Introduction

Individuals with Autistic Spectrum Disorders (ASD) are characterized by restricted interests, repetitive rigid behavior, and differences in communication and social interaction (American Psychiatric Association, 2013). Difficulties in interpersonal interactions and friendships are among the areas that most affect these individuals' daily functioning and quality of life (Deschamps et al., 2014). These difficulties may arise in part from differences in moral development and subsequent moral reasoning compared with individuals with typical development (TD) (e.g., see Bellesi et al., 2018; Gleichgerrcht et al., 2012; Schaller et al., 2019; Ros & Stokes, 2018). Indeed, it has been suggested that individuals with ASD exhibit more rational decision-making than TD individuals, including moral judgments in emotional scenarios (“enhanced rationality” hypothesis in ASD; see Rozenkrantz et al., 2021, for a recent review). The scarce literature on this issue has focused on adults with ASD; little is known about the development of moral reasoning from childhood to adolescence in ASD (see Dempsey et al., 2020, for a review). The present study aims to fill this gap by analyzing the influence of emotion on moral decision-making in individuals with ASD from childhood to late adolescence, thus providing valuable insights into their moral reasoning development.

The most studied dilemmas that examine the influence of emotion on moral decision-making are the trolley dilemma's impersonal and personal scenarios (Foot, 1967; Thomson, 1976). In the impersonal scenario (“switch dilemma”), the participant has to deal with a runaway trolley that cannot be stopped. The trolley would undoubtedly kill five people on the track unless a lever that diverts the trolley to another track is pulled. Critically, if diverted to the other track, only one person would be killed by the trolley. In the personal scenario (“footbridge dilemma”), the participant has to decide whether to push a large man off a footbridge to stop a train from killing five people on the track. In the personal and impersonal scenarios, the decision is the same: would you decide to sacrifice one person to save five others (i.e., utilitarian decision), or would you let the trolley hit the group of five people (i.e., deontological decision)? However, the two scenarios differ in emotional engagement: switching a lever conveys a low emotional charge (impersonal scenario), whereas pushing a person to her/his death implies a substantial emotional charge (personal scenario) (Greene et al., 2001). In the impersonal scenario, most children and adults with TD choose to pull the switch (i.e., the utilitarian decision). However, in the personal scenario, the majority of individuals with TD would not push the person onto the tracks, adopting a deontological decision instead (e.g., Greene et al., 2001; Moore et al., 2008; Pellizzoni et al., 2010). Indeed, most TD adults feel a strong emotional aversion to the utilitarian decision in personal dilemmas (e.g., Petrinovich et al., 1993; Skulmowski et al., 2014).

Rationalist theories posit that moral decision-making relies on cognitive reasoning to engage in both a cost–benefit analysis (Kohlberg, 1969) and an emotional reaction to others’ distress (Damasio, 1994; Turiel, 1983). Dual-process theories (Greene, 2009; Greene et al., 2001, 2004; Moll & de Oliveira-Souza, 2007) combined these concepts to understand moral judgments in impersonal and personal scenarios. On the one hand, controlled cognitive processes are usually involved in low emotionally charged moral dilemmas (i.e., impersonal scenarios). Therefore, they promote utilitarian judgments, which means deciding to approve a harmful action when it serves a greater good. On the other hand, intuitive, emotional processes are usually involved in high emotionally salient moral dilemmas (i.e., personal scenarios), promoting deontological judgments (i.e., the decision to favor a person's rights) even though they do not lead to a greater good.

Clearly, the development of moral maturity requires cognitive and emotional abilities. An important issue is whether this development is delayed or atypical in ASD. Indeed, differences in empathy (i.e., a response to another individual based on their psychological or contextual circumstances; Hoffman, 2001) among individuals with ASD may reflect delayed or atypical moral development (see Dempsey et al., 2020; Mathersul et al., 2013). Empathy can be distinguished into the cognitive empathy system (i.e., the ability to infer the internal mental state of another individual) and the affective empathy system (i.e., the capacity to automatically experience an appropriate emotional response to another individual's emotional state) (Blair, 2005, 2008). While individuals with ASD may have a lower cognitive empathy system than their TD peers, their affective empathy system appears to be preserved (Dziobek et al., 2008; Rueda et al., 2015) or even heightened (Baron-Cohen, 2009; Smith, 2009a)—this has been named the "empathy imbalance hypothesis" (Smith, 2009a). This is particularly problematic, as differences in empathy levels have been found to hurt the personal well-being of individuals with ASC (Ros & Stokes, 2018).

Regarding prior research on personal and impersonal dilemmas in ASD, three studies have examined the moral decisions in adults with ASD (e.g., Gleichgerrcht et al., 2012; Patil et al., 2016; Schaller et al., 2019). These studies showed that, in the impersonal switch dilemma, the percentage of utilitarian responses was similar for adults with ASD and adults with TD (e.g., 78% of adults with ASD and 72% of TD adults decided to pull the switch in the Gleichgerrcht et al., 2012, study; see also Patil et al., 2016, and Schaller et al., 2019, for converging evidence). The evidence of across-group differences in personal dilemmas in adults with ASD is not entirely conclusive. For instance, Gleichgerrcht et al. (2012) found that 36% of adults with ASD decided to push the person from the bridge, and this figure was reduced to 14% for TD adults. While Schaller et al. (2019) found, numerically, a similar pattern as Gleichgerrcht et al. (2012), Patil et al. (2016) found a non-significant difference in the opposite direction.

Importantly, Schaller et al. (2019) also examined the switch and footbridge dilemmas with a group of 16 adolescents with ASD and 22 with TD (14–18 years old). In the impersonal switch dilemma, adolescents with and without ASD showed a similar percentage of utilitarian responses (60% of adolescents with ASD, 57% of TD adolescents). In the personal footbridge dilemma, adolescents with ASD chose the utilitarian response more often (20%) than TD adolescents (9%). However, this difference did not reach significance, probably due to the small sample size. To explain their findings, Schaller et al. (2019) suggested that the differences in utilitarian decisions in moral dilemmas for individuals with ASD could reflect a delay in developing social skills that could extend to adulthood. Consistent with this view, Gleichgerrcht et al. (2012) reported an association between utilitarian decisions in moral dilemmas and difficulties in social skills for adult individuals with ASD.

Clearly, to fully understand the development of moral judgments in ASD, it is necessary to study a wide range of ages from childhood to late adolescence. Previous research on moral development in TD children has shown that children between 9 and 10 years old typically choose the utilitarian decision in moral dilemmas regardless of their emotional engagement (Bucciarelli, 2015). As children grow up, deontological judgments increase gradually because of an upturn in their cognitive and emotional resources and the capacity to represent and keep in mind the models of alternative options (see Bara et al., 2001; Bucciarelli, 2015; Bucciarelli et al., 2008). As adults and adolescents with ASD appear to show a different pattern of moral decisions in personal moral dilemmas than TD individuals (Gleichgerrcht et al., 2012; Schaller et al., 2019), the roots of this effect would presumably originate at a younger age.

Besides age, the underlying emotional and cognitive processes for the development of moral decision-making in ASD should be considered. To shed some light on these processes, previous research has assessed individuals’ ratings of the appropriateness of their decisions (i.e., a measure of cognitive reasoning) and how they felt about their decisions (i.e., a measure of emotional reaction). Gleichgerrcht et al. (2012) found that adults with ASD expressed that their utilitarian responses to the personal dilemma were inappropriate but felt less emotional arousal than TD participants. However, whereas Schaller et al. (2019) reported that the two groups did not differ in the personal dilemma, individuals with ASD rated their utilitarian decisions more permissible and felt less emotional arousal in the impersonal dilemma. Conversely, Patil et al. (2016) asked how emotionally arousing participants found the scenarios (not their decisions) and found that, regardless of the dilemma type, adults with ASD reported more emotional arousal than TD participants. These conflicting findings may be because individuals with ASD show constricted emotional functioning characterized by a difficulty in identifying and reporting their feeling states (i.e., alexithymia; see Griffin et al., 2016; Hill et al., 2004; Uljarevic & Hamilton, 2013). Prior research has shown that especially expressing their own emotions was particularly difficult for children with ASD relative to TD children (Costa et al., 2017; Lartseva et al., 2015). To minimize these interpretive issues, we chose pictograms to measure the individuals' emotional arousal and appropriateness in the present experiment. The reason is that pictograms have been considered more appropriate than verbal self-reports in ASD individuals (see Bird & Cook, 2013; Frith & Happé, 1999; Uljarevic & Hamilton, 2013) (see Appendix 1, for a depiction of the pictograms).

In sum, the present experiment applied two well-studied dilemmas (i.e., the impersonal switch dilemma and the personal footbridge moral dilemma) to children and adolescents with ASD. Thus, this research fills the gaps in knowledge about moral judgments because: (1) it includes young individuals with ASD ranging from childhood to late adolescence; (2) it includes assessing the individuals’ emotional arousal and the rating of acceptability of their moral decision employing pictograms; and (3) it includes a full assessment of ASD symptomatology (i.e., social interaction, language, communication, restrictive interests, and repetitive behavior), and their association with moral decisions.

The predictions are as follows. First, taking as reference the model on moral development in TD children (e.g., Bucchiarelli, 2015) and the “enhanced rationality” hypothesis in ASD (Rozenkrantz et al., 2021), we expected a less steep decrease in utilitarian decisions as a function of age for ASD children than TD children, especially in the personal moral dilemma. Second, based on the empathy imbalance hypothesis in individuals with ASD (Smith, 2009a), we expected children and adolescents with ASD to feel more emotional arousal due to a heightened affective empathy (Baron-Cohen, 2009; Smith, 2009a) but, at the same time, to rate their utilitarian decision as more appropriate than TD children and adolescents due to a reduced cognitive empathy (Blair, 2008; Smith, 2009a). Third, based on the importance of social cognition for moral reasoning (see Bellesi et al., 2018; Gleichgerrcht et al., 2012; Schaller et al., 2019), we expect to find a positive correlation between social interaction difficulties and utilitarian judgments in personal moral dilemmas.

Method

Participants

A final sample of 127 children and adolescents between 6 and 18 years of age took part in the study. The participants were 66 ASD out-patients from the Department of Psychiatry and Clinical Psychology, and 61 children with TD recruited in two local primary schools comparable in sex, age, and Intelligence Quotient. All individuals in the clinical group fulfilled the DSM-5 criteria (American Psychiatric Association, 2013) for ASD. The demographic and clinical details are shown in Table 1.

Eligible participants from the ASD group were children and adolescents who had been diagnosed by the referring clinicians before the study. Additionally, a trained clinical psychologist confirmed ASD diagnosis by individually interviewing parents using the Autism Diagnostic InterviewRevised (Lord et al., 1994), which measuring scales are according to the DSM-5 criteria (Reciprocal Social Interactions, Language/Communication, Repetitive Behaviors/ Interests, and evidence of onset before 36 months of age) (American Psychiatric Association, 2013).

Exclusion criteria were other psychiatric diagnoses based on the case note review in ASD children. TD children did not have a psychiatric history, as reported by their parents. Additionally, participants were excluded if they had a verbal IQ below 80 in the Kaufman Brief Intelligence Test (K-BIT; Kaufman, 1997), neurological history, major medical disorders, or medication use that could influence cognition (e.g., psychotropic medicines, treatment with corticosteroids).

Furthermore, parents completed the Child Behavior Check List (CBCL; Achenbach, 1991) to control the subclinical symptomatology in TD children and the severity of any syndromes in ASD children. CBCL assesses information on problem behavior in children between 6 and 18 through eight syndrome scales (Anxious/Depressed, Withdrawal, Somatic Complaints, Social Problems, Thought Problems, Attention Problems, Rule-Breaking Behavior, Aggressive Behavior). A standardized T-score>70 in the control group for any syndrome scale was considered an exclusion criterion. See Fig. 1 for the selection process.

Table 1

Figure 1

Procedure

After signing an informed consent form, parents answered to CBCL and ADI-R interview (in the case of ASD children). Children were assessed individually in a quiet room. Subsequently, the children received the non-moral dilemma (as a control condition) and the two moral dilemmas in a counterbalanced order. The text of the dilemmas is presented in Appendix 2. The moral dilemmas were the standard trolley dilemma and the footbridge dilemma (Greene et al., 2001, following Foot, 1967; Thomson, 1976). Participants had to choose whether to harm one person to save five people in both moral scenarios. After each decision, participants had to report on a 6-point Likert scale from 1= totally inappropriate/completely restless to 6=totally appropriate/completely calm: 1) if they thought their decision was appropriate [How appropriate is it to [take the train, switch the lever, push the person]?]”), and 2) if they were calm after their decision [“How calm do you feel about your decision?”]). The order of presentation of these questions was counterbalanced.

Data Analysis

Continuous variables were summarized using the mean (standard deviation) and median (1st, 3rd quartiles). Categorical variables were summarized using absolute and relative frequencies (in percentages). We created a Bayesian mixed logistic regression model on the utilitarian responses with the fixed factors Age, Group (ASD, TD), and Dilemma (personal, impersonal, neutral), including individuals as a random intercept. The model's reference levels were the neutral dilemma for the factor Dilemma and TD for the factor group. To assess the effect of group and dilemma on the two response variables after the decision (i.e., appropriateness, calmness), we created Bayesian mixed ordinal regression models with Appropriateness (or Calmness), Age, and Group as fixed factors, including individuals as a random intercept. A Bayesian logistic regression was conducted in the ASD group to assess the associations between utilitarian response and ADI-R. Weakly regularization priors were used for all fixed effects in the models. Interpretation of the results was performed using the 95% credible interval. An effect was considered significant when the 95% credible interval of its estimate (an Odds Ratio, OR) did not contain 1. All statistical analyses were performed using R (R Core Team, 2021) using the brms (Bürkner, 2017) and clickR (Fornes & Hervas, 2020) packages.

Results

Descriptive data showing the percentage of utilitarian responses and their assessment in terms of appropriateness and calmness for each dilemma in each group are displayed in Table 2.

Table 2

Response to the Dilemmas

The estimates and the 95% credible intervals of the Bayesian mixed logistic regression model with the response (utilitarian vs. non-utilitarian [deontological]) as a dependent variable, and age, group, and dilemma are presented in Table 3—note that the 95% credible interval of the threeway interactions between dilemma, group and age did not contain 1. Since the two three-way interactions between dilemma, group, and age make the individual estimated parameters of the models difficult to interpret, we focused on partial dependence plots where these interactions can be easily understood. These plots also include 95% credible interval bands.

Figure 2 displays the partial dependence plot for the probability of utilitarian response over Dilemma, Age, and Group (left panel: TD group; right panel: ASD group). For younger children, the probability of a utilitarian response is high in both groups regardless of the dilemma. The key differences correspond to the interplay between age and group in the personal dilemma. As shown in Fig. 2, the effect of age on utilitarian responses in the personal dilemma is much less pronounced in the ASD group than in the TD group (OR=8.83, CI 95% [1.62, 115]). Specifically, in the TD group, but not in the ASD group, the probability of utilitarian response in the personal dilemma sharply decreases as age increases (OR=0.087, CI 95% [0.005, 0.49]).

Table 3

Figure 2

Appropriateness

The estimates and 95% credible intervals for each of the effects of the Bayesian mixed ordinal regression to assess the effect of group and dilemma on appropriateness for the utilitarian responses are shown in Table 4. As can be seen in the left panel of Fig. 3, the ASD individuals who made a utilitarian response in the personal dilemma thought that their decision was more appropriate than the TD individuals (OR=3.36, CrI 95% [1.12, 10.35]).

Table 4

Calmness

Table 5 displays the estimates and 95% credible intervals for each of the effects of the Bayesian mixed ordinal regression to evaluate the effect of group and dilemma on calmness for utilitarian responses. As shown in the right panel of Fig. 3, ASD individuals felt less calm than TD individuals (OR=0.194, CrI 95% [0.062, 0.595]) after taking a utilitarian decision in personal dilemma.

Table 5

Figure 3

ADI‑R Scores and Moral Dilemmas

Regarding the association between ADI-R scores in the ASD group and the probability of utilitarian responses for the personal dilemma, only the Social Interaction score showed a significant association (see Table 6). Specifically, the score on social interaction difficulties was related to the probability of utilitarian responses in the personal dilemma (OR=1.49, CrI 95% [1.066, 2.266]; see Fig. 4, for depiction).

Table 6

Figure 4

Discussion

The present study examined the response to moral dilemmas (impersonal vs. personal) in a relatively large sample of children and adolescents with ASD vs. TD matched controls. As expected, we found that the probability of a utilitarian response decreased with age. Critically, this effect was modulated by group and type of dilemma (personal vs. impersonal). The probability of a utilitarian response decreased steeply with age for personal dilemmas for TD children. Conversely, the probability of a utilitarian response to the dilemmas barely decreased with age for children with ASD. Another notable finding is that, after making a utilitarian decision in a personal dilemma, the individuals with ASD not only rated their decisions as more appropriate than the TD controls but also felt less calm about their decisions. Finally, we found that choosing the utilitarian response in the personal dilemma in ASD individuals was associated with more difficulties in social interaction.

Regarding the moral development in ASD, we found more utilitarian decisions in the personal footbridge dilemma than in the impersonal switch dilemma, thus extending previous findings with adults (Gleichgerrcht et al., 2012; Patil et al., 2016) and adolescents (Schaller et al., 2019) to a children population. Critically, the inclusion of children and adolescents between the ages of 6 to 18 allows us to make inferences on moral reasoning development in individuals with ASD. First, the probability of a utilitarian response for 6-year-old children was nearly 100% for TD and ASD children. Second, in the range of 8 to 12 years of age, the probability of a utilitarian decision in a personal dilemma sharply decreased as age increased for TD children (see Bucchiarelli, 2015, for a model); however, this was not the case for ASD children. Indeed, for children with ASD, a decrease in the probability of a utilitarian decision in a personal dilemma only became visible from 14 years onwards. At the age of 18, although the probability of a utilitarian decision decreased with age in both groups, individuals with ASD showed a higher probability of a utilitarian response in a personal moral dilemma than TD individuals. Thus, the moral judgment of children with ASD develops atypically from 8 years of age onwards, in the way that it is delayed from the one of TD children. This delay relative to TD individuals remains at least up to young adulthood. These differences in moral judgments fit well with the “enhanced rationality” hypothesis (Rozenkrantz et al., 2021), which proposes that individuals with ASD use more rational and bias-free decision-making processes than TD individuals. Although one might argue that moral judgments could be hard to categorize in terms of rationality, utilitarian choices most often represent the option in which the greater good is maximized (i.e., most people are saved, despite the ethical consequences of the decision). Thus, an enhanced rational way of thinking in the development of ASD children may account for the increasing differences in moral decisions between young individuals with ASD and TD individuals in the present study.

Another novel feature of the present study was the individuals' evaluation of their choice (cognitive: appropriateness; emotional: calmness) after making a utilitarian judgment in the personal dilemma (i.e., pushing a person to death to save five others). While children with ASD rated their utilitarian decisions as more appropriate than the TD children, they also felt less calm about these judgments. The fact that individuals with ASD showed an incongruent assessment in these two domains favors those theories on ASD that posit that individuals with ASD have a decreased cognitive empathy system (Blair, 2008) but a heightened affective empathy system (Baron-Cohen, 2009; Smith, 2009a). Of note, previous studies showed conflicting findings on the rating of appropriateness and calmness in individuals with ASD, possibly due to difficulties in expressing the felt emotions (Costa et al., 2017; Lartseva et al., 2015). To assist the children with ASD, we chose pictograms to measure the individual level of calmness and appropriateness of the answers, so that possible difficulties in the expression of emotions merely have a minimal effect on our data.

Further evidence for a heightened emotional reaction to others’distress among individuals with ASD is coming from neuroimaging studies. During affective empathy tasks, a hyper-functioning of the amygdala occurs in individuals with ASD (Baron-Cohen et al., 2000), which, in turn, prompts individuals with ASD to act in a hyper-reactive way in social and emotional contexts (Intense World Hypothesis; Markram, Rinaldi, & Markram, 2007). This overstimulation would be reflected as less calmness after making a utilitarian response in ASD individuals than in TD controls. Consequently, when confronted with an emotionally charged moral decision (especially the case in personal moral dilemmas), a hyper-functioning affective empathy system would lead to overwhelmingly intense processing of the situation (BaronCohen et al., 2000). Following the Intense World Theory, a defective bottom-up modulation also has paradoxical cognitive consequences (Markram & Markram, 2010). Due to the excessive distress in potentially overwhelming situations, individuals with ASD may tend to withdraw from them by seeking less emotionally charged reasoning (Markram et al., 2007; Smith, 2009a, 2009b). In this way, individuals with ASD may manage to avoid overwhelming emotional processes by choosing a utilitarian decision that is interpreted as more “appropriate” (i.e., it saves the most lives). The current findings lead to new insights into empathy processing in individuals with ASD. Following the idea of the “double empathy problem” (Milton, 2012), the current findings help to reshape the view that a different empathy processing in ASD individuals is seen as a “deficit”, but rather as a different way of perceiving empathy and social interactions (Milton, 2012).

The present study also revealed that more reported social interaction difficulties in children/adolescents with ASD were associated with a higher probability of a utilitarian response in the personal moral dilemma. Thus, the high choice of utilitarian judgments in emotional scenarios may be related to the abnormal behavior in situations such as emotional sharing, offering and seeking comfort, social smiling, and responding to other children, which is often observed in individuals with ASD (Lord et al., 1994). In sum, the correlation between social interaction difficulties and utilitarian responses in children and adolescents with ASD extends previous evidence with adults. For instance, Gleichgerrcht et al. (2012) reported that, in personal moral dilemmas, a higher probability of utilitarian decisions was associated with more difficulties in social cognition in adults with ASD.

To our knowledge, the present study is the first that examined the development of moral judgments in individuals with ASD ranging from childhood to late adolescence, thus filling the gaps in knowledge about young individuals with ASD. Another novel element is that we measured the participants' emotional and cognitive states after making a utilitarian decision. Despite the strengths of our study, certain limitations have to be taken into consideration. First, we investigated moral judgments in children and adolescents using a cross-sectional rather than a longitudinal design. We chose this option because a longitudinal study from early childhood to late adolescence would have interpretive issues (e.g., same dilemmas at multiple times) and potential methodological problems (e.g., participants' dropping out of the study). Second, our study only employed two moral dilemmas: the personal “footbridge” dilemma and the impersonal “switch” dilemma. Although these are by far the two most studied moral dilemmas and serve as the basis for leading theories of moral judgments (e.g., Greene et al., 2001, 2004), including a wider variety of moral dilemmas in future research would have led to greater generalizability (e.g., see Patil et al., 2016).

Altogether, the present study revealed that emotionbased decision-making in ASD individuals does not develop similarly to TD individuals (i.e., probability of a utilitarian response in a moral dilemma decreased much more shallowly with age for children with ASD). Moreover, ASD individuals also show differences in the underlying cognitive beliefs and emotional responses to their behavioral decision. This difference was particularly manifest in an emotionally charged scenario (i.e., pushing a person from the bridge to save five lives). Notably, individuals with ASD felt less calm than their TD peers (exaggerated affective empathy system) after choosing a utilitarian decision, but they rated their utilitarian decisions as more appropriate (less developed cognitive empathy system). This finding suggests that by deciding on a utilitarian course of action 'for the greater good', those with ASD would place themselves in a position that caused them greater distress due to their awareness of the affective consequences of their decision. Furthermore, utilitarian judgments in personal dilemmas were related to more social interaction difficulties. Thus, therapeutic interventions that aim to improve the social skills of children with ASD should focus on both the development of their cognitive empathy system and the design of efficient coping strategies for emotionally charged situations.

Appendix 1. Example of pictogram for self‑report

See Fig. 5

Appendix 2. Text of the dilemmas used in the study

(a) Trolley dilemma A runaway trolley is coming down a railway track. The trolley is approaching five persons in a sidetrack who will not be able to leave in time before the trolley will hit them. If the trolley will continue it will certainly overrun the five persons and will kill all of them. You are standing beside a switch of a railway track. The only way for you to save the five persons is to throw the switch. This will cause the trolley to go to another track where it will overrun and kill one person, while the five persons on the first track will survive. Would you decide whether to flip a switch to redirect a trolley onto the man or whether to allow the trolley to hit the five people?

(b) Footbridge dilemma In this scenario, the trolley is approaching five persons on the railway track. This time you are on a bridge above the track, and a man is beside you. The only way for you to save the five persons is to push the man off the bridge so that his body would stop the trolley from hitting five people further down the tracks. Would you decide whether to push the man off the bridge or whether to allow the trolley to hit the five people?

(c) Neutral dilemma You are traveling from Valencia to Barcelona in order to attend a very important meeting that starts at 12:00. To travel, you can take either the train or the bus. The train will get you there just in time for the meeting: 5 min before. The bus is scheduled to arrive one hour before the meeting, but can sometimes be several hours late because of traffic. It would be great to arrive an hour before the meeting and walk around town, but you can't afford to be late. Would you take the train instead of the bus in order to guarantee you won't be late for the meeting, even if you arrive just in time?

Introduction

Individuals with Autistic Spectrum Disorders (ASD) often display restricted interests, repetitive behaviors, and distinct patterns in communication and social interaction. Difficulties in social interactions and friendships significantly impact daily functioning and overall quality of life for these individuals. These challenges may stem from differences in moral development and reasoning compared to individuals with typical development. Some research suggests that individuals with ASD exhibit more rational decision-making, including moral judgments in emotional situations, a concept sometimes referred to as the "enhanced rationality" hypothesis in ASD. Much of the existing research on this topic has focused on adults with ASD, leaving a gap in understanding how moral reasoning develops from childhood to adolescence in this population. This study aims to address this gap by examining how emotion influences moral decision-making in individuals with ASD across childhood and late adolescence, thereby offering insights into their moral reasoning development.

The "trolley dilemma" and its variations are commonly used to study how emotion affects moral decision-making. In the impersonal "switch dilemma," a runaway trolley is heading towards five people, and the only way to save them is to pull a lever, diverting the trolley to another track where it will kill one person. In the personal "footbridge dilemma," the decision involves pushing a large man off a footbridge to stop a train from killing five people. Both scenarios pose the same ethical choice: sacrificing one person to save five (a utilitarian decision) or allowing five people to die (a deontological decision). The key difference lies in emotional involvement: switching a lever is less emotionally charged (impersonal), while directly pushing someone to their death is highly emotional (personal). Typically, most children and adults without ASD choose the utilitarian option in the impersonal scenario. However, in the personal scenario, most individuals without ASD would not push the person, instead making a deontological choice, often driven by a strong emotional aversion to direct harm.

Moral decision-making is understood through various theories. Rationalist theories propose that moral judgments involve both a logical cost-benefit analysis and an emotional response to another person's distress. Dual-process theories combine these ideas, suggesting that low-emotion dilemmas (impersonal) engage controlled cognitive processes, leading to utilitarian judgments (approving harm for a greater good). High-emotion dilemmas (personal) activate intuitive, emotional processes, promoting deontological judgments (prioritizing individual rights, even if it doesn't save more lives). The development of moral maturity clearly relies on both cognitive and emotional abilities. A crucial question is whether this development is delayed or atypical in individuals with ASD. Differences in empathy, defined as a response to another individual based on their situation, may indicate delayed or atypical moral development in ASD. Empathy can be separated into cognitive empathy (understanding another's mental state) and affective empathy (experiencing an appropriate emotional response to another's emotions). While individuals with ASD may have lower cognitive empathy, their affective empathy might be preserved or even heightened, a concept known as the "empathy imbalance hypothesis." This imbalance can significantly impact personal well-being.

Previous research on moral dilemmas in adults with ASD has shown that utilitarian responses in impersonal scenarios are similar to those of typically developing adults. However, findings for personal dilemmas in adults with ASD have been inconsistent. Studies on adolescents with ASD also showed similar utilitarian responses in impersonal dilemmas but a tendency towards more utilitarian responses in personal dilemmas compared to typically developing adolescents, although these differences were not always statistically significant, possibly due to small sample sizes. Researchers have suggested that differences in utilitarian decisions in ASD might reflect a delay in social skill development that extends into adulthood. To fully understand moral judgments in ASD, it is essential to study a broad age range from childhood to late adolescence. Research on typically developing children indicates that younger children often make utilitarian decisions regardless of emotional engagement, with deontological judgments gradually increasing with age as cognitive and emotional resources develop. Given that adults and adolescents with ASD appear to show a different pattern of moral decisions in personal dilemmas, the origins of this effect likely begin at younger ages.

Beyond age, understanding the emotional and cognitive processes behind moral decision-making in ASD is important. Previous studies have assessed how individuals rate the appropriateness of their decisions and their emotional feelings about those decisions. Findings have been conflicting, possibly because individuals with ASD may experience alexithymia, a difficulty in identifying and reporting their emotional states. To minimize these interpretive challenges, this study utilized pictograms to measure emotional arousal and appropriateness. This approach is considered more suitable than verbal self-reports for individuals with ASD. This study, therefore, applied two well-known moral dilemmas to children and adolescents with ASD to fill knowledge gaps by: including young individuals across a wide age range; assessing emotional arousal and decision acceptability using pictograms; and evaluating ASD symptomatology and its link to moral decisions. Predictions included a less steep decrease in utilitarian decisions with age for ASD children, especially in personal dilemmas, supporting the "enhanced rationality" hypothesis. It was also predicted that children and adolescents with ASD would report more emotional arousal due to heightened affective empathy but rate utilitarian decisions as more appropriate due to reduced cognitive empathy, aligning with the "empathy imbalance hypothesis." Finally, a positive correlation was expected between social interaction difficulties and utilitarian judgments in personal moral dilemmas.

Method

Participants

A total of 127 children and adolescents, aged 6 to 18, participated in the study. This included 66 out-patients with ASD from a Department of Psychiatry and Clinical Psychology, and 61 children with typical development (TD) recruited from two local primary schools. The two groups were comparable in terms of sex, age, and Intelligence Quotient. All individuals in the clinical group met the diagnostic criteria for ASD.

Eligible participants from the ASD group had received a prior diagnosis from clinicians. A trained clinical psychologist further confirmed the ASD diagnosis by interviewing parents using a standardized diagnostic interview that aligns with diagnostic criteria, covering reciprocal social interactions, language and communication, repetitive behaviors and interests, and evidence of onset before 36 months of age.

Exclusion criteria for ASD children included other psychiatric diagnoses identified during a review of case notes. TD children had no reported psychiatric history. Additionally, participants were excluded if their verbal IQ was below 80, if they had a neurological history, major medical disorders, or were using medication that could affect cognition, such as psychotropic drugs or corticosteroids. Parents also completed a behavior checklist to control for subclinical symptoms in TD children and to assess the severity of any syndromes in ASD children. A standardized score above 70 on any syndrome scale in the control group led to exclusion.

Procedure

After providing informed consent, parents completed the behavior checklist and the diagnostic interview (for children with ASD). Children were assessed individually in a quiet room. They then received a non-moral dilemma (as a control condition) and the two moral dilemmas in a randomized order. The moral dilemmas were the standard trolley dilemma and the footbridge dilemma. Participants were asked to choose whether to harm one person to save five people in both moral scenarios. Following each decision, participants used a 6-point scale, from "totally inappropriate/completely restless" to "totally appropriate/completely calm," to report: 1) how appropriate they felt their decision was, and 2) how calm they felt about their decision. The order of these questions was also randomized.

Data Analysis

Continuous variables were summarized using mean and standard deviation, and median with interquartile ranges. Categorical variables were presented as absolute and relative frequencies. A Bayesian mixed logistic regression model was developed for utilitarian responses, incorporating age, group (ASD, TD), and dilemma type (personal, impersonal, neutral) as fixed factors, with individuals as a random intercept. To evaluate the impact of group and dilemma on the post-decision variables (appropriateness, calmness), Bayesian mixed ordinal regression models were created, using appropriateness (or calmness), age, and group as fixed factors, with individuals as a random intercept. For the ASD group, a Bayesian logistic regression assessed the links between utilitarian responses and diagnostic interview scores. Weakly regularizing prior distributions were used for all fixed effects. Results were interpreted using the 95% credible interval, with an effect considered significant if the 95% credible interval of its estimate (an Odds Ratio) did not include 1. All statistical analyses were performed using specialized software packages.

Results

Descriptive data showing the percentage of utilitarian responses and their assessment in terms of appropriateness and calmness for each dilemma in each group were recorded.

Response to the Dilemmas

The estimated effects and their 95% credible intervals from the Bayesian mixed logistic regression model, with the type of response (utilitarian versus non-utilitarian or deontological) as the outcome variable and age, group, and dilemma as factors, were analyzed. Since the interactions between dilemma, group, and age were complex, the findings were best understood through visual plots. These plots showed that for younger children, the likelihood of a utilitarian response was high in both groups regardless of the dilemma. Crucially, the key differences appeared in the interplay between age and group in the personal dilemma. The effect of age on utilitarian responses in the personal dilemma was significantly less pronounced in the ASD group compared to the TD group. Specifically, for the TD group, the probability of a utilitarian response in the personal dilemma decreased sharply as age increased, but this pattern was not observed in the ASD group.

Appropriateness

Estimates and 95% credible intervals from the Bayesian mixed ordinal regression model, assessing the effect of group and dilemma on the appropriateness ratings for utilitarian responses, were examined. Individuals with ASD who made a utilitarian response in the personal dilemma perceived their decision as significantly more appropriate than individuals with TD.

Calmness

The estimates and 95% credible intervals from the Bayesian mixed ordinal regression model, evaluating the effect of group and dilemma on calmness ratings for utilitarian responses, were also analyzed. Individuals with ASD reported feeling less calm than individuals with TD after making a utilitarian decision in a personal dilemma.

ADI‑R Scores and Moral Dilemmas

Regarding the relationship between diagnostic interview scores within the ASD group and the likelihood of utilitarian responses in the personal dilemma, only the social interaction score showed a significant association. Specifically, higher scores indicating greater social interaction difficulties were linked to a higher probability of utilitarian responses in the personal dilemma.

Discussion

This study investigated responses to moral dilemmas (impersonal versus personal) in a substantial sample of children and adolescents with ASD and matched typically developing (TD) controls. As predicted, the likelihood of a utilitarian response generally decreased with age. Importantly, this effect varied based on the group and the type of dilemma. For personal dilemmas, the probability of a utilitarian response decreased steeply with age for TD children. In contrast, for children with ASD, the probability of a utilitarian response to these dilemmas showed a much less significant decrease with age. Another notable finding was that after making a utilitarian decision in a personal dilemma, individuals with ASD not only rated their decisions as more appropriate than TD controls but also reported feeling less calm about their choices. Finally, among individuals with ASD, choosing a utilitarian response in the personal dilemma was associated with greater difficulties in social interaction.

Regarding moral development in ASD, the study observed more utilitarian decisions in the personal footbridge dilemma compared to the impersonal switch dilemma, extending previous findings from adults and adolescents to a child population. The inclusion of participants aged 6 to 18 allowed for inferences about moral reasoning development in individuals with ASD. For 6-year-old children, the likelihood of a utilitarian response was nearly 100% in both TD and ASD groups. However, between 8 and 12 years of age, the probability of a utilitarian decision in a personal dilemma sharply declined for TD children, a pattern not seen in ASD children. A decrease in utilitarian decisions in personal dilemmas for children with ASD only became noticeable from 14 years onward. By age 18, although the probability of a utilitarian decision decreased in both groups, individuals with ASD still showed a higher probability of utilitarian responses in personal moral dilemmas than TD individuals. This suggests that the moral judgment of children with ASD develops atypically from around 8 years of age, exhibiting a delay compared to TD children, a delay that persists into young adulthood. These differences in moral judgments align with the "enhanced rationality" hypothesis, which suggests that individuals with ASD employ more rational and less biased decision-making processes. Utilitarian choices often represent the option that maximizes the greater good (saving more lives), and thus, an enhanced rational way of thinking in children with ASD may explain the increasing divergence in moral decisions between young individuals with ASD and TD individuals.

A novel aspect of this study was the assessment of individuals' cognitive evaluation (appropriateness) and emotional reaction (calmness) after making a utilitarian judgment in a personal dilemma. While children with ASD rated their utilitarian decisions as more appropriate than TD children, they also reported feeling less calm about these judgments. This incongruent assessment across cognitive and emotional domains supports theories suggesting that individuals with ASD have a decreased cognitive empathy system but a heightened affective empathy system. The use of pictograms to measure calmness and appropriateness helped to mitigate potential difficulties in expressing felt emotions, which has been an issue in previous research. Further evidence for heightened emotional reactions to others' distress in individuals with ASD comes from neuroimaging studies, which have shown increased activity in brain regions associated with emotion during affective empathy tasks. This overstimulation, consistent with the "Intense World Hypothesis," might lead to hyper-reactive responses in social and emotional situations. Such overwhelming intensity could cause individuals with ASD to withdraw or seek less emotionally charged reasoning, leading them to choose a utilitarian decision that they interpret as more "appropriate" to manage distress. These findings contribute to a refined understanding of empathy in ASD, moving beyond viewing different empathy processing as simply a "deficit" towards recognizing it as a distinct way of perceiving empathy and social interactions, as proposed by the "double empathy problem."

The study also found that greater reported social interaction difficulties in children and adolescents with ASD were associated with a higher likelihood of making a utilitarian response in the personal moral dilemma. This suggests that the prevalence of utilitarian judgments in emotional scenarios may be linked to behaviors often observed in individuals with ASD, such as challenges in emotional sharing, offering and seeking comfort, social smiling, and responding to peers. This correlation between social interaction difficulties and utilitarian responses in younger individuals with ASD extends previous findings observed in adults. Despite the strengths of this study, including being the first to examine moral judgment development in ASD across such a wide age range and using pictograms for emotional and cognitive states, certain limitations must be acknowledged. The study employed a cross-sectional rather than a longitudinal design, and only two types of moral dilemmas were used. While these are widely studied dilemmas, incorporating a broader variety in future research could enhance the generalizability of the findings.

In conclusion, this study revealed that emotion-based decision-making in individuals with ASD does not develop in the same way as in TD individuals; specifically, the probability of a utilitarian response in a moral dilemma decreased much less significantly with age for children with ASD. Moreover, individuals with ASD exhibited differences in the underlying cognitive beliefs and emotional responses to their behavioral decisions, particularly in emotionally charged scenarios. Individuals with ASD felt less calm than their TD peers (suggesting an exaggerated affective empathy system) after choosing a utilitarian decision, yet they rated these utilitarian decisions as more appropriate (suggesting a less developed cognitive empathy system). This indicates that by deciding on a utilitarian course of action "for the greater good," individuals with ASD might place themselves in a situation that causes them greater distress due to their awareness of the emotional consequences of their decision. Furthermore, utilitarian judgments in personal dilemmas were associated with more social interaction difficulties. These findings suggest that therapeutic interventions aimed at improving social skills in children with ASD should address both the development of their cognitive empathy system and the creation of effective coping strategies for emotionally challenging situations.

Introduction

Individuals diagnosed with Autistic Spectrum Disorders (ASD) often exhibit specific patterns of restricted interests, rigid repetitive behaviors, and distinct approaches to communication and social interaction. Challenges in forming friendships and engaging in interpersonal interactions can significantly affect their daily lives and overall well-being. These difficulties may partly stem from differences in moral development and how they engage in moral reasoning compared to individuals with typical development (TD). Some research suggests that individuals with ASD might make more rational decisions, including moral judgments in emotionally charged situations, aligning with an "enhanced rationality" hypothesis. However, most existing research has focused on adults with ASD, leaving a gap in understanding how moral reasoning develops from childhood through adolescence in this population. The current study aims to address this by examining the role of emotion in moral decision-making among individuals with ASD from childhood to late adolescence, offering insights into their moral development.

The "trolley dilemma" is a widely studied scenario for exploring the influence of emotion on moral decisions, presented in both impersonal and personal forms. In the impersonal "switch dilemma," a runaway trolley is headed for five people, and the only way to save them is to pull a lever, diverting the trolley to another track where it would kill one person. The personal "footbridge dilemma" involves pushing a large man off a bridge to stop a trolley from killing five people. Both scenarios pose the same fundamental choice: sacrifice one person to save five (a utilitarian decision) or allow five people to die (a deontological decision). The key difference lies in the emotional engagement: pulling a lever is considered less emotionally charged than physically pushing a person to their death. Typically, in the impersonal scenario, most TD children and adults choose the utilitarian option (pulling the switch). However, in the personal scenario, most TD individuals opt for the deontological decision, choosing not to push the person, often due to a strong emotional aversion to the utilitarian choice.

Theories of moral decision-making propose that individuals rely on both logical reasoning to weigh costs and benefits and emotional responses to others' distress. Dual-process theories combine these ideas, suggesting that controlled cognitive processes guide judgments in less emotionally charged impersonal dilemmas, promoting utilitarian choices that serve a greater good. In contrast, intuitive emotional processes tend to dominate highly emotional personal dilemmas, leading to deontological judgments that prioritize individual rights, even if it does not save the most lives.

Developing moral maturity requires both cognitive and emotional capabilities. A relevant question is whether this development is delayed or atypical in individuals with ASD, particularly given observed differences in empathy. Empathy is often divided into cognitive empathy (the ability to understand another person's mental state) and affective empathy (the ability to experience an appropriate emotional response to another's emotions). While individuals with ASD may show reduced cognitive empathy compared to their TD peers, their affective empathy might be preserved or even heightened, a concept known as the "empathy imbalance hypothesis." These differences in empathy levels can negatively impact the personal well-being of individuals with ASD.

Previous research examining moral decisions in adult and adolescent ASD populations using these dilemmas has yielded mixed findings. Some studies suggest similar utilitarian responses in impersonal dilemmas for both ASD and TD adults. However, in personal dilemmas, some studies show that adults with ASD are more likely to make utilitarian choices than TD adults, while others report no significant differences. For adolescents, one study found that ASD adolescents chose utilitarian responses more often in personal dilemmas than TD adolescents, though this difference was not statistically significant, possibly due to small sample sizes. These observations suggest that differences in moral judgments in ASD might reflect a delay in social skill development that could persist into adulthood. The current study seeks to provide clarity by examining a broader age range and assessing both emotional and cognitive aspects of decision-making.

Method

Participants

The study included a total of 127 children and adolescents, aged 6 to 18 years. This sample comprised 66 individuals with ASD, recruited from a psychiatric and clinical psychology department, and 61 TD children from two local primary schools. The two groups were comparable in terms of sex, age, and Intelligence Quotient. All individuals in the clinical group met the diagnostic criteria for ASD, as outlined in the DSM-5. (See Table 1 for demographic and clinical details).

Eligible participants in the ASD group had received a prior clinical diagnosis. Additionally, a trained clinical psychologist confirmed the ASD diagnosis by interviewing parents using the Autism Diagnostic Interview-Revised, which assesses social interaction, language/communication, repetitive behaviors/interests, and evidence of onset before 36 months of age, consistent with DSM-5 criteria.

Exclusion criteria for ASD participants included other psychiatric diagnoses. TD children had no reported psychiatric history. All participants were excluded if they had a verbal IQ below 80, a history of neurological issues, major medical disorders, or were taking medications that could affect cognition. Parents also completed the Child Behavior Check List (CBCL) to monitor subclinical symptoms in TD children and syndrome severity in ASD children. A standardized T-score above 70 on any CBCL syndrome scale for the control group was an exclusion criterion. (See Fig. 1 for the selection process).

Procedure

After providing informed consent, parents completed the CBCL and, for ASD children, the ADI-R interview. Children were then assessed individually in a quiet room. They received a non-moral dilemma (as a control) and two moral dilemmas in a counterbalanced order. The moral dilemmas used were the standard trolley dilemma and the footbridge dilemma. Participants were asked to choose whether to harm one person to save five in both moral scenarios. Following each decision, participants used a 6-point Likert scale (from 1 = totally inappropriate/completely restless to 6 = totally appropriate/completely calm) to rate: 1) how appropriate their decision was, and 2) how calm they felt about their decision. The order of these questions was counterbalanced.

Data Analysis

Continuous variables were summarized using mean, standard deviation, median, and interquartile ranges. Categorical variables were described using absolute and relative frequencies. A Bayesian mixed logistic regression model was constructed to analyze utilitarian responses, with age, group (ASD, TD), and dilemma type (personal, impersonal, neutral) as fixed factors, and individuals as a random intercept. Bayesian mixed ordinal regression models were used to assess the impact of group and dilemma on appropriateness and calmness, with age, group, and dilemma type as fixed factors and individuals as a random intercept. For the ASD group, a Bayesian logistic regression explored the relationship between utilitarian responses and ADI-R scores. Weak regularization priors were applied to all fixed effects. Results were interpreted using 95% credible intervals, with an effect considered significant if its 95% credible interval for the Odds Ratio (OR) did not include 1. All statistical analyses were performed using R statistical software.

Results

Descriptive data, including the percentage of utilitarian responses and their assessment of appropriateness and calmness for each dilemma in both groups, are presented in Table 2.

Response to the Dilemmas

Table 3 shows the estimates and 95% credible intervals from the Bayesian mixed logistic regression model, which examined utilitarian versus non-utilitarian responses based on age, group, and dilemma type. The 95% credible interval of the three-way interactions between dilemma, group, and age did not include 1, indicating significant interactions. Given the complexity of interpreting individual parameters in these interactions, the focus was placed on partial dependence plots for clearer understanding. (These plots also included 95% credible interval bands).

Figure 2 illustrates the probability of a utilitarian response across dilemma type, age, and group. For younger children, the probability of a utilitarian response was high in both groups regardless of the dilemma. Crucially, differences emerged from the interaction between age and group in the personal dilemma. The effect of age on utilitarian responses in the personal dilemma was considerably less pronounced in the ASD group than in the TD group. Specifically, in the TD group, the probability of a utilitarian response in the personal dilemma sharply declined with increasing age; this pattern was not observed in the ASD group.

Appropriateness

The estimates and 95% credible intervals for the effects of group and dilemma on the appropriateness ratings for utilitarian responses are presented in Table 4. As depicted in the left panel of Fig. 3, individuals with ASD who made a utilitarian response in the personal dilemma rated their decision as more appropriate compared to TD individuals.

Calmness

Table 5 provides the estimates and 95% credible intervals for the effects of group and dilemma on calmness ratings for utilitarian responses. As shown in the right panel of Fig. 3, individuals with ASD reported feeling less calm than TD individuals after making a utilitarian decision in a personal dilemma.

ADI‑R Scores and Moral Dilemmas

Regarding the association between ADI-R scores within the ASD group and the likelihood of utilitarian responses in the personal dilemma, only the Social Interaction score showed a significant relationship (see Table 6). Specifically, higher scores indicating greater social interaction difficulties were associated with a higher probability of utilitarian responses in the personal dilemma. (See Fig. 4 for a visual representation).

Discussion

The study investigated responses to moral dilemmas (impersonal vs. personal) in a substantial sample of children and adolescents with ASD and matched TD controls. As hypothesized, the likelihood of a utilitarian response generally decreased with age. Importantly, this effect varied depending on the group and the type of dilemma. For TD children, the probability of a utilitarian response in personal dilemmas decreased sharply with age. In contrast, for children with ASD, this probability showed a much less pronounced decline with age. Another notable finding was that after making a utilitarian decision in a personal dilemma, individuals with ASD rated their decision as more appropriate than TD controls but simultaneously reported feeling less calm about it. Furthermore, among individuals with ASD, choosing the utilitarian response in personal dilemmas was associated with greater difficulties in social interaction.

Regarding moral development in ASD, the study observed more utilitarian decisions in the personal "footbridge" dilemma than in the impersonal "switch" dilemma, extending previous findings from adults and adolescents to a child population. The inclusion of participants from 6 to 18 years of age allowed for inferences about moral reasoning development in ASD. At age 6, both TD and ASD children showed a nearly 100% probability of a utilitarian response. However, between 8 and 12 years of age, the probability of a utilitarian decision in personal dilemmas declined sharply for TD children but not for ASD children. For children with ASD, a noticeable decrease in utilitarian decisions in personal dilemmas only became apparent from age 14 onwards. By age 18, although the probability of utilitarian decisions decreased in both groups with age, individuals with ASD still showed a higher likelihood of utilitarian responses in personal moral dilemmas than TD individuals. This suggests that the moral judgment of children with ASD develops atypically, showing a delay compared to TD children from age 8, a delay that persists into young adulthood. These differences in moral judgments align with the "enhanced rationality" hypothesis, which proposes that individuals with ASD use more rational and bias-free decision-making processes. Utilitarian choices often represent the option that maximizes the greater good, suggesting that an enhanced rational approach in ASD development may contribute to the observed differences in moral decisions.

A unique aspect of the study was the assessment of participants' cognitive (appropriateness) and emotional (calmness) evaluations after making a utilitarian judgment in the personal dilemma (e.g., pushing a person to save five others). While children with ASD rated their utilitarian decisions as more appropriate than TD children, they also reported feeling less calm about these judgments. This incongruent assessment supports theories suggesting that individuals with ASD may have a reduced cognitive empathy system but a heightened affective empathy system, known as the "empathy imbalance hypothesis." The use of pictograms to measure calmness and appropriateness helped minimize potential difficulties ASD children might have in expressing emotions, which could have confounded findings in previous studies.

Further evidence for heightened emotional reactions to others' distress in individuals with ASD comes from neuroimaging studies, which have shown increased amygdala activity during affective empathy tasks in this population. This hyper-functioning could lead to intense processing of social and emotional situations. According to the "Intense World Hypothesis," such overwhelming stimulation might lead individuals with ASD to withdraw from these situations by seeking less emotionally charged reasoning. In this way, they might manage intense emotional processes by choosing a utilitarian decision, which they then interpret as more "appropriate" because it saves more lives. These findings contribute to a refined understanding of empathy in ASD, viewing it not merely as a "deficit" but as a distinct way of processing empathy and social interactions, consistent with the "double empathy problem" perspective.

The study also revealed a significant association between greater reported social interaction difficulties in children and adolescents with ASD and a higher probability of making utilitarian responses in personal moral dilemmas. This suggests that the frequent choice of utilitarian judgments in emotional scenarios may be linked to observed atypical behaviors in emotional sharing, comfort-seeking, social smiling, and responding to peers, commonly seen in individuals with ASD. This correlation between social interaction challenges and utilitarian responses in younger individuals with ASD extends previous findings reported in adults.

The current study is significant as it provides the first examination of moral judgment development across a broad age range (childhood to late adolescence) in individuals with ASD. The inclusion of emotional arousal and appropriateness ratings using pictograms further enhanced the understanding of underlying processes. Despite these strengths, some limitations exist. The study employed a cross-sectional design rather than a longitudinal one, though this approach was chosen to mitigate interpretive and methodological issues associated with long-term studies. Additionally, only two moral dilemmas (the impersonal "switch" and personal "footbridge") were used. While these are foundational to moral judgment theories, incorporating a wider variety of dilemmas in future research could enhance generalizability.

In summary, the study demonstrates that emotion-based decision-making in individuals with ASD develops differently from TD individuals, with the probability of utilitarian responses in moral dilemmas decreasing much more gradually with age for those with ASD. Furthermore, individuals with ASD show distinct cognitive beliefs and emotional responses to their decisions. This was particularly evident in emotionally charged personal dilemmas, where ASD individuals reported feeling less calm than their TD peers (suggesting an exaggerated affective empathy system) yet rated their utilitarian decisions as more appropriate (indicating a less developed cognitive empathy system). This suggests that in pursuing the "greater good" through a utilitarian choice, individuals with ASD might experience greater distress due to their awareness of the affective consequences of their actions. Moreover, utilitarian judgments in personal dilemmas were linked to more social interaction difficulties. These findings underscore the importance of therapeutic interventions for children with ASD that target both the development of cognitive empathy and the creation of effective coping strategies for emotionally charged situations.

Introduction

Individuals with Autism Spectrum Disorder (ASD) often show specific patterns in their behavior, such as very focused interests, repetitive actions, and distinct ways of communicating and interacting with others. These differences can make daily life challenging, particularly when it comes to social interactions and friendships. Research suggests that these difficulties might be partly linked to how moral reasoning develops in individuals with ASD compared to those without the disorder. Some theories propose that individuals with ASD may make decisions based more on logic than emotion, even in emotionally charged situations—a concept known as "enhanced rationality." This study aims to understand how emotions influence moral decision-making in individuals with ASD from childhood through late adolescence.

To study the influence of emotion on moral choices, researchers often use dilemmas like the "trolley problem." This involves two main scenarios: the impersonal "switch dilemma" and the personal "footbridge dilemma." In the switch dilemma, a runaway trolley threatens five people, and the only way to save them is to pull a lever, diverting the trolley to another track where it will kill one person. This scenario generally involves low emotional engagement. In the footbridge dilemma, to save five people, a person must decide whether to push a large man off a footbridge to stop the trolley. This scenario is highly emotional. For individuals without ASD, the majority of children and adults typically choose to pull the switch (a utilitarian decision to save more lives). However, in the more emotional footbridge dilemma, most individuals without ASD would not push the person, opting for a deontological decision (favoring an individual's rights over the greater good).

Moral decision-making involves both logical thought processes and emotional responses. Cognitive processes usually guide choices in situations with low emotional impact, often leading to utilitarian judgments (decisions that achieve the greatest good). In contrast, intuitive emotional processes typically guide choices in highly emotional dilemmas, promoting deontological judgments (decisions that uphold individual rights, even if they don't save the most lives). The development of mature moral reasoning requires a balance of both cognitive and emotional abilities.

Empathy, which is the ability to understand and respond to another person's situation or feelings, is central to moral development. Empathy can be divided into cognitive empathy (understanding someone else's mental state) and affective empathy (experiencing an appropriate emotional response to another's emotions). While individuals with ASD may show lower levels of cognitive empathy, their affective empathy might be maintained or even heightened. This concept is sometimes referred to as the "empathy imbalance hypothesis."

Previous studies on moral dilemmas in adults and adolescents with ASD have yielded varied results. In impersonal dilemmas, adults with ASD tend to make similar utilitarian choices as adults without ASD. However, in personal, highly emotional dilemmas, findings are less consistent; some studies suggest that adults and adolescents with ASD might be more likely to make utilitarian choices than those without ASD. These findings highlight the importance of studying moral reasoning development across a wider age range in individuals with ASD.

This study investigates moral decisions in children and adolescents with ASD using the impersonal switch dilemma and the personal footbridge dilemma. The research aims to understand how emotion affects these choices across different ages, measure individuals' emotional arousal and how appropriate they found their decisions, and examine how specific ASD symptoms relate to moral choices. The researchers predicted that utilitarian decisions would decrease less with age for individuals with ASD, particularly in the personal dilemma. They also expected individuals with ASD to feel more emotional but rate their utilitarian decisions as more appropriate than individuals without ASD. Finally, a link was predicted between social interaction difficulties and utilitarian judgments in personal moral dilemmas.

Method

Participants

The study included 127 children and adolescents between 6 and 18 years of age. This sample consisted of 66 individuals with ASD and 61 children without ASD (referred to as typically developing, or TD). The ASD group included patients diagnosed with ASD by clinicians, with the diagnosis further confirmed by interviewing parents using a standard diagnostic tool. Individuals were excluded from the study if they had other psychiatric diagnoses, a low verbal IQ, neurological problems, or took medications that could affect thinking. Parents of individuals without ASD reported no psychiatric history. Both groups were similar in terms of age, gender, and general intelligence.

Procedure

After parents provided informed consent, they completed questionnaires about their children's behavior and, for the ASD group, participated in an interview to confirm diagnosis. Children were then assessed individually in a quiet room. Each child was presented with a non-moral dilemma (as a control) and then the two moral dilemmas (the trolley dilemma and the footbridge dilemma) in a random order. For each moral dilemma, participants had to choose whether to harm one person to save five others. After making their decision, participants used a 6-point scale with pictures (pictograms) to report how appropriate they thought their decision was and how calm they felt about it.

Data Analysis

Statistical models were used to analyze the results. These models compared how often each group made utilitarian choices across the different dilemmas and how these choices changed with age. They also examined how appropriateness ratings and calmness levels differed between the groups after a utilitarian decision. Special statistical methods were used to determine if observed effects were significant, meaning they were unlikely to have happened by chance.

Results

Response to the Dilemmas

The study found that the likelihood of making a utilitarian choice generally decreased as individuals got older. However, this pattern varied significantly between the ASD and TD groups, especially in the personal dilemma. For individuals without ASD, the probability of making a utilitarian choice in the personal dilemma sharply decreased with increasing age. In contrast, for individuals with ASD, this decrease in utilitarian choices in the personal dilemma was much less pronounced across the age range studied.

Appropriateness

When individuals with ASD made a utilitarian choice in the personal dilemma, they rated their decision as more appropriate compared to individuals without ASD who made the same choice.

Calmness

After making a utilitarian choice in the personal dilemma, individuals with ASD reported feeling less calm than individuals without ASD.

ADI‑R Scores and Moral Dilemmas

In the ASD group, a higher score on social interaction difficulties was linked to a greater chance of making a utilitarian choice in the personal moral dilemma.

Discussion

This study found that the likelihood of making a utilitarian choice generally decreased with age, but this pattern was significantly different for individuals with Autism Spectrum Disorder (ASD), especially in emotionally challenging situations. For those without ASD, utilitarian choices sharply declined with age in personal dilemmas, but this decrease was much less pronounced for individuals with ASD. This suggests that moral judgment development in individuals with ASD is different from around age 8 onwards, possibly because they use more logical, less emotion-driven decision-making, which is known as the "enhanced rationality" hypothesis.

The study also looked at how individuals felt about their decisions. When individuals with ASD made a utilitarian choice in an emotional dilemma, they rated it as more appropriate, but they also reported feeling less calm than those without ASD. This suggests a conflict: while their logical assessment found the choice suitable, their emotional response was one of distress. This finding supports the idea that individuals with ASD may have strong emotional reactions to others' distress (heightened affective empathy) but a different way of understanding others' thoughts and feelings (decreased cognitive empathy). It is possible that individuals with ASD choose a logical, utilitarian path as a way to manage or avoid overwhelming emotional situations, viewing it as the "more appropriate" choice despite their internal discomfort.

A notable finding was the link between social interaction difficulties and moral decisions in individuals with ASD. Those who reported more challenges in social interactions were more likely to make utilitarian choices in emotional dilemmas. This supports previous research that has found a connection between social cognitive difficulties and utilitarian judgments in adults with ASD, extending these observations to younger individuals.

A key strength of this study is its inclusion of a wide range of ages, from childhood to late adolescence, which helps provide insights into the development of moral judgments in individuals with ASD. The use of pictograms was also important for accurately measuring emotional and cognitive states, helping to overcome challenges individuals with ASD might face in describing their feelings verbally. However, the study has limitations; it looked at a single point in time rather than following individuals over many years, and it focused on only two specific moral dilemmas.

In conclusion, the study shows that moral decision-making, especially when emotions are involved, develops differently in individuals with ASD compared to others. Individuals with ASD may exhibit a blend of strong emotional responses to situations and a logical approach to problems, leading them to choose actions they deem "for the greater good" despite personal distress. These findings highlight the importance of understanding this unique development. They suggest that therapeutic approaches for children with ASD should focus on improving social skills, developing cognitive empathy, and providing strategies to manage intense emotional situations.

Introduction

People with Autism Spectrum Disorder (ASD) often have specific ways of thinking and acting. They might also have different ways of communicating and interacting with others. Because of these differences, it can sometimes be hard for them to make friends or get along with people. This can affect their daily life and how happy they feel.

Some researchers believe that people with ASD may think about right and wrong differently than others. For example, some studies suggest that people with ASD might make choices based more on logic than on feelings, even when faced with difficult situations. However, most past research looked at adults with ASD. We don't know much about how their ideas about right and wrong develop from when they are children to when they are teenagers.

This study looked at how feelings affect choices about right and wrong for people with ASD from childhood to their late teen years. The goal was to better understand how their moral thinking grows over time.

To study how feelings affect moral choices, researchers often use "dilemmas" or difficult stories. Two common stories are the "trolley dilemma" and the "footbridge dilemma." In the trolley dilemma, a train is headed for five people. A person can pull a lever to make the train go onto another track where it will hit only one person. In the footbridge dilemma, a person is on a bridge above a track where a train is headed for five people. The only way to stop the train is to push a large person off the bridge.

In both stories, the choice is to save five lives by causing one death. However, pulling a lever feels less emotional than pushing someone. Most people without ASD would pull the lever to save five. But in the footbridge dilemma, most people without ASD would not push the person, because it feels too emotional. They would rather let the five people die than push someone to their death.

Method

Participants

The study included 127 children and teenagers between 6 and 18 years old. This group had 66 patients with ASD from a mental health clinic and 61 children without ASD (called the "TD" group) from two schools. The two groups were similar in age and intelligence. All the people in the ASD group had been diagnosed with ASD by doctors.

To make sure the ASD diagnosis was correct, a trained expert talked to the parents. This expert asked about how the children interacted with others, their communication, and any set ways of behaving or interests. Children were not included if they had other mental health problems, a low IQ, a brain problem, or were taking medicines that could affect their thinking.

Parents of all children also filled out a checklist about their child's behavior. This helped to make sure that children without ASD did not have other serious problems.

Procedure

Parents first signed a form saying their child could be in the study. Then, they filled out some forms about their child. Children met individually in a quiet room. They were given three stories to consider: a control story (about taking a train or bus) and the two moral dilemmas (the trolley and footbridge stories). The order of the stories was mixed up for each child.

After hearing each story, children had to decide what they would do. For the moral stories, this meant choosing whether to harm one person to save five. After each choice, children answered two questions using a simple picture scale. They rated how "right" their decision felt and how "calm" they felt about their decision.

Data Analysis

Researchers used special computer programs to look at the choices made by the children. They checked how age, group (ASD or TD), and the type of story (personal, impersonal, or neutral) affected decisions. They also looked at how the groups rated their choices for how "right" they felt and how "calm" they were. Finally, for the ASD group, they checked if there was a link between their social interaction scores and their choices in the stories.

Results

Response to the Dilemmas

At younger ages, both groups of children (ASD and TD) often chose to save the most lives, no matter which story they heard. The main differences appeared in the "personal" dilemma (the footbridge story) as children got older. In the TD group, as children grew older, they became much less likely to choose to push the person off the bridge. This was not as true for children with ASD. For them, the choice to push the person did not go down as much with age.

Appropriateness

When children made the "save five by harming one" choice in the personal dilemma, those with ASD thought their decision was more "right" than the TD children did.

Calmness

However, after making the "save five by harming one" choice in the personal dilemma, children with ASD felt less calm than the TD children.

ADI‑R Scores and Moral Dilemmas

For the children with ASD, the researchers found a link between their social interaction scores and their choices. Children with ASD who had more trouble with social interaction were more likely to make the "save five by harming one" choice in the personal dilemma.

Discussion

This study looked at how children and teenagers with ASD make choices about right and wrong compared to those without ASD. The study found that as children get older, they generally choose to save the most lives less often. However, this change was different for children with ASD, especially in stories where feelings were strongly involved.

For children without ASD, the choice to save five by harming one in the emotional footbridge story dropped quickly as they got older. But for children with ASD, this choice did not drop as much with age. Even at 18 years old, people with ASD were more likely to make this choice than those without ASD. This suggests that the way children with ASD think about right and wrong develops differently and might be delayed compared to others. This fits with the idea that people with ASD might make choices based more on logic, trying to save the most lives, even if it feels wrong.

Another interesting finding was how people felt about their choices. After making the "save five by harming one" decision in the emotional story, children with ASD felt that their decision was more "right" but also felt less "calm." This suggests that people with ASD might be very sensitive to others' feelings (leading to less calmness), but they might also have a different way of understanding what is socially right or wrong (leading them to think their decision was "right"). Using pictures to ask about their feelings helped get clearer answers.

This study also found that children and teenagers with ASD who had more trouble interacting with others were more likely to make the "save five by harming one" choice in the emotional story. This shows that how people with ASD interact with others may be connected to how they make moral decisions.

This study is important because it looked at how moral choices develop in people with ASD from childhood to their late teens. It also measured their feelings and thoughts about their choices in a new way. However, the study looked at different age groups at one time, rather than following the same children over many years. Also, only two moral stories were used. More research with different stories could give an even wider view.

Overall, this study shows that people with ASD think about right and wrong in a unique way as they grow up. They may make choices that seem more logical, even if those choices cause them more emotional upset. This suggests that helpers should focus on improving social skills for children with ASD and helping them manage strong feelings in difficult situations.