need personal and sexual relationships. However, despite enjoying greater autonomy, people with ID still find it difficult to develop the relationships they aspire to (Department of Health 2009). Their sexual and intimate lives often remain 'public affairs', overseen by parents, family members and/or carers (Rogers and Tuckwell 2016), and their rates of relationships and marriage are much lower than the wider population (Emerson et al. 2005). Within Scotland, only 5.1% of people with ID were married compared with 45.7% of the general population. (Scottish Learning Disability Observatory 2011).
The literature regarding people with ID's sexuality has predominantly focused upon sex education, sexual knowledge and sexual abuse, specifically within the context of risk and vulnerability (Fitzgerald and Withers 2013). It has been increasingly acknowledged that the views of people with ID should be considered (McDonald et al. 2016), and there has been an increase in research focused upon their own perceptions of their sexuality and relationships. This research has highlighted the barriers to establishing relationships that these individuals face. These barriers include their high dependency on others, limited privacy, restricted social opportunities and others' prejudice (Wilkinson et al. 2015). Thus, there is a continuing disconnect between the recognition of people with ID's rights to have the sexual experiences and the opportunities, freedom and support to practise these rights.
Meeting prospective partners is difficult when people have more limited social networks or fail to enter social spaces like work places (Emerson and Hatton 2008). With regard to what is desirable in a relationship, research has found that people with ID want to have romantic partners and live as couples and that they value kindness and companionship rather than financial security, social status or intelligence (Bates et al. 2016;Rojas et al. 2016). Bates et al. (2016) also reported that participants in their study appeared to hold less conventional views of physical attraction, such as preferring shortness in men. In fact, little is known about people with ID's views of attractiveness, and no research has specifically explored what people with ID consider to be attractive in romantic partners. However, there have been numerous studies that have explored attraction and partner selection for people without ID.
One area of research in the general population has concerned facial preferences. Little et al. (2011b) noted that preferences for faces can have an impact on a range of social outcomes, such as decisions about relationships, both romantic and platonic, employability and social exchanges. Furthermore, while 'good looks' have been identified as important in potential partners by both men and women (Buss and Barnes 1986), more recent evidence suggests that men are considerably more influenced by physical attractiveness than women (van Hooff, Crawford & Van Vugt, 2011). Judgements of facial attractiveness are influenced by both personal and shared preferences of attraction (Hönekopp 2006). Qualities such as symmetry and averageness appear to be preferred by adults from diverse cultures, suggesting people may use similar cues to judge attractiveness (Langlois et al. 2000;Little et al. 2011b). Due to the lack of research regarding people with ID's views of attraction, it remains unknown if they are also in agreement with the wider population and using the same cues to rate attraction. Evidence from other areas of research, such as the perception of emotional cues, would suggest that people with ID are less sensitive to some facial cues compared with the general population, as they appear to have more difficulty with the recognition of complex emotions like 'fear', 'disgust' and 'surprise' (Matheson and Jahoda 2005: 5-12).
The search for a romantic partner does not solely rest on our evaluation of others. As prospective partners, we are also the subject of evaluation by others. To find a partner, a person needs to identify people that fit their criteria of attractiveness. This is a mutual process and, in turn, they need to be attractive to the other person (Campbell and Wilbur 2009). Identifying a partner therefore involves making a social comparison about our position as a prospective partner.
Social comparison theory proposes that how we evaluate ourselves in comparison with others is key to the development of our sense of worth (Festinger 1954). These comparisons are influenced by our interpersonal experiences and relationships. As a population, people with ID are often subject to negative experiences, such as bullying and discrimination (Emerson 2010). Additionally, their relationship opportunities may be impeded by conventional ideas about attractiveness and stereotypes of disabled people held by society (Rojas et al. 2016). There is contradictory evidence about whether or not such negative experiences result in people with ID internalising a negative or stigmatised view of themselves (Dagnan and Waring 2004;Jahoda and Markova 2004). Moreover, if individuals engage in downward social comparison, in order to promote a positive view of self compared with peers, then they may wish to promote an optimistic view of their attractiveness (Monteleone and Forrester-Jones 2016). Therefore, it is unclear how people with ID's devalued social status will influence their sense of being desirable to others.
This exploratory study is the first to draw on innovative methods from attractiveness research in the general population, to compare how a group of participants with ID and a group of participants without ID view facial attractiveness. The study also aimed to explore if people with and without ID consider themselves as desirable to others and the nature of people's perceptions of themselves and others as romantic partners. For example, exploring people's attraction to different faces is a step towards understanding the role played by physical attraction in people with ID's relationship aspirations and choice of partner. Considering people with ID's sense of their own attractiveness also has important practical implications. An awareness that one's attraction to others may not be reciprocated could result in feelings of rejection and disappointment.

Participants
Twenty-nine adults with ID and 29 adults without ID were recruited from further education institutions and voluntary community organisations from urban settings in the central belt of Scotland. All participants were aged 16-40 years. This is typical of the age groups recruited for attractiveness and sexuality studies (Bale and Archer 2013;Katsena and Dimdins 2015;Wincenciak et al. 2015;Rojas et al. 2016). Those without ID were recruited from a range of college courses, including police services, politics, history and social sciences. To determine if potential participants with ID had sufficient expressive and receptive language to complete all components of the study, they were selected using criteria from the Adaptive Behaviour Scale (Nihira et al. 1993). These criteria ascertained whether they could (1) talk to others about sports, family, group activities and so forth; (2) use complex sentences containing 'because', 'but' and so forth; and (3) answer simple questions such as 'What is your name?' or 'What are you doing?' Participants were also excluded if they had sensory impairments that had an impact on their ability to take part in any component of the study. Attempts were made to have similarly matched groups with regard to age, gender and socio-economic status.
Following data collection, four participants recruited to the ID group were excluded from the analyses as their scores on the Wechsler Abbreviated Scale of Intelligence -Second Edition (WASI-II; Wechsler 2011) were above 70 and outwith the ID range, as defined by the British Psychological Society (2015).

Experimental tasks, interview and measures
The experimental tasks and interview were based on attraction research within the general population (Bale and Archer 2013;Wincenciak et al. 2015). All components of the study were piloted and delivered in the manner described below.

Background information
Background information was collected about participants' age, gender, relationship status, sexual orientation and socio-economic status. Socio-economic status was measured by the Scottish Index Multiple Deprivation (SIMD; Scottish Government 2016). A person's deprivation status is rated on a scale of one to five based upon their postcode, where one represents the most deprived areas and five represents the least deprived.
Attractiveness rating Control task. The aim of the control task was to establish whether participants could follow the instructions to complete the experimental tasks appropriately and understand the Likert rating scale used in the experimental tasks. Participants were asked to rate how much they liked a set of images (television programmes or food) using a five-point Likert scale. They were then asked to give reasons for their choices. Time was taken to check the participants' understanding of the rating scale. If required, the instructions and tasks were repeated to ensure that the participants understood what to do. Participants had the opportunity to ask questions.
Attractiveness rating task. Depending on their sexual orientation, participants were presented with a set of 50 images of either men or women's faces. Participants who identified as bisexual were asked to state their current preference. The faces were of 50 white men (mean age = 24.2 years, SD = 3.99 years) and 50 white women (mean age = 24.3 years, SD = 4.01 years), posed front-on to the camera with direct gaze and neutral expressions to control for possible effects of gaze and emotion cues on responses to faces. Images were aligned on pupil position and cropped so that clothing was not visible. These images have been used in other recent facial attractiveness studies (Fisher et al. 2014;Wincenciak et al. 2015). Although the faces are all white Eastern European, they represent systematic variations in key facial characteristics such as symmetry, which have been found to be consistent markers of attractiveness cross-culturally (Rhodes 2006). Participants were asked to rate how attractive they thought the images were on a five-point Likert scale by placing each image in to one of five boxes labelled: not at all, a wee bit, ok, quite, or very. The scale was visually represented using blocks of increasing size. To account for order effect, the order that images were presented in was alternated.

Semi-structured 'romantic partner' interview
The aim of this exploratory interview was to establish a dialogue with participants about their selections on the attractiveness task and their self-perceptions about being desirable to others. Participants were presented with the images they had rated the highest in the attractiveness task. They were then instructed to select the image they thought was the most attractive from the set, and asked 'Tell me what made you think this person is attractive'. This was followed by the closed questions 'Do you think this person would ask you out on a date?' and 'Tell me what you think they would say if you asked them out on a date?', after which participants' reasons for their answers were explored, 'Can you tell me the reasons that made you think that?'. This task was repeated for the set of images the participants rated as least attractive. To avoid order effects, the sets of the highest and lowest rated images were presented in a different order to each successive participant.

Wechsler Abbreviated Scale of Intelligence -Second Edition
To establish that recruited participants were in the correct groups, the WASI-II (Wechsler 2011) was administered as a measure of cognitive ability. The two-subtest form of the WASI-II was used to provide an estimate of full-scale IQ. Psychometric properties include good to excellent test-retest reliability across subtests (0.83-0.94) and composite scores (0.90-0.96), a high level of internal reliability (0.90-0.92), and acceptable (0.71) to excellent (0.92) concurrent validity.

Procedure
Both the participant information sheet and consent form were provided in an accessible format, and great care was taken to ensure that participants were aware of the purpose of the study and that participation was voluntary.
The researcher met with the participants in a private room at their college or day service, at a time convenient to them. Time was taken at the start of the session to establish rapport with participants to promote engagement. Participants were seen on their own, with four exceptions where, at the participant's request, staff joined the session to provide support with communication. Participants were asked for their socio-demographic details before the control task was carried out. They were then asked to complete the attractiveness task, followed by the semi-structured 'romantic partner' interview. The WASI-II was administered last because it is a test of ability, contrary to the spirit of the other tasks where the aim was to elicit the participants' views as experts. If the WASI-II had been administered first, it could have inhibited the participants' engagement. The semi-structured interview was audio recorded.

Pilot phase
Prior to the main interviews, the attractiveness rating task and semi-structured romantic partner interview were piloted with two adults with ID and two adults without ID. The pilot established that it was possible for the participants to sort 50 images within the proposed 1 hour timeframe and showed that the Likert rating scale was comprehensible for the participants. Moreover, the interview questions about the participants' attractiveness ratings were piloted. As a result of the pilot, some of the language used was simplified and closed yes/no options were used for the 'dating questions', instead of open-ended questions. One individual did not want to answer questions about dating because they were already in a relationship. It was therefore made clear to participants that their responses did not reflect upon or impact their current relationship status.

Analysis
Spearman's rho correlation coefficient was used to establish the association between ratings of attractiveness by people with and without ID. Prototype composite images were manufactured using specialist computer graphic software to visually represent the most and least attractive facial characteristics (i.e. average shape, colour, and texture information) as determined by the average ratings made by the participants from each group. These methods were designed for this purpose and are commonly used in facial attractiveness research. For a full account of the method, refer to Tiddeman et al. (2001).
Chi-square analyses were undertaken to examine group d1ifferences within the 'dating scenario' questions. Where the conditions for chi-square were not met, the Fisher's exact test was used.
In addition, recordings of the romantic partner interview were transcribed verbatim and content analysed (Strauss 1987). This process involved identifying the reasons that emerged from the participants' transcripts in relation to acceptance or rejection in dating scenarios and preferences in a romantic partner. Categories were then developed that reflected the type of views expressed. An independent rater was asked to assign the participants' reasons within each question into the type of categories that were developed. Agreement was evaluated by Cohen's kappa coefficient and indicated a strong level of agreement for all questions (McHugh 2012). The kappa values were (1) being asked on a date = 0.89 and (2) offer of a date accepted or rejected = 0.805. All analyses were two tailed as the study was exploratory in nature.
Two women with ID (one who identified as heterosexual and one who identified as lesbian) did not appear to understand the rating scale used for the experimental task; as such, their data were deemed unreliable and excluded from the analysis. Table 1 shows the socio-demographic characteristics of the 54 participants who took part in the study.

Participant characteristics
Groups were similar in terms of gender, age range and socio-economic status. Of the 14 women with ID who took part, three identified as lesbian, two of whom were in a relationship together. One man with ID identified as bisexual, and all other participants identified as heterosexual. All participants without ID identified as heterosexual, with the exception of one man who identified as bisexual. Both individuals who identified as bisexual expressed a preference for women at the time of participation. Participants with ID identified as either single (n = 14) or in a relationship (n = 11). The majority of participants without ID identified as single (n = 22), with six stating they were in a relationship and one was married. Most participants from both groups were living in their family home. The WASI-II scores indicated that the cognitive abilities of the two groups were at the expected levels of ability (mild to moderate ID or average ability) for their age group. Four participants without ID refused to complete the WAIS-II. Their level of ability was indicated by their enrolment on a mainstream college course, such as social sciences. SIMD was not calculated for nine participants with ID because they did not provide a postcode. Participants' socio-economic status, in both groups, were spread across the range of SIMD quintiles from the most deprived areas in Scotland to the most advantaged.

Ratings of attractiveness
The following findings represent ratings of attractiveness by heterosexual participants. For each image, the mean attractiveness rating was calculated by collapsing (averaging) scores across participants in each group. Meaningful comparisons could not be made for lesbian or bisexual participants due to the small number of participants recruited.

Within group
Agreement of the ratings of attractiveness across heterosexual participants were highly consistent within groups, as indicated by Cronbach's alpha; women with ID = 0.94, men with ID = 0.90, women without ID = 0.95, men without ID = 0.96. Figure 1 illustrates the most and least attractive prototype composite images, from the heterosexual participants' average ratings per group. The high attractiveness composite images for both groups have skin colouration that has a healthy glow. They are slimmer and have a more positive demeanour, for example, a slight smile. In comparison, the low attractiveness composite images have an unhealthier pallor, their faces appear heavier and they have a more negative demeanour. These differences were consistent across all participant groups, which suggests that they all used similar visual cues to form impressions of attractiveness.

Between group comparison
The consistency between groups demonstrated by the prototype images was further supported by highly correlated ratings of attractiveness. Spearman's rho correlation coefficient indicated there was a statistically significant association between both groups' ratings of attractiveness for heterosexual men (rho = 0.42, P = 0.002) and women (rho = 0.70, P < 0.001), suggesting that there was some shared idea of attractiveness between groups. There was a stronger association between women than men. Scatterplots (Fig. 2) illustrate the associations between group ratings for women and men. Accepted or rejected for a date. A statistically significant difference was found between people with ID and those without ID's view about whether the person they had rated the most attractive would ask them on a date (χ 2 (1) = 8.295, P = 0.004). Twenty (80%) participants with ID said they would get asked out, compared with 12 (41.4%) participants without ID. When broken down by gender, a statistically significant difference was found between men (P = 0.001, two tailed, Fisher's exact test), with all 11 men with ID stating they would be asked out compared with five (35.7%) men without ID. No statistically significant difference was found between women (χ 2 (1) = 0.909, P = 0.340), although the findings were in the same direction. Nine (64.3%) women with ID said they would be asked out compared with seven (46.7%) women without ID.

Perceived attractiveness to others
Reasons for 'yes' responses. A third of those with and without ID who thought they would be asked on a date by someone they found attractive attributed it to their 'personality'. 'Physical attraction' was also a common reason given by participants with and without ID. It is noteworthy that only those with ID discussed 'companionship' as a reason for being asked out, either because the other person needed company or as a benefit to themselves. Participants without ID spoke about 'type' being a motivation for dating, suggesting they considered others would be attracted to a certain type of person, for which they might or might not be a good fit.
Reasons for 'no' responses. Over half of the participants without ID and over a third of those with ID spoke about 'attractiveness' as the reason they would not be asked out on a date. Participants discussed attraction in terms of social rank, referring to the image being more attractive than them and therefore 'out of their league'. Age difference was another common reason those without ID gave for not being asked on a date. One person with an ID thought the other person's lack of confidence would be the issue. In contrast, one person without an ID believed their own 'confidence issues' would stop them from approaching someone.

Offer of a date
Accepted or rejected. There were no statistically significant differences between the two groups' views that their own offer of a date to the person they found most attractive would be accepted or rejected (χ 2 (1) = 5.172, P = 0.023). However, more people with ID stated that their offer would be accepted. Twenty-one (84%) people with ID thought their offer of a date would be accepted compared with 16 (55.2%) people without ID. When breaking the groups down by gender, there were no statistically significant differences between men's (P = 0.020, two sided, Fisher's exact test) or women's (χ 2 (1) = 1.007, P = 0.316) responses. All 11 men with ID thought their offer of a date would be accepted compared with eight (57.1%) men without ID. While 10 (71.4%) women with ID thought their offer of a date would be accepted compared with eight (53.3%) women without ID.
Reasons for 'yes' responses. Half of those without ID attributed their offer of a date being accepted to 'personality'. The most common reason given by participants with ID was 'physical attraction', with 'companionship' the second most common reason. None of those without ID mentioned companionship as a reason for their offer of a date being accepted.
Reasons for 'no' responses. The most common reasons people without ID believed their offer of a date would be rejected was due to their own 'attractiveness' and 'age'. Participants with ID discussed 'availability' and mentioned practical reasons such as where they lived. Women in both groups made reference to 'traditional views' about dating and it being a man's role to ask someone out on a date.

Discussion
The findings show an association between what men and women with ID and those without ID considered attractive in romantic partners. Agreement on high and low attractiveness ratings suggested that individuals were using similar visual cues to form impressions of attractiveness. These findings fit with a wealth of literature that suggests that different factors produce powerful common stereotypes of attraction (Langlois et al. 2000). As people with ID are exposed to the same cultural norms of attractiveness expressed by society and the media, it was therefore not surprising that their views were consistent with those in general population. Little et al. (2011a) suggest that exposure to and learning about what is found attractive by others lead individuals to search for these desirable traits in prospective partners.
Although an evolutionary perspective proposes that shared views of attractiveness are a mechanism for ensuring gene survival (Little et al. 2011b), it should also be noted that findings of very similar judgements of facial attractiveness by the participants with and without ID are somewhat surprising. The particular sets of pictures used were very similar, with slight differences on characteristics thought to be key in determining attractiveness (e.g. skin tone). Hence, the participants were being asked to make very subtle judgements about the visual cues. Past research has suggested that people with ID can have significant difficulties with differentiating between more complex facial emotions such as disgust or surprise (Matheson and Jahoda 2005). These findings show that people with ID are able to make fine-grained discriminations in relation to attractiveness. With regard to self-perceived desirability as a romantic partner, more of those with ID thought they would be invited on a date. In addition, a higher proportion of those with ID also said their offers of a date would be accepted by the person they found most attractive. These findings suggest that the participants with an ID were more likely to consider themselves desirable or attractive than their nondisabled peers. Thus, despite their devalued social status, they retained a positive sense of self. This was a surprising finding because even though there has been a significant movement towards addressing prejudice and negative stereotyping faced by people with ID, it has been found that negative societal attitudes are still commonplace. When McCarthy et al. (2020) asked people with ID about their experiences of using mainstream dating agencies, they found evidence of people being 'ridiculed' and 'ignored'.
The relative optimism of the participants with an ID may have been due to how they interpreted the research questions. As a marginalised group, people with ID typically have fewer opportunities to develop informal social relationships, compared with their non-disabled peers (Scior et al. 2020). Therefore, they may have had less experience of forming intimate relationships (Pownall et al. 2012). This could have made it more difficult for them to judge what would happen in a dating scenario, particularly when posed with a hypothetical question. In addition, it could be argued that these findings relate merely to participants without ID being more influenced by social desirability bias. Their attempts to be viewed positively by the researcher may have led them to moderate their responses to try not appear overconfident or boastful.
A lack of social or inter-personal awareness may not be the only explanation for the participants' optimism about being viewed favourably by others. There is strong evidence about people with ID's sensitivity to stigma and social rejection (Jahoda et al. 2010). Promoting a favourable view of themselves, in relation to others, could have been a deliberate rejection of their stigmatised status and a way of promoting a positive sense of self (Jahoda and Markova 2004).
Social comparison was a dominant theme in participants without ID's reasons for expecting a negative outcome in the dating scenario questions. The view that attractive individuals were 'out of my league' suggested that they based partner selection on assessing their own attractiveness to others in comparison with social norms. This fits with the notion that people look for a partner who is similarly socially desirable to themselves, a consistent theme in the attraction literature. For example, evolutionary theory suggests that partners seek a mate with equivalent value (Buss and Shackelford 2008), while the 'matching hypothesis' suggests a matched socially desirable partner can offer a more successful relationship outcome (Taylor et al. 2011). It was unclear whether people with ID were making the same type of social comparison about their own desirability. Given their marginalised status within society, further research is required to explore how self-worth may influence partner selection within this population.

Limitations
The findings from this exploratory study need to be interpreted with considerable caution. Participants in both groups remarked on how the photographed faces used in the study differed from their own. The set of photographs were of individuals from Eastern Europe. Although research within the general population has indicated cross-cultural norms in attraction, the ethnicity of the images may have influenced participants' responses. A further drawback was the failure to collect data on the ethnicity of the study participants.
The set of photographed faces used within the study were originally models used as avatars for computer games. All the faces were therefore relatively good looking, and certainly, none appeared to have a visible disability. The lack of variance within the image set is also a limitation of the study and has not been taken into account when interpreting the findings. As such, an interesting area of future research would be to investigate attitudes towards people who have visible disabilities or are more markedly less attractive. However, as previously stated, the comparatively 'good looking' set of photographs used in this study meant that the participants were being asked to make subtle judgements about the relative attractiveness of the photos.
The experimental tasks and interview employed within this exploratory study were based on attraction research within the general population (Bale and Archer 2013;Wincenciak et al. 2015). These methods would benefit from further research to examine their validity when used with people who have an ID.

Further research and implications for practice
This exploratory study suggests that people with ID view themselves as desirable to others who are attractive. An important area of future research would be to explore who people with ID actually date, within their available dating scene. This should include those who have left the family home and have more autonomy around meeting and dating potential partners. It would also be interesting to explore whether partner preferences are related to actual partner choice and what happens when people's attraction to others is not reciprocated.
Although this study was inclusive of all sexual identities, the sample was predominantly heterosexual, which limited the analysis of lesbian and bisexual participants' data. In line with the growing recognition of diverse sexual identities within the ID population (Abbott and Howarth 2007), further research exploring the preferences of those with non-heterosexual sexual identities is required.
This research highlights that physical attraction is likely to play a role in people with ID's romantic partner choice, just as it does with other people. Clearly attraction goes beyond physical features and may include other characteristics like personality and what people say and do. The notion of attraction is central to the development of positive romantic and intimate relationships. For example, when is it okay to approach someone you are attracted to? What are the emotional consequences if you are rejected? There may also be the need to challenge some of the normative notions of physical attractiveness to support people with ID to retain a positive sense of themselves and their partnership choices (Löfgren-Mårtenson 2013). Thus, physical attraction should be included in sexual and relationship education. Talking more openly about attraction with peers, family or support professionals may facilitate more sensitive support with romantic relationships, alongside a consideration of the variety of factors that may influence partner choice. The participants with ID were the only ones to suggest that going out on a date would offer the opportunity for companionship.

Conclusions
This exploratory study offers an initial step towards incorporating people with ID into the attraction literature and towards advancing the evidence-base surrounding people with ID's intimate relationships. The participants with ID in this study held the same views about attractiveness as their non-disabled peers. The finding that they were more inclined to view themselves as desirable than their non-disabled peers might demonstrate more relaxed views about social comparisons and social status as determined by attractiveness. However, it remains unclear whether such views would influence how people with ID behave in everyday life.