Perceiving the facial trustworthiness: Facial age,emotional expression,and attractiveness

Abstract

People can make trustworthiness judgements based on facial characteristics. However, the previous findings regarding whether facial age influences interpersonal trust are inconsistent. Using the trust game, the current study investigated the interactions of facial age with attractiveness and emotional expression in regard to trustworthiness judgements. In Experiments 1 and 2, younger participants were asked to invest in either younger or older faces that were shown for 2,000 and 33 ms, respectively. The results showed that people trust the faces of older people more than they do younger people. There was also an interaction between facial age and attractiveness. Participants invested more money in older faces than in younger faces only when they perceived the faces to be less attractive. However, the interaction between facial age and emotional expression was slightly inconsistent in the two experiments. Participants invested more money in older faces that were shown for 2,000 ms when they perceived the happy and sad emotions, but they invested more money in older faces that were shown for 33 ms when they perceived the happy emotion. These results reveal that people make trustworthiness judgements based on multiple facial cues when they view strangers of different ages.

Keywords

Trust game trustworthiness facial age attractiveness emotional expression

Introduction

Interpersonal trust refers to the degree of trust that is present in a certain interpersonal relationship (Johnson-George & Swap, 1982), which can lead to cooperative behaviours (Mayer et al., 1995). When interacting with a stranger, people can form their first impression of trustworthiness based on the limited available information (e.g., facial appearance) and decide whether to cooperate with this person (Krueger et al., 2008; Rusman et al., 2012; Zhang et al., 2019). Trust and trustworthiness, which both matter and are closely related, can benefit both the individual and the community. First, interpersonal trust is adaptive, which facilitates prosocial behaviour (Poulin & Haase, 2015). Thus, trust has a positive effect on job satisfaction (Guinot et al., 2014), improves the quality of life (Tokuda et al., 2008), and longitudinally predicts subjective well-being (Bailey et al., 2015; Bailey & Leon, 2019; Poulin & Haase, 2015). Second, trust is a vital ingredient in the formation and maintenance of stable social relations (Beldad et al., 2010). Social exchanges involving material and nonmaterial goods are dependent on trust for continuation and completion (James, 2002). For instance, social trust can help farmers promote beneficial cooperation and ecological consciousness (Wang et al., 2019). An increased level of general trust can enhance social capital and economic growth (Guiso et al., 2004). Many previous studies have found that people can make social inferences based on the characteristics of a stranger’s face, such as gender (Carragher et al., 2018; Oosterhof & Todorov, 2008), age (Bailey et al., 2015; Bailey & Leon, 2019), emotion (Caulfield et al., 2014, 2016), and attractiveness (Kaisler & Leder, 2017; Ma et al., 2016; South Palomares & Young, 2018). People judge others based on their facial appearance in hundreds or even tens of milliseconds (Todorov et al., 2009; Willis & Todorov, 2006), and these judgements can subsequently and indirectly influence social decision making. Among these judgements, the fundamental salient type is that of facial trustworthiness (Oosterhof & Todorov, 2008). Trustworthiness judgements based on facial photos are found cross-culturally (Jones et al., 2021). Previous studies have found that facial trustworthiness judgements can affect trust behaviour in experiments (Qi et al., 2018; Rezlescu et al., 2012; Stirrat & Perrett, 2010; van’t Wout & Sanfey, 2008).

Previous studies have examined which factors can influence trustworthiness judgements, including two types of responses, namely self-reported and behavioural trust (Bailey & Leon, 2019; Zhang et al., 2019). In the studies that employ self-reported trust response, participants need to rate the facial trustworthiness (Bailey et al., 2015; Wernick & Manaster, 1984). In the studies that use the behavioural trust response, different paradigms are used to prompt people to make trustworthiness judgements, including the trust game (TG, Krueger et al., 2008; Qi et al., 2018; Stirrat & Perrett, 2010) and the ultimatum game (UG, Wu et al., 2018). In the trust game, the trustor (or investor) determines an amount of money to invest in a trustee (or partner), and this money is multiplied by some factor, such as three or four (Bailey et al., 2015; Qi et al., 2018). Then, the trustee needs to return some of the money to the trustor, and the amounts of investment from both trustor and trustee serve as quantitative indicators of trust behaviour (Rilling & Sanfey, 2011). In the ultimatum game, participants can serve as either the proposer or responder. The proposer needs to divide a fixed monetary reward with the responder, and the responder needs to decide whether he or she would like to accept or reject this offer. In general, participant serves as the responder in the ultimatum game or the trustor in the trust game. When participants play trust games with different facial photos, some researchers make the faces trustworthy or not by controlling changeable facial features, such as hair style, eye gaze, facial width ratio, or emotion (Stirrat & Perrett, 2010; van’t Wout & Sanfey, 2008). In other studies, facial trustworthiness is a subjective rating based on the holistic processing of the facial photo (e.g., Qi et al., 2018; Wilson & Rule, 2015, 2016). The more trustworthy a face appears to be, the more money people are willing to invest in it (Qi et al., 2018; Rezlescu et al., 2012). Some studies have found that facial trustworthiness judgements can even affect trust behaviours in real life. Researchers have found that criminals with lower levels of trustworthiness are more likely to be sentenced to death (Wilson & Rule, 2015, 2016). Therefore, facial trustworthiness has an important influence on people’s trust behaviour.

Many previous studies have focused on the influence of static morphological features, such as emotional expression and attractiveness, on trustworthiness judgements (Slepian & Carr, 2019; South Palomares & Young, 2018). Based on the hypothesis of emotion generalisation, different types of facial trustworthiness correspond to different emotional valences and thus affect people’s cooperative behaviours (Oosterhof & Todorov, 2008). For example, researchers have found that people give more trustworthy ratings to faces in which facial emotions are more positive and vice versa (Calvo et al., 2019; Todorov, 2008). Rezlescu et al. (2012) also found that individuals tend to attach positive emotions more to trustworthy faces compared with untrustworthy faces. Previous studies on facial expressions have found that people generally consider faces with positive emotions to be trustworthy and faces with negative emotions to be untrustworthy (Chen & Bargh, 1999; Oosterhof & Todorov, 2008). Individuals perceive people with happy faces as having traits associated with high levels of affiliation and dominance, and perceive people with sad faces as having traits associated with a moderate level of affiliation and a low level of dominance (Zebrowitz & Montepare, 2008). We expect people to reduce their actual trust behaviour when faced with negative expressions. Compared with emotional expression, researchers have also found a positive correlation between attractiveness and trustworthy judgements of the trustee (Ma et al., 2016; Oosterhof & Todorov, 2008). People associate high levels of attractiveness with high levels of trustworthiness (Todorov et al., 2009); another study found that people perceive an attractive face as being more positive and social, and vice versa (Zebrowitz & Montepare, 2008).

In addition to common facial characteristics, facial age can also moderate people’s evaluations on a stranger and thus affect the subsequent trust behaviours (Bailey et al., 2016; Bailey & Leon, 2019; Kiiski et al., 2016) for both trustors and trustees (Bailey et al., 2015; Greiner & Zednik, 2019). However, the previous empirical findings regarding the impressions of older faces make distinct predictions regarding interpersonal trust. On one hand, some studies on the negative stereotypes of age have indicated that individuals tend to evaluate older people more negatively than they do younger people (e.g., Ebner, 2008; Ward, 1988). Compared with younger adults, older adults have been perceived as lower levels of competence (Kite & Johnson, 1988), lower levels of attractiveness (Ebner, 2008; Wernick & Manaster, 1984), lower levels of dominance (Batres et al., 2015), and increased levels of sad and angry emotions (Craig & Lipp, 2018; Sacco & Hugenberg, 2009). Therefore, in this situation, young adults cannot cooperate with older adults, who are identified as lacking competence and thus may fail to return money. Moreover, the attractiveness halo effect (Ma et al., 2016; Zebrowitz & Montepare, 2008) and the hypothesis of emotion generalisation (Oosterhof & Todorov, 2008) also predict that older individuals would be judged as being less trustworthy. Thus, young adults would invest less money in older adults. On the other hand, although there is a negative stereotype about older adults, people can still trust older people more because of the positive social characters present in older individuals. A previous study found that young adults commonly hold a positive stereotype about older adults, which might enhance young adults’ feeling of social closeness with older adults (Slessor et al., 2014). In addition to self-reported results, a similar result has been found in regard to trust behaviour. With increasing of age, the individuals serving as trustees show higher levels of trustworthiness to the trustors (Greiner & Zednik, 2019). The individuals rate the older faces as being more trustworthy (Bailey et al., 2015). This inconsistent conclusion regarding facial trustworthiness judgements could also be due to cultural differences (Jones et al., 2021). Boduroglu et al. (2006) employed an open-ended approach to collect descriptors of youth and old age from American and Chinese groups and then compared the proportions of descriptors. The authors found that people from the Chinese culture evaluate older individuals more positively, and they suggested that individuals of Chinese culture are more respectful and considerate of older individuals than younger individuals. Thus, in the present case, when playing trust game, it is likely that participants with a Chinese culture background will invest more in older individuals. However, distinguishing the relationships between facial age, attractiveness, and emotional expression is one of the main prerequisites for clarifying the effect of facial age on trust behaviour.

In line with previous evidence, most experiments that employ the age-related trust game task do not take facial attractiveness and emotional expression into account; in this situation, researchers might fail to clearly determine the effect of facial age. Therefore, the present study was designed to examine the impact of facial age on trust behaviour by considering facial attractiveness and emotional expression. We used multiple facial cues (e.g., facial age, attractiveness, and emotional expression) to further explore whether trust behaviour based on facial age is moderated by different facial cues when people make judgements on faces.

The study consisted of two experiments. In Experiment 1, we investigated whether facial age could influence trustworthiness judgements on younger and older faces and whether the results would be consistent among different levels of facial attractiveness and emotional expressions. We proposed that people would invest more money in happy faces and more attractive faces. We expected that there would be facial age bias towards trustworthiness and this bias would be consistent among different levels of attractiveness and emotional expression. Notably, Craig and Lipp (2018) recruited young adults to categorise emotion or facial age when seeing younger and older faces with different emotional levels. They found that young adults process younger faces more slowly than they do older faces. This prolonged processing time might allow participants to have enough time to over-generalise neutral faces. As Hester (2019) found, people over-generalise the emotion of neutral faces and perceive negative or positive emotion from neutral facial photos. Sometimes, young adults perceive the younger happy emotional faces to be fake smiles (Okubo et al., 2012). Considering the differences in the processing time of younger and older faces, Experiment 2 was designed to explore whether the impacts of facial features (e.g., facial age and emotional expression) on facial trustworthiness would still exist at a shortened processing time. The faces in Experiment 2 were presented for 33 ms, which is the threshold of making trustworthiness judgements based on facial appearances (Todorov et al., 2009).

Experiment 1

Materials and methods

Participants

This study was approved by the internal review board of the Institute of Psychology, Chinese Academy of Sciences. In Experiments 1 and 2, we conducted a priori power analysis with an estimate of the effect size was 0.25, and 18 participants were needed with a statistical testing power of 0.95 (α = .05). In Experiment 1, 52 undergraduate and graduate students (30 females, age = 21.42 ± 2.56) were recruited and offered monetary compensation for their time. All participants were right-handed and reported to have normal colour vision.

Stimuli

Previous studies used computer-generated faces to test the facial trustworthiness (Bailey & Leon, 2019; Ma et al., 2016), but there were more studies using true faces to test (Bailey et al., 2015; Craig & Lipp, 2018; Kaisler & Leder, 2017; Lindeberg et al., 2018; Qi et al., 2018). Given the lack of fidelity of computer-generated faces, a total of 96 frontal photographs of unfamiliar Chinese faces were used in the current study, including 48 older faces and 48 younger faces with three different emotional expressions (happy/neutral/sad). Half of the face photos were males, and all faces of the women were not wearing makeup which might have the effects on age perceptions (Russell et al., 2019). The images of younger faces were selected from Chinese Affective Picture System (CAPS, Bai et al., 2005). The age of young people in the photos ranged from 17 to 23 years, and the age of old people ranged from 60 to 70 years. To our best knowledge, there was no such database that included enough old Chinese faces. Thus, referring to the procedure of the previous study (Bai et al., 2005), we asked older people to make happy, neutral, and sad emotional expressions and then took the pictures to form the materials of the older faces. All the face photos were converted to grayscale and placed on a black background 260 × 300 pixels in size according to the direction of the CAPS.

Considering rating both attractiveness and trustworthiness might introduce common method bias, the emotional expressions and attractiveness of each face were rated in the pilot studies (Lindeberg et al., 2018). Fifty-six participants rated emotions with a 9-point Likert-type scale, with higher scores indicating more positive valence. The mean score for young happy faces was 5.92 (SD = 0.15), neutral faces was 4.89 (SD = 0.15), and sad faces was 3.06 (SD = 0.15); follow-up pairwise comparison with Bonferroni correction demonstrated that young happy faces were rated as more positive valence, then the young neutral faces, and finally the young sad faces (all ps < .001); for old happy faces was 7.34 (SD = 0.15), neutral faces was 4.98 (SD = 0.15), and sad faces was 3.61 (SD = 0.15), follow-up pairwise comparison with Bonferroni correction demonstrated that old happy faces were rated as the most positive, followed by the young neutral faces, and the young sad faces were the most negative ones (all ps < .001). Another 44 participants rated the facial attractiveness with a 7-point Likert-type scale, with higher scores indicating more attractive. Each group of face photos was divided into two groups and labelled as high- or low-attractiveness according to the median ratings. The mean rating of attractiveness for young high-attractiveness faces was 3.36 (SD = 0.06) and low-attractiveness faces was 2.77 (SD = 0.06, p < .001); for old high-attractiveness faces was 2.95 (SD = 0.06) and low-attractiveness faces was 2.49 (SD = 0.06, p < .001).

Moreover, we analysed the inter-rater agreements to explore the validity of rating for emotional expression and attractiveness by these participants. We calculated the intraclass correlation coefficient (ICC), which we chose the two-way random model with absolute agreement type. Finally, we got the ICC = 0.449 (p < .001) in single measures of emotional expression rating and ICC = 0.117 (p < .001) in single measures of attractiveness rating, which indicates acceptable repeatability and it is suitable to use these faces sorted as different levels of emotional valence and attractiveness in Experiments 1 and 2.

Procedure

The stimuli presentation and data collection were controlled by E-Prime. As shown in Figure 1, the sequence of a typical trial in Experiment 1 was similar as the procedure in Berg et al. (1995), Bailey et al. (2015), and Qi et al. (2018). Participants were told to play trust games with different people online. In each trial, they would have 10 yuan to invest on their partner in this trial. In a typical trial, the fixation point “+” was first presented in the centre of the screen for 1,000–1,500 ms and then a face photo was displayed in the centre of the screen for 2,000 ms. Participants were asked to decide the amount of money (from 1 to 10 yuan) they would like to invest in this partner in this trial. After that, their partner would receive quadruple the amount of money invested by participant and distribute it either fairly (5:5) or unfairly (3:1 or the partner keeps all the money). Participants would see the distribution results at the end of each trial. Before the real experiment, participants were encouraged to use any strategy they wanted to maximise their amount of points. Participants were informed that the monetary bonus would be determined by the actual result of a randomly selected trial.

Figure 1.

The sequence and timing of one typical trial in Experiment 1.

Each face was presented twice, and a practice block with eight practice trials was completed before the formal experiment to familiarise participants with the task. The experiment contained 4 blocks with a total of 192 trials. Each block contained only one type of the age, the block sequence was counterbalanced between subjects. At the end of the experiment, one trial would be selected randomly and the actual result of the trial would be the bonus money that participant would receive.

Design

There were three independent variables of interest: facial age (young vs. old), emotional expression (happy, neutral, and sad), and facial attractiveness (high vs. low). And the dependent variable was the amount of money to invest in trustee. We proposed that people invest more money in older faces, and this facial age bias is consistent among different levels of attractiveness and emotional expression. And we employed repeated analyses of variance (ANOVAs) based on subjects to test our hypotheses.

Results

The mean money distributed to partners was shown in Figure 2. A 2 (facial attractiveness: high vs. low) × 3 (emotional expression: happy vs. neutral vs. sad) × 2 (facial age: young vs. old) repeated ANOVAs revealed a main effect of facial age, F (1, 51) = 17.63, p < .001, $η_{p}^{2} = . 257$ . Compared with younger faces (4.51 ± 0.19), people invested more money in the older faces (4.98 ± 0.21). There was a main effect of emotional expression, F (2, 102) = 23.67, p < .001, $η_{p}^{2} = . 317$ . Compared with sad faces (4.31 ± 0.19), people invested more money in happy faces (5.02 ± 0.22, p < .001) and the neutral faces (4.91 ± 0.19, p < .001). And the investment in neutral and happy faces were similar (p = .902). Regarding facial attractiveness, F (1, 51) = 77.99, p < .001, $η_{p}^{2} = . 605$ , compared with low-attractive faces (4.50 ± 0.19), people invested more money in high-attractive faces (4.99 ± 0.20).

Figure 2.

Mean investments on different types of faces in Experiment 1: (a) the mean money to the young faces and (b) the mean money to the old faces. Error bars indicate the SEMs.

The facial attractiveness × facial age interaction was significant, F (1, 51) = 90.08, p < .001, $η_{p}^{2} = . 639$ . Compared with low-attractive younger faces (4.10 ± 0.19), participants invested more money in low-attractive older faces (4.89 ± 0.21, p < .001), but investments were similar in high-attractive young (4.91 ± 0.20) and older faces (5.07 ± 0.22, p = .145), as shown in Figure 3.

Figure 3.

The interaction of facial attractiveness and facial age in Experiment 1. Error bars indicate the SEMs.

The facial age × emotional expression interaction was also significant, F (2, 102) = 33.92, p < .001, $η_{p}^{2} = . 399$ . Compared with younger faces (4.46 ± 0.22; 4.11 ± 0.19), participants invested more money in older faces (5.57 ± 0.24; 4.51 ± 0.22) with happy emotion (p < .001) and sad emotion (p = .010). But the investments were similar in faces with neutral emotion when seeing young (4.96 ± 0.20) and older faces (4.85 ± 0.21, p = .399), as shown in Figure 4.

Figure 4.

The interaction of emotional expression and facial age in Experiment 1. Error bars indicate the SEMs.

The facial attractiveness × emotional expression interaction was significant, F (2, 102) = 9.91, p < .001, $η_{p}^{2} = . 163$ . In three levels of emotion expression, people always invested more in high-attractiveness faces compared with low-attractiveness faces (ps < .001). The three-way interaction was not significant, F (2, 102) = 1.09, p = .339, $η_{p}^{2} = . 021$ .

Experiment 2

Experiment 1 showed that people are willing to invest more money in older faces, especially at levels of low attractiveness, happy, and sad, which indicates that adults trust older people more than younger people. Thus, our hypothesis that people trust older faces more than they trust younger faces has been confirmed. Moreover, the effect of older faces on trust behaviour is moderated by facial attractiveness and emotional expressions.

The facial age superiority effect has only been found at levels of low-attractiveness level and non-neutral emotional conditions. The weights of facial features are not consistent when making trustworthiness judgements, which might cause the phenomenon that there is no facial age bias at high-attractiveness level. In the age classification task, Craig and Lipp (2018) found that young adults process faces of the same age more carefully and make slower responses, which is called own-age bias. In view of this situation, we recruited young participants and used younger and older faces as the materials, the likely reason why the amount of money they invested in younger and older neutral faces was not different was the stimuli presentation. Hester (2019) found that people can process neutral faces over-generalisation, and perceive negative or positive emotion in neutral faces. Therefore, we hypothesised that participants in Experiment 1 had enough time to process young neutral faces over-generalisation and then perceive positive emotion in young neutral faces. Thus, Experiment 2 further explored whether presentation time influenced the age superiority effect on trust behaviours.

Considering that people distinguish between trustworthy and untrustworthy faces in just 33 ms (Freeman et al., 2014), Experiment 2 further investigated whether the interaction of facial age with facial attractiveness and emotional expressions could still exist if the duration of the faces viewing time was reduced to 33 ms.

Materials and methods

Participants

In this experiment, another 52 undergraduate and graduate students (30 females, age = 21.25 ± 2.51) were recruited and offered monetary compensation for their time. All of them were right-handed and had no colour blindness or colour weakness.

Stimulus materials

The face photos used in Experiment 2 were identical to Experiment 1.

Procedure

The procedure of Experiment 2 was identical to Experiment 1 except that the duration of the faces was changed to 33 ms.

Results

The mean money distributed to partners was shown in Figure 5. A 2 (facial attractiveness: high vs. low) × 3 (emotional expression: happy vs. neutral vs. sad) × 2 (facial age: young vs. old) repeated ANOVAs revealed a main effect of facial age, F (1, 51) = 1.18, p = .283, $η_{p}^{2} = . 023$ . There was a main effect of emotional expression, F (2, 102) = 21.78, p < .001, $η_{p}^{2} = . 299$ . Compared with sad faces (5.28 ± 0.28), people invested more money in happy faces (5.98 ± 0.27, p < .001) and neutral faces (5.74 ± 0.26, p < .001). And in this experiment, the investment amounts in neutral faces and happy faces were different (p = .031). The effect of facial attractiveness was also significant, F (1, 51) = 45.49, p < .001, $η_{p}^{2} = . 471$ , compared with low-attractive faces (5.49 ± 0.27), people invested more money in high-attractive faces (5.84 ± 0.26).

Figure 5.

Mean investments to different types of faces in Experiment 2: (a) the mean money to the young faces and (b) the mean money to the old faces. Error bars indicate the SEMs.

The facial attractiveness × facial age interaction was significant, F (1, 51) = 27.91, p < .001, $η_{p}^{2} = . 354$ . Compared with low-attractive younger faces (5.27 ± 0.29), participants invested more money in low-attractive older faces (5.71 ± 0.29, p = .035), but investments were similar in the high-attractive younger faces (5.84 ± 0.26) and older faces (5.84 ± 0.29, p = .992), as shown in Figure 6.

Figure 6.

The interaction of facial attractiveness and facial age in Experiment 2. Error bars indicate the SEMs.

The interaction of facial age × emotional expression was significant, F (2, 102) = 13.99, p < .001, $η_{p}^{2} = . 215$ . Compared with younger faces (5.75 ± 0.29), participants invested more money in older faces (6.20 ± 0.30) when perceiving happy emotion (p = .037). But the investments were similar in young (5.82 ± 0.26; 5.10 ± 0.30) and older faces (5.66 ± 0.29; 5.47 ± 0.31) when perceiving neutral (p = .428) and sad emotion (p = .105), as shown in Figure 7.

Figure 7.

The interaction of emotional expression and facial age in Experiment 1. Error bars indicate the SEMs.

The interaction of facial attractiveness × emotional expression was also significant, F (2, 102) = 8.04, p = .001, $η_{p}^{2} = . 136$ . And the simple effect test showed that in three levels of emotion expression, people always invested more in high-attractiveness faces than in low-attractiveness faces (ps < .01). The three-way interaction, F (2, 102) = 2.71, p = .072, $η_{p}^{2} = . 050$ , was not significant.

Even though the result of interactions between facial age and attractiveness was similar in Experiments 1 and 2, participants just only invested more in older faces with happy emotion in Experiment 2. So, we calculated the effect of facial age on the differences of investment amount evoked by emotional expressions and analysed whether this effect would be modified by the duration of face. The magnitudes of investments were calculated by using the difference in mean investments in happy and sad faces divided by their sum, $\frac{{Money}_{happy} - {Money}_{sad}}{{Money}_{happy} + {Money}_{sad}}$ . A 2 (facial age: young vs. old) × 2 (duration time: 33 ms vs. 2,000 ms) ANOVAs revealed a main effect of facial age, F (1, 102) = 10.72, p = .001, $η_{p}^{2} = . 095$ . More importantly, there was an interaction of facial age × duration time, F (1, 102) = 6.07, p = .015, $η_{p}^{2} = . 056$ . Only in 2,000 ms, the difference of investments between the happy and sad levels in older faces (0.11 ± 0.02) would be larger than younger faces (0.04 ± 0.02, p < .001). These results indicated that the facial cue of age would hardly influence people’s trustworthiness judgements under a relatively shorter presentation time.

Discussion

In this study, we found an effect of facial age and the interaction of facial age with emotional expressions and attractiveness on trustworthiness judgements. People’s trust behaviours increase with facial age and are moderated by different facial cues when they judge faces from different facial age groups. According to the trust game used in previous studies about facial perception (Qi et al., 2018; Stirrat & Perrett, 2010), the present study broadened the age of the utilised face materials to older faces and explored the influence of facial age, attractiveness, and emotional expressions on trust behaviours. This study found that compared with younger faces, people trust older faces more under conditions of both abundant (2,000 ms) and limited (33 ms) time for face processing. Specifically, compared with younger faces, people trust older faces more when they perceive the happy emotion under both 2,000 and 33 ms conditions; however, people trust older faces (vs. younger faces) more when they perceive the sad emotion under the 2,000 ms condition. Moreover, people trust older faces (vs. younger faces) more under the low-attractiveness condition.

First, different from the expectation based on negative stereotypes of age (Ebner, 2008), the current study was consistent with Bailey et al. (2015) and found that young people trust older faces more than younger faces in the trust game. A previous study found that with the increase of age, people trust others more and want to let others believe themselves (Greiner & Zednik, 2019), and people may refer to this preconceived notion to judge trustworthiness. The stereotypes of older adults are that older individuals have high levels of warmth and low levels of competence (Fiske et al., 2002); thus, people may invest more money in older individual because they are less likely to betray them in a trust game. Even though we used the trust game in this study, trust behaviours were relevant not only to trustworthiness but also to the reciprocity concerns during the trust game (Bailey et al., 2015). At the same time, this outcome may be due to the influence of Chinese culture (Boduroglu et al., 2006). Respecting older individuals has long been a national tradition, and young people would be ashamed of distrusting older people. Therefore, the young participants in the current study invested more money in older faces.

Second, the current research found an interaction between facial age and attractiveness. First, young adults were found to trust older faces more at the low-attractiveness level, but there was no difference at the high-attractiveness level. According to previous studies, older faces are associated with low attractiveness, according to the asymmetric facial features (Sacco & Hugenberg, 2009; Wernick & Manaster, 1984). The association of older faces and high attractiveness might seem unexpected and thus lower the expectation of reciprocity. Another possibility is that people do not take facial age into account when seeing high-attractiveness faces. For instance, Calvo et al. (2018) proposed that the use of attractiveness could minimise the effort of making trustworthiness judgements when dealing with information complexity or uncertainty. As one of the critical facial cues, facial attractiveness might lack usefulness in trustworthiness judgements regarding older faces. Future studies are encouraged to explore the effect of different facial cues on social evaluations in different age groups.

Moreover, people place more trust in older faces with the non-neutral emotions, especially happy emotion. In Experiment 1, we found that people invested more in older faces (vs. younger faces) at both happy and sad levels. However, in Experiment 2, people only invested more in older faces at happy level. Craig and Lipp (2018) found that the facial age cue influenced emotional perception (happy and angry faces) but the emotional cue failed to influence age categorisation. Even though it is different from emotional perception, the facial age cue amplifies the effect of non-neutral emotions on trustworthiness judgements in trust games. Considering that trustworthiness judgements rely on emotional cue (Todorov et al., 2009; Zebrowitz & Montepare, 2008), the interaction between facial age and emotional expressions that we found supported the idea that facial age does influence trustworthiness judgements based on emotional expressions.

Specifically, people process faces of the same age more carefully and slowly (Anastasi & Rhodes, 2006; Craig & Lipp, 2018). According to the cross-experimental result, only in Experiment 1 (faces shown for 2,000 ms), would the difference in investments between the happy and sad levels for older faces be larger than that for younger faces; the results indicated that people might perceive younger happy face as a fake smile (Okubo et al., 2012), which reduces the perceived trustworthiness of younger happy faces. However, in Experiment 2, people failed to make over-generalisation in a limited time. By shorting the duration of the face viewing to 33 ms, there was still an interaction between facial age and emotional expressions. From the results, people invested more in older faces (vs. younger faces) only at happy level, which means that people could still integrate facial age cues from happy faces even at 33 ms when engaged in the trustworthiness judgements. A previous study indicated that happy faces are rated as older than other emotions due to smile-associated wrinkles (Ganel, 2015). They could not over-interpret the young happy faces because the duration of the facial photos was shortened to 33 ms.

As mentioned above, in addition to the impact of facial age, many studies have focused on the impact of participants’ age on trustworthiness judgements (Bell et al., 2013; Suzuki et al., 2018). For example, previous studies have presented both younger and older faces to participants from different age groups (Bailey et al., 2015; Li & Fung, 2013), and they have found the own-age bias (Anastasi & Rhodes, 2006). Specifically, people of different ages are more likely to identify and recognise the faces that are similar to their own age in comparison to people of other ages. However, these studies failed to control the facial features of the stimuli, such as emotional expression and attractiveness. In our current study, we controlled the facial features, and the own-age bias disappeared for young adults. One possibility is that the own-age bias can only be found in the face recognition task. When people are making trustworthiness judgements, the own-age bias does not work. Future studies are needed to further explore whether the own-age bias emerges when the older people make trust judgements based on facial photos. Moreover, the time needed to perceive and interpret a specific feature of the face remains unclear. Future studies using Electroencephalogram methods are encouraged to use a certain amount of time for processing every specific facial feature (e.g., age, attractiveness, emotion) when doing trustworthiness judgements.

Conclusion

The present study broadened the range of facial age in the trust game and explored how the faces of younger and older people with different levels of emotional expressions and attractiveness affect people’s trust behaviours. People trust older faces more than younger faces. More importantly, the current research found that people employ different facial cues when making trust decisions based on faces from different age groups. Specifically, facial age is a valid cue used only in the judgements based on the low-attractiveness and on non-neutral emotional levels. Considering that the current study recruited only undergraduate and graduate students as participants, future studies could explore the difference in age of trustors to further understand the relationship between the multiple facial features of trustees and the factors of the trustors.

Footnotes

Author contributions

Y.L. and Z.C. made equal contributions to this manuscript. Z.C. performed the experiments and wrote the paper. Y.L. and Y.Q. designed the experiments, analysed the data, and revised this manuscript. X.L. revised the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the fund for building world-class universities (disciplines) of Renmin University of China (RUCPSY0007).

ORCID iDs

Ziwei Chen

Yue Qi

References

Anastasi

J. S.

Rhodes

M. G.

(2006). Evidence for an own-age bias in face recognition. North American Journal of Psychology, 8(2), 237–252.

Bai

Huang

Y. X.

Luo

Y. J.

(2005). The development of native Chinese affective picture system: A pretest in 46 college students. Chinese Mental Health Journal, 19, 719–722. https://doi.org/10.3321/J.ISSN:1000-6729.2005.11.001

Bailey

P. E.

Leon

(2019). A systematic review and meta-analysis of age-related differences in trust. Psychology and Aging, 34(5), 674–685. https://doi.org/10.31219/osf.io/we7vh

Bailey

P. E.

Slessor

Rieger

Rendell

P. G.

Moustafa

A. A.

Ruffman

(2015). Trust and trustworthiness in young and older adults. Psychology and Aging, 30(4), 977–986. https://doi.org/10.1037/a0039736

Bailey

P. E.

Szcap

McLennan

S. N.

Slessor

Ruffman

Rendell

P. G.

(2016). Age-related similarities and differences in first impressions of trustworthiness. Cognition and Emotion, 30(5), 1017–1026. https://doi.org/10.1080/02699931.2015.1039493

Batres

Perrett

D. I.

(2015). Influence of perceived height, masculinity, and age on each other and on perceptions of dominance in male faces. Perception, 44(11), 1293–1325. https://doi.org/10.1177/0301006615596898

Beldad

de Jong

Steehouder

(2010). How shall I trust the faceless and the intangible? A literature review on the antecedents of online trust. Computers in Human Behavior, 26(5), 857–869. https://doi.org/10.1016/j.chb.2010.03.013

Bell

Giang

Mund

Buchner

(2013). Memory for reputational trait information: Is social-emotional information processing less flexible in old age? Psychology and Aging, 28(4), 984–995. https://doi.org/10.1037/a0034266

Berg

Dickhaut

McCabe

(1995). Trust, reciprocity and social history. Games and Economic Behavior, 10, 122–142. https://doi.org/10.1006/game.1995.1027

10.

Boduroglu

Yoon

Luo

Park

D. C.

(2006). Age-related stereotypes: A comparison of American and Chinese cultures. The Gerontologist, 52(5), 324–333. https://doi.org/10.1159/000094614

11.

Calvo

M. G.

Gutiérrez-García

Beltrán

(2018). Neural time course and brain sources of facial attractiveness vs. trustworthiness judgment. Cognitive Affective & Behavioral Neuroscience, 18, 1233–1247. https://doi.org/10.3758/s13415-018-0634-0

12.

Calvo

M. G.

Krumhuber

E. G.

Fernandez-Martin

(2019). Visual attention mechanisms in happiness versus trustworthiness processing of facial expressions. Quarterly Journal of Experimental Psychology, 73(12), 2328–2347. https://doi.org/10.1177/1747021820957575

13.

Carragher

D. J.

Thomas

N. A.

Nicholls

M. E. R.

(2018). Is trustworthiness lateralized in the face? Evidence from a trust game. Laterality, 23(1), 20–38. https://doi.org/10.1080/1357650X.2017.1298120

14.

Caulfield

Ewing

Bank

Rhodes

(2016). Judging trustworthiness from faces: Emotion cues modulate trustworthiness judgments in young children. British Journal of Psychology, 107(3), 503–518. https://doi.org/10.1111/bjop.12156

15.

Caulfield

Ewing

Burton

Avard

Rhodes

(2014). Facial trustworthiness judgments in children with ASD are modulated by happy and angry emotional cues. PLOS ONE, 9(5), Article e97644. https://doi.org/10.1371/journal.pone.0097644

16.

Chen

Bargh

J. A.

(1999). Consequences of automatic evaluation: Immediate behavioral predispositions to approach or avoid the stimulus. Personality and Social Psychology Bulletin, 25(2), 215–224. https://doi.org/10.1177/0146167299025002007

17.

Craig

B. M.

Lipp

O. V.

(2018). Facial age cues and emotional expression interact asymmetrically: Age cues moderate emotion categorisation. Cognition and Emotion, 32, 350–362. https://doi.org/10.1080/02699931.2017.1310087

18.

Ebner

N. C.

(2008). Age of face matters: Age-group differences in ratings of young and old faces. Behavior Research Methods, 40(1), 130–136. https://doi.org/10.3758/BRM.40.1.130

19.

Fiske

S. T.

Cuddy

A. J. C.

Glick

(2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social, 82(6), 878–902. https://doi.org/10.1037//0022-3514.82.6.878

20.

Freeman

J. B.

Stolier

R. M.

Ingbretsen

Z. A.

Hehman

E. A.

(2014). Amygdala responsivity to high-level social information from unseen faces. The Journal of Neuroscience, 34(32), 10573–10581. https://doi.org/10.1523/JNEUROSCI.5063-13.2014

21.

Ganel

(2015). Smiling makes you look older. Psychonomic Bulletin & Review, 22(6), 1671–1677. https://doi.org/10.3758/s13423-015-0822-7

22.

Greiner

Zednik

(2019). Trust and age: An experiment with current and former students. Economics Letters, 181(4), 37–39. https://doi.org/10.1016/j.econlet.2019.04.004

23.

Guinot

Chiva

Roca-Puig

(2014). Interpersonal trust, stress and satisfaction at work: An empirical study. Personal Review, 43, 96–115. https://doi.org/10.1108/PR-02-2012-0043

24.

Guiso

Sapienza

Zingales

(2004). The role of social capital in financial development. The American Economic Review, 94, 526–556. https://doi.org/10.1257/0002828041464498

25.

Hester

(2019). Perceived negative emotion in neutral faces: Gender-dependent effects on attractiveness and threat. Emotion, 19(8), 1490–1494. https://doi.org/10.1037/emo0000525

26.

James

H. S.

(2002). The trust paradox: A survey of economic inquiries into the nature of trust and trustworthiness. Journal of Economic Behavior & Organization, 47, 291–307. https://doi.org/10.1016/S0167-2681(01)00214-1

27.

Johnson-George

Swap

W. C.

(1982). Measure of specific interpersonal trust: Construction and validation of a scale to assess trust in a specific other. Journal of Personality and Social Psychology, 43(6), 1306–1317. https://doi.org/10.1037/0022-3514.43.6.1306

28.

Jones

B. C.

DeBruine

L. M.

Flake

J. K.

Liuzza

M. T.

Antfolk

Arinze

N. C.

. . . Coles

N. A.

(2021). To which world regions does the valence–dominance model of social perception apply? Nature Human Behaviour, 5(1), 159–169. https://doi.org/10.1038/s41562-020-01007-2

29.

Kaisler

R. E.

Leder

(2017). Combined effects of gaze and orientation of faces on person judgments in social situations. Frontiers in Psychology, 8, Article 259. https://doi.org/10.3389/fpsyg.2017.00259

30.

Kiiski

H. S. M.

Cullen

Clavin

S. L.

Newell

F. N.

(2016). Perceptual and social attributes underlining age-related preferences for faces. Frontiers in Human Neuroscience, 10, Article 437. https://doi.org/10.1037/0022-3514.43.6.1306

31.

Kite

M. E.

Johnson

B. T.

(1988). Attitudes toward older and younger adults: A meta-analysis. Psychology and Aging, 3(3), 233–244. https://doi.org/10.1037/0882-7974.3.3.233

32.

Krueger

Mccabe

Moll

Kriegeskorte

Zahn

Strenziok

. . . Grafman

(2008). Neural correlates of trust. Proceedings of the National Academy of Sciences of the United States of America, 104(50), 20084–20089. https://doi.org/10.1073/pnas.0710103104

33.

Fung

H. H.

(2013). Age differences in trust: An investigation across 38 countries. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 68, 347–355. https://doi.org/10.1093/geronb/gbs072

34.

Lindeberg

Craig

B. M.

Lipp

O. V.

(2018). You look pretty happy: Attractiveness moderates emotion perception. Emotion, 19(6), 1070–1080. https://doi.org/10.1037/emo0000513

35.

Luo

(2016). Children’s facial trustworthiness judgments: Agreement and relationship with facial attractiveness. Frontiers in Psychology, 7, Article 499. https://doi.org/10.3389/fpsyg.2016.00499

36.

Mayer

R. C.

Davis

J. H.

Schoorman

F. D.

(1995). An integrative model of organisational trust. Academy of Management Review, 20(3), 709–734. https://doi.org/10.2307/258792

37.

Okubo

Kobayashi

Ishikawa

(2012). A fake smile thwarts cheater detection. Journal of Nonverbal Behavior, 36(3), 217–225. https://doi.org/10.1007/s10919-012-0134-9

38.

Oosterhof

N. N.

Todorov

(2008). The functional basis of face evaluation. Proceedings of the National Academy of Sciences of the United States of America, 105(32), 11087–11092. https://doi.org/10.1073/pnas.0805664105

39.

Poulin

M. J.

Haase

C. M.

(2015). Growing to trust: Evidence that trust increases and sustains well-being across the life span. Social Psychological and Personality Science, 6, 614–621. https://doi.org/10.1177/1948550615574301

40.

(2018). Are rich people perceived as more trustworthy? Perceived socioeconomic status modulates judgments of trustworthiness and trust behavior based on facial appearance. Frontiers in Psychology, 9, Article 512. https://doi.org/10.3389/fpsyg.2018.00512

41.

Rezlescu

Duchaine

Olivola

C. Y.

Chater

(2012). Unfakeable facial configurations affect strategic choices in trust games with or without information about past behavior. PLOS ONE, 7(3), Article e34293. https://doi.org/10.1371/journal.pone.0034293

42.

Rilling

J. K.

Sanfey

A. G.

(2011). The neuroscience of social decision-making. Annual Review of Psychology, 62, 23–48. https://doi.org/10.1146/annurev.psych.121208.131647

43.

Rusman

Bruggen

J. V.

Sloep

Valcke

Koper

(2012). Can I trust you? Personal profile for a first impression of trustworthiness in virtual project teams. International Journal of Information Technology Project Management, 3(1), 15–35. https://doi.org/10.4018/jitpm.2012010102

44.

Russell

Batres

Courreges

Kaminski

Soppelsa

Morizot

Porcheron

(2019). Differential effects of makeup on perceived age. British Journal of Psychology, 110, 87–100. https://doi.org/10.1111/bjop.12337

45.

Sacco

D. F.

Hugenberg

(2009). The look of fear and anger: Facial maturity modulates recognition of fearful and angry expressions. Emotion, 9(1), 39–49. https://doi.org/10.1037/a0014081

46.

Slepian

M. L.

Carr

E. W.

(2019). Facial expressions of authenticity: Emotion variability increases judgments of trustworthiness and leadership. Cognition, 183, 82–98. https://doi.org/10.1016/j.cognition.2018.10.009

47.

Slessor

Phillips

L. H.

Ruffman

Bailey

P. E.

Insch

(2014). Exploring own-age biases in deception detection. Cognition and Emotion, 28, 493–506. https://doi.org/10.1080/02699931.2013.839438

48.

South Palomares

J. K.

Young

A. W

. (2018). Facial first impressions of partner preference traits: Trustworthiness, status, and attractiveness. British Journal of Psychology, 9(8), 990–1000. https://doi.org/10.1177/1948550617732388

49.

Stirrat

Perrett

D. I.

(2010). Valid facial cues to cooperation and trust: Male facial width and trustworthiness. Psychological Science, 21(3), 349–354. https://doi.org/10.1177/0956797610362647

50.

Suzuki

Ueno

Ishikawa

Kobayashi

Okubo

Nakai

(2018). Age-related differences in the activation of the mentalizing- and reward-related brain regions during the learning of others’ true trustworthiness. Neurobiology of Aging, 73, 1–8. https://doi.org/10.1016/j.neurobiolaging.2018.09.002

51.

Todorov

(2008). Evaluating faces on trustworthiness: An extension of systems for recognition of emotions signaling approach/avoidance behaviors. Annals of the New York Academy of Sciences, 1124(1), 208–224. https://doi.org/10.1196/annals.1440.012

52.

Todorov

Pakrashi

Oosterhof

N. N.

(2009). Evaluating faces on trustworthiness after minimal time exposure. Social Cognition, 27(6), 813–833. https://doi.org/10.1521/soco.2009.27.6.813

53.

Tokuda

Jimba

Yanai

Fujii

Inoguchi

(2008). Interpersonal trust and quality-of-life: A cross-sectional study in Japan. PLOS ONE, 3(12), Article e3985. https://doi.org/10.1371/journal.pone.0003985

54.

van’t Wout

Sanfey

A. G

. (2008). Friend or foe: The effect of implicit trustworthiness judgments in social decision-making. Cognition, 108(3), 796–803. https://doi.org/10.1016/j.cognition.2008.07.002

55.

Wang

Zhang

Tong

(2019). The impacts of social trust and group norms on farmers ecological consciousness. Research of Agricultural Modernization, 40(2), 215–225. https://doi.org/10.13872/j.1000-0275.2019.0001

56.

Ward

(1988). Stereotypes of old persons in Singapore. Journal of Cross-Cultural Gerontology, 3(2), 95–103. https://doi.org/10.1007/BF00116753

57.

Wernick

Manaster

G. J.

(1984). Age and the perception of age and attractiveness. The Gerontologist, 24(4), 408–414. https://doi.org/10.1093/geront/24.4.408

58.

Willis

Todorov

(2006). First impressions making up your mind after a 100-ms exposure to a face. Psychological Science, 17(7), 592–598. https://doi.org/10.1111/j.1467-9280.2006.01750.x

59.

Wilson

J. P.

Rule

N. O.

(2015). Facial trustworthiness predicts extreme criminal-sentencing outcomes. Psychological Science, 26(8), 1325–1331. https://doi.org/10.1177/0956797615590992

60.

Wilson

J. P.

Rule

N. O.

(2016). Hypothetical sentencing decisions are associated with actual capital punishment outcomes: The role of facial trustworthiness. Social Psychological and Personality Science, 7(4), 331–338. https://doi.org/10.1177/1948550615624142

61.

Gao

Wan

Wang

Yang

. . . Pan

(2018). Effects of facial trustworthiness and gender on decision making in the ultimatum game. Social Behavior and Personality: An International Journal, 46(3), 499–516. https://doi.org/10.2224/sbp.6966

62.

Zebrowitz

L. A.

Montepare

J. M.

(2008). Social psychological face perception: Why appearance matters. Social and Personality Psychology Compass, 2(3), 1497–1517. https://doi.org/10.1111/j.1751-9004.2008.00109.x

63.

Zhang

Luo

(2019). The time course and influence factors of facial trustworthiness processing. Advances in Psychological Science, 27(8), 1394–1403. https://doi.org/10.3724/SP.J.1042.2019.01394