Social Decision Making in Adolescents and Young Adults: Evidence From the Ultimatum Game and Cognitive Biases

Abstract

During adolescence and early adulthood, individuals deal with important developmental changes, especially in the context of complex social interactions. Previous studies demonstrated that those changes have a significant impact on the social decision making process, in terms of a progressive increase of intentionality comprehension of others, of the sensitivity to fairness, and of the impermeability to decisional biases. However, neither adolescents nor adults reach the ideal level of maximization and of rationality of the homo economicus proposed by classical economics theory, thus remaining more close to the model of the “bounded rationality” proposed by cognitive psychology. In the present study, we analyzed two aspects of decision making in 110 participants from early adolescence to young adulthood: the sensitivity to fairness and the permeability to decisional biases (Outcome Bias and Hindsight Bias). To address these questions, we adopted a modified version of the Ultimatum Game task, where participants faced fair, unfair, and hyperfair offers from proposers described as generous, selfish, or neutral. We also administered two behavioral tasks testing the influence of the Outcome Bias and of the Hindsight Bias in the evaluation of the decision. Our behavioral results highlighted that the participants are still partially consequentialist, as the decisional process is influenced by a complex balance between the outcome and the psychological description of the proposer. As regards cognitive biases, the Outcome Bias and the Hindsight Bias are present in the whole sample, with no relevant age differences.

Keywords

Social decision making Theory of Mind adolescence fairness Ultimatum Game Outcome Bias Hindsight Bias

Introduction

The adaptive ability that allows people to select an optimal course of action from several alternatives in highly complex social environments is called social decision making (Sanfey, 2007). Behavioral studies conducted within the framework of the Economic Game Theory with bargaining games (Camerer, 2003) have demonstrated that people’s decisions are not exclusively oriented to the maximization of self-gain, but that they also take into account the related outcomes for others, especially in those settings characterized by cooperative exchanges and reciprocity (Camerer & Fehr, 2006). Neuroeconomics research focused on the behavioral and neural components of social decision making has highlighted that, during social interactions, adults are not merely motivated by self-interest, but also by self versus other comparisons (Fehr & Fischbacher, 2003). This is particularly evident with the Ultimatum Game (UG) task (Güth, Schmittberger, & Schwarze, 1982), where one player (responder) receives a monetary offer from another player (proposer), who has to divide the amount. The responder has the option of accepting or rejecting the offer. If the responder accepts the offer, the sum of money will be really split between the two participants; on the contrary, if the responder rejects the offer, both players will receive nothing. If individuals were motivated simply by self-interest, according to the self-gain maximization principle, it would be expected that responders accept every type of offer, while the proposer will offer the smallest possible amount. However, offers of less than 20% of the total amount are rejected about half of the time, showing that when the division of the stake is unfair, responders reject it, preferring that both players get nothing (Camerer, 2003). The consistency of such results in western industrialized countries has suggested that a key psychological component of social decision making is the sensitivity toward fairness, conceived as “strong reciprocity” by Rabin (1993), as “inequity aversion” by Fehr and Schmidt (1999), or as the sensitivity toward the violation of a social norm by Bicchieri (2006) and Bicchieri and Chavez (2010).

From this brief overview, it seems well established that the effective behavior of adult individuals in social decision making does not confirm the self-gain maximization principle, but rather it is the result of many psychological components of the decisional process in an interactive situation. The first component, the sensitivity to fairness, is a key factor in the interpersonal context because it arises when people make a subjective comparison between self-interest and needs or preferences of others. Such a comparative component implies the integration of context-related information and of social–emotional aspects in the decisional process, including also the evaluation of the partner’s intentions, which consists in inferences about the mental state of the partner (Güroğlu, Van Den Bos, Rombouts, & Crone, 2010; Sanfey, Rilling, Aronson, Nystrom, & Cohen, 2003).

As a matter of fact, in real-life decision making, people have often to cope with situations that either present both objective information (for instance, monetary offers) and subjective information (such as the features of the partner of the interaction) or often a dissonance between them. Literature about social decision making has provided good evidences about framing effects in social decision making concerning either the offers or the partners. In this sense, Solnick and Schweitzer (1999) investigated the role of physical appearance in social exchange, and found that participants usually make higher offers to more attractive players, but that at the same time physically attractive participants are expected to offer more. As far as psychological features are concerned, Sutter (2007) showed that the responder may focus not only on the outcome, but also on the intention of the proposer. He found that adults mainly focus on the intention, thus explaining the high rates of rejections of unfair offers, whereas younger participants (children and adolescents) mainly focus on the outcome, thus explaining the higher rates of acceptances of unfair offers in these populations. Delgado et al. (2005) showed that knowledge about the personality of one’s partner in a Trust Game can bias decision making substantially. In this study, the researchers provided a vignette about each partner, which broadly described them as either a “good” person (i.e., someone who in the past has rescued victims from a burning building) or a “bad” person (i.e., someone who has engaged in moderately unethical behavior). Results showed that the mental representations about the degree of morality of the partners affected decision making, and more money was entrusted to partners who had received a “good” moral designation. Yet a different type of psychological framing relates to social roles. For example, Burnham et al. (2000) showed that the label of the other as either “partner” or “opponent” affects players’ behavior in the Trust Game (Berg et al., 1995), and Hoffman et al. (1994) found that if players are assigned as “sellers” (i.e., proposers) or “buyers” (i.e., responders) based on performance in a general knowledge quiz before playing the UG, proposers tend to make more selfish offers and responders tend to accept them. Finally, Marchetti, Castelli, Harlé, and Sanfey (2011) showed that different descriptions of the proposer on the “psychological level” (John is a generous person, who takes into account the needs of other people…), on the “physical level” (John is a tall man with dark eyes and hair…), or the absence of any description combined with fair/unfair offers generate a framing effect that affects the final decision.

Recent works have focused on the development of the sensitivity to fairness, suggesting that children’s willingness and ability to distribute goods in a fair and equal manner emerge in infancy, around the second year of life (Geraci & Surian, 2011; Schmidt & Sommerville, 2011; Sloane, Baillargeon, & Premack, 2012). However, children under the age of seven tend to prefer distribution in their favor (Overgaauw, Güroğlu, & Crone, 2012), while an inequitable distribution of resources becomes aversive from seven to eight years old (Güroğlu, Van Den Bos, Van Dijk, Rombouts, & Crone, 2011). Thus, the judgment of fairness undergoes significant changes during late childhood and early adolescence (Güroglu, Van Den Bos, & Crone, 2009): the preference for fair distribution increases with age, and it is partially related to the acquisition of perspective-taking abilities (Castelli, Massaro, Sanfey, & Marchetti, 2010, 2014; Fehr, Bernhard, & Rockenbach, 2008; Güroglu et al., 2009; Overgaauw et al., 2012; Sutter, 2007). Relevant developmental changes have been observed also for the strategic decision making process. For example, Steinbeis, Bernhardt, and Singer (2012) claim that there is an age-related increase in strategic decision making, combined with strategic reasoning. Based on what has been said so far, decision making is outlined as a complex process during which relational aspects play an important role. At the same time, in a development perspective, very young children—though showing a premature sensitivity to fainess—appear to be consequentialists, that is, they pay much more attention to the material consequences of the decision. Only over the years, they acquire a more articulated attitude that—while not completely forgetting the outcome of the decision—integrates more psychological aspects. This strong propensity for outcomes in social decision making may represent the same mechanism that induces people to be “captives of knowledge”: when a knowledge is gained, it will contaminate our decision making abilities, preventing us from reasoning properly on the prospect of a less-informed person (Birch, Brosseau-Liard, Haddock, & Ghrear, 2017).

Shifting attention to adults, they use complex reasoning during decision making more than children (Huizenga, Crone, & Jansen, 2007), and they are less prone to cognitive biases than children, specifically to the Outcome Bias and to the Hindsight Bias, two different expressions of the so-called “curse of knowledge” (Massaro, Castelli, Sanvito, & Marchetti, 2014). The “curse of knowledge” is defined as “the tendency to be biased by one’s own knowledge when attempting to appreciate a more naïve or uninformed perspective” (Birch & Bloom, 2004, p. 255). Particularly, the Outcome Bias evaluates the tendency to judge a past decision on the basis of its outcome. In other words, resisting to this bias means to judge the outcome of such past choice as not relevant for the judgment of the choice itself, and to consider only the premises as essential for this judgment. However, decisions with negative outcomes are usually undervalued, while decisions with positive outcomes are usually overvalued (Baron & Hershey, 1988). The Hindsight Bias represents the tendency to view events that have occurred as more predictable than they were before they took place (Fischhoff, 1975). It involves the inability to recapture the feeling of uncertainty that preceded an event, providing the illusion of understanding the past. For these reasons, it is often termed “knew-it-all-along” effect (Roese & Vohs, 2012). There is some evidence for a developmental decline in cognitive biases (Massaro et al., 2014), especially in Hindsight Bias from childhood to adulthood (Pohl, Bayen, & Martin, 2010), but the topic is still controversial (Bernstein, Erdfelder, Meltzoff, Peria, & Loftus, 2011).

Adolescence is usually defined as a transition phase between childhood and adulthood, which involves many physical, cognitive, socio-emotional, and contextual changes (Steinberg & Morris, 2001). In this period, the nature of social interaction changes qualitatively, becoming more complex and more oriented outside the family toward peer relationships (Crone & Dahl, 2012; Zanolie, De Cremer, Güroğlu, & Crone, 2015). Those developmental changes have a significant impact on cognitive and affective components of the decisional process, especially in the context of social interaction, which also depends on the concurrent choices of others. By examining the maturity of teenagers’ judgment on three psychosocial factors—responsibility, temperament, and perspective—Steinberg and Cauffman (1996) emphasize the need for studies that take into account both cognitive and noncognitive aspects. From a neurocognitive point of view, the brain of teenagers does not seem to have reached a level of maturity comparable to that of adults. Nonlinear changes in cortical gray matter seem to affect specific areas of the brain—crucial for decision making—between 12 and 16 years of age, making the adolescence a potential significant benchmark for decision making abilities. Although different from that of adults, the brains and the decision making capabilities of teenagers are not necessarily limited; rather, they are unique, that is, expression of a complex and delicate phase of development and thus characterized by operating mechanisms that experience the integrated evaluation of different elements of decision making such as, for example, intertemporal choice, prospective evaluation, and integration of positive and negative feedback (Hartley & Somerville, 2015).

Indeed, the transition from childhood to adolescence seems to be characterized by the development of more and more efficient and normatively oriented reasoning skills, as well as by the appearance of decision making strategies contaminated by biases of reasoning (Jacobs & Klaczynski, 2002). On the other hand, limited to the Hindsight Bias, Bernstein et al. (2011) have highlighted how this kind of bias evolves throughout life and how children and the elderly are most likely to be affected by it. Finally, the study of decision making abilities in this critical phase of the life becomes even more relevant if one takes into account the fact that adolescents are especially concerned with implementing risk behaviors and that a better understanding of the factors involved might also have positive effects on social policies (Reyna & Farley, 2006).

Purpose of the study

On the basis of the previous overview, the study of social decision making and its potential correlations with Hindsight Bias and Outcome Bias across adolescence and early adulthood may provide a further understanding of how decision making abilities in social situations evolve and consolidate (Crone & Van Der Molen, 2004; Crone, Bunge, Latenstein, & Van Der Molen, 2005; Güroglu et al., 2009; Van Duijvenvoorde, Peters, Braams, & Crone, 2016; Will, Crone, Van Lier, & Güroğlu, 2016). The aims of this study are: (a) to analyze the effect of fairness sensitivity on the interplay between psychological and material aspects of the decisional process, and (b) to explore the permeability to cognitive biases in the evaluation of a decision. Participants faced different type of offers (from hyperfair to unfair) from different descriptions of the proposer (a psychological description as selfish or generous, and a neutral one), which can influence the sensitivity toward fairness and the final decision. Furthermore, they were tested for the decisional biases through the Outcome Bias and the Hindsight Bias. We explore the possibility that the type of psychological description of the proposer may influence the decision combined with the fairness/unfairness/hyperfairness of the offer, and that the cognitive biases may decrease with advancing ages.

Method

Participants

A total of 110 students participated in this study. According to the Italian school system, they were divided into three groups: (a) 47 students from 14 to 16 years of age, attending the first three (3) years of High School (3HS group); (b) 42 students from 17 to 18 years of age, attending the last two (2) years of High School (2HS group); (c) 21 undergraduate students (UGS group) from 19 years of age and above, attending the Catholic University of the Sacred Heart in Milano. The age groups were sex-matched and none of the participants reported any learning or social problems. Demographics are illustrated in Table 1.

Table 1.

Demographic characteristics of the sample.

	3HS	2HS	UGS	Group comparison
N	47	42	21
Age
Mean ± SD	14.90 ± 0.47	17.79 ± 0.34	22.62 ± 1.87	<0.001
Range	14.24–16.09	17.24–18.87	19.06–26.43
Gender
M/F	25/22	25/17	10/11

A written informed consent and the informed parental consent for the adolescents were obtained from each participant. The study was approved by the local ethical committee, according to APA ethical standards.

Procedure

The tasks were administered at the High School for the 3HS group and the 2HS group, and at the University for the UGS groups. All age groups were screened for executive function (Clock Test) in order to have homogeneous samples under the cognitive profile. In addition, all participants were tested for fairness sensitivity (UG task as responder), and for decisional biases (Outcome Bias and Hindsight Bias).

Executive function task. Executive function was assessed with the Clock Test (Fabio, Antonietti, & Pravettoni, 2008; Moron, 1997), a visual-attention task evaluating the automation process and the return to voluntary control through the presentation of four tables. A total of 400 watches are showed on each table, among which there are 40 targets. In the first, second, and third table the participants had to identify all the clocks showing 04:00. Thus, the activation of an automation process is expected; in the last table, the participants had to identify all the clocks that show 05:00, which requires a return to voluntary control. The time demanded to complete the task was two minutes for each table. To evaluate the subject’s performance, a number of indexes were calculated: the total of uncorrected answers and omissions (Inaccuracy Index, range 0/160), the Automation Index (range −1/+1) and the Rigidity Index (range −1/+1). The basic idea is that the automation process develops with exercise and involves an increase in the executive accuracy (with a decrease in the number of errors) and an increase in speed of execution (Larson & Saccuzzo, 1986; Szymura & Nęcka, 1998). On the other hand, high automation index involves high rigidity in coding, so a change of target may result in a drop in speed and accuracy (Szymura, Slabosz, & Orzechowski, 2001).

Social decision making task. All participants played a modified version of UG. According to the original version, the UG is a simple game that examines fairness and more strategic thinking in the context of two-player bargaining. In the UG, the proposer and responder are asked to divide a sum of money, with the proposer specifying how this sum should be divided between the two. The responder has the option of accepting or rejecting the offer. If the offer is accepted, the sum is divided as proposed. If it is rejected, neither player receives anything. In either event the game is over, that is, there are no subsequent rounds in which to reach agreement. The decision to reject an unfair offer may be considered a form of altruistic punishment because the responder chooses to receive no money rather than the amount offered by the proposer, presumably to punish the proposer for making a miserly offer. If people are motivated purely by self-interest, the responder should accept any offer and, knowing this, the proposer will offer the smallest nonzero amount. All participants played in the role of responder facing all the possible sharing offers between 1–9 and 9–1 from an anonymous partner (proposer) described as generous, selfish, or neutral. Each participant received three offers for each typology of sharing, for a total of 27 offers. The innovative aspects of this procedure concern the description of the proposer with a “psychological label”, so that each offer was preceded by the initial of the name of the proposer and a written label describing the proposer as generous, selfish, or neutral. This type of manipulation of the proposer description was derived from a previous work by Marchetti et al. (2011), showing that the label generates a framing effect that affects the final decision along with the sensitivity toward fairness. The UG was played for real, so, in the case of acceptance, participants received the money converted into gift card.

Cognitive biases. All participants completed the Outcome Bias task (Baron & Hershey, 1988) and the Hindsight Bias task (Fischhoff, 1975). The Outcome Bias task is defined as the tendency to judge a past decision based on its outcome. Participants read two stories about an opportunity for a man to undergo a surgical operation, one with a good outcome (the patient survives) and one with a bad outcome (the patient dies). They had to answer two questions about the goodness of the decision of both the patient and the doctor on a Likert scale ranging from −3 to 3. The magnitude of the bias is the difference between the two judgment conditions (successful vs. unsuccessful), both for the doctor and for the patient. The Hindsight Bias task evaluates the tendency, after an event has occurred, to consider such event as more predictable, despite having little or no objective basis for predicting it. Participants were asked to read a story about a war scenario. At the end of the story, four possible outcomes were provided, and participants had to estimate the probability of occurrence of each outcome (the sum of the four probabilities should be 100). Then, the real outcome was provided. At this point, participants were asked to estimate again the probability of occurrence of each of the possible outcomes, that is, to review their previous estimation, and then to state what probability other participants would assign to the four outcomes. The more participants revise their previous estimations on the basis of the new information (real outcome), the more they are affected by the bias. The bias is greater if the knowledge of the outcome misleads also the estimate attributed to another person. In addition to the three estimates, the subject was asked whether, in the light of the outcome: (a) he/she believed that the initial estimate should be revised (coded as dichotomous) (HB2); (b) he/she believed that the initial estimate was a good one (seven points Likert scale) (HB4).

Statistical analyses

Behavioral data were analyzed with Statistical Package for Social Sciences (SPSS), version 21.0. If not differently specified, we conducted an analysis of variance for the variables with normal distribution, and nonparametric analyses (Kruskal–Wallis H. with post-hoc correction for multiple comparisons) for the variables with not normal distribution.

Results

We found no significant differences in the Automation Index (p = 0.20) and in the Rigidity Index (p = 0.24) of the Clock task between groups, so these abilities remain constant across the ages. Results are showed in detail in Table 2.

Table 2.

Performances to the Clock task.

	3HS	2HS	UGS	p	F	η²p	θ
	Mean ± SD	Mean ± SD	Mean ± SD	p	F	η²p	θ
Clock Test
Correct responses (total/160)	122.26 ± 18.55	127.88 ± 18.81	125.50 ± 18.15	0.36	1.04 (2 128)
Inaccuracy (total)	50.66 ± 21.99	43.52 ± 19.75	48.52 ± 22.63	0.29	1.27 (2 128)
Automation index	0.22 ± 0.30	0.13 ± 0.37	0.08 ± 0.22	0.08	2.58 (2 128)
Rigidity index	− 0.11 ± 0.22	−0.17 ± 0.26	−0.02 ± 0.31	0.24 (2HS < UGS)	1.44 (2 107)

As regards the UG, data were analyzed using a mixed-design analysis of variance with two three-levels within-subjects factors (offer type × proposer’s description), one three-levels between-subject factor (age), and one two-levels between-subject factor (gender). Mauchly’s test indicated that the assumption of sphericity had been violated (χ²₍₂₎ = 21.9, p < .001). Since Greenhouse–Geisser epsilon was higher than .75, degrees of freedom were corrected using Huynh–Feldt estimates of sphericity (ɛ = 0.89) (Field, 2013). Only a main effect of offer type was found (F_{(2, 208)} = 270.39 p < .001, ηp²= .733, θ = 1.0). Pairwise comparisons using the Bonferroni correction revealed that the acceptance rates were different according to the type of offer, decreasing significantly from hyperfair to fair to unfair offers (7.86 ± .20 6.92 ± .16 1.88 ± .22, all p values ≤ .001).

A post-hoc power analysis was conducted using the software package, G*Power 3.1 (Faul, Erdfelder, Lang, & Buchner, 2007) revealing that the sample size may have played a role in limiting the significance of within and between factors interactions. In fact, analyses showed that for the largest effect size F observed (.20) in the present study for this kind of interactions (e.g., offer type × age × gender), an N of approximately 420 would be needed to obtain statistical power at the recommended .80 level (Cohen, 1988).

Descriptive statistics of the acceptance rate are showed in the Table 3.

Table 3.

Mean and standard deviation of the accepted responses in the UG.

Condition	Label	Group	Mean	SD
hyperfair	Neutral	3HS	80,000	186,501
		2HS	80,000	253,752
		UGS	80,952	184,132
		Total	80,182	212,448
	Generous	3HS	76,170	250,697
		2HS	80,000	238,900
		UGS	78,095	267,617
		Total	78,000	247,853
	Selfish	3HS	74,043	241,066
		2HS	80,952	223,944
		UGS	78,095	231,558
		Total	77,455	232,827
Fair	Neutral	3HS	64,255	163,846
		2HS	73,810	195,001
		UGS	77,619	167,047
		Total	70,455	183,936
	Generous	3HS	63,617	200,462
		2HS	67,381	249,936
		UGS	74,762	172,102
		Total	67,182	218,053
	Selfish	3HS	62,340	204,528
		2HS	74,286	182,320
		UGS	78,095	193,956
		Total	69,909	203,860
Unfair	Neutral	3HS	10,000	191,107
		2HS	20,238	280,668
		UGS	25,238	321,899
		Total	16,818	260,533
	Generous	3HS	20,851	299,877
		2HS	17,381	283,755
		UGS	22,381	282,674
		Total	19,818	288,616
	Selfish	3HS	15,957	249,050
		2HS	23,333	326,599
		UGS	20,952	207,135
		Total	19,727	274,433

In order to better understand the decision behavior as a function of offer and of proposer profile, given the nature of variables distribution, we carried out a series of nonparametric analyses (correcting for multiple comparisons) on individual offers and comparing the three profiles. Regarding the effect of the proposer’s description on the acceptance rate compared to the corresponding neutral condition, our analyses show that offers from neutral proposer were more accepted than offers from a generous proposer when the offer of sharing was 5–5 (p = .027); the neutral proposer offers gained a greater acceptance rate compared to those of a selfish proposer when the offer of sharing was 1–9 (p = 0.01).

As regards the cognitive biases, our results showed the existence of the Outcome Bias across all age groups. In particular, this bias is observed both in the evaluation of the medical perspective (Z = 5.46 p < 0.001) and in the evaluation of the patient’s perspective (Z = 4.91 p < 0.001). No differences were found in the extent of the Outcome Bias between age groups. However, we found a significant difference in the doctor’s decision evaluation item, that is, the evaluation of the goodness of the doctor’s decision: in the unsuccessful condition (X²= 12.99 p = 0.002) young adults (UGS group) showed higher scores compared to adolescents (2HS group, p = 0.002) and young adolescents (3HS, p = 0.005), and in the successful condition (X²= 6.33 p = 0.042) young adults (UGS group) showed higher scores compared to young adolescents (3HS group, p = 0.039). Results are showed in detail in Table 4. Our results demonstrated also the existence of the Hindsight Bias across all age groups (Friedman test = 54.11 p < 0.001). More specifically, this bias is present only when the subject, who now knows the outcome, is invited to revise her/his previous predictions (p < 0.001); it does not appear when the subject must assume the predictions of a third person still unaware of the outcome (p = 0.86). The amplitude of the bias does not change among the three age groups (p = 0.47). Considering more specifically the task questions, significant evidences emerged in the HB2 (X²= 8.31 p = 0.016), and HB4 (X²= 7.67 p = 0.022). In the first case, adolescents (2HS group) tend to be more frequent among those who believe that the initial estimate should not be revised unlike young adolescents (3HS group) and young adults (UGS group), that are distributed fairly; in the other case—consistent with the result just described—adolescents (2HS group) evaluate the initial estimate significantly better than the young adolescents (3HS group) and showed higher scores compared to the other two groups (3HS and UGS groups). Results are showed in detail in Table 4.

Table 4.

Performaces to the Outcome Bias task and to the Hindsight Bias task.

	3HS	2HS	UGS	Group comparison (p)	X ²	Multiple comparison
	Rank med/ Frequency (%)	Rank med/ Frequency (%)	Rank med/ Frequency (%)	Group comparison (p)	X ²	Multiple comparison
Outcome Bias
Unsuccessful
DDE^a	60.83	57.46	77.32	0.029	7.09	(UGS > 2HS)
Successful
DDE^a	58.11	62.69	76.85	0.043	6.28	(UGS > 3HS)
Hindsight Bias
HB3C–HB1C^a	61.64	78.30	58.58	0.023	7.61	(2HS > 3HS)
HB2(%)	56/44	27/73	54/46	0.02	8.69	(2HS > 3HS, UGS)
HB4^a	57.87	76.04	55.22	0.016	8.24	(2HS > 3HS, UGS)

DDE: doctor’s decision evaluation and chi-square analyses for nominal variables (the frequency are expressed in percentage).

Kruskal–Wallis analyses for scale variables.

Discussion and conclusions

The present study aimed to investigate in young adolescents, adolescents, and young adults: (a) the social decision making in the UG facing different type of offers combined with different descriptions of the proposers (b) the permeability to cognitive biases in the evaluation of a decision. Our behavioral data highlight that in the UG, the acceptance rate varied with the type of offer. Specifically, all participants were obviously more prone to accept hyperfair offers and to reject unfair ones, as typically happens in UG experiments, where the acceptance rates decrease as the offers became less fair (Sanfey et al., 2003). Regarding the effect of the psychological labels, our results showed that the fair offer (5–5) and the hyperfair one (1–9) from neutral participants were more accepted than offers from proposers described as generous and selfish respectively. Our findings highlight that when people face a conflict between “hot (psychological labels)” and “cold (amount of the offer)” aspects of decision making, they lower the acceptance rating. In fact, from a proposer described as selfish, it is likely that people do not expect to receive a so incoherent—because incredibly high—offer of sharing; conversely, from a proposer described as generous, it is likely that people hope to receive much more than a perfectly equal distribution of the good. So, a neutral proposer seems to “neutralize” one of the two conflicting variables. Probably, the same 5–5 offer is accepted more from a neutral proposer than from a generous proposer because perfect equality constitutes the exemplary division. This aspect clearly emerges also from developmental studies, where children move from an initial conception of fairness as their own advantage, to a conception of fairness as perfect equality (see, e.g., Castelli et al., 2014). The same mechanism should work also for hyperfair offers, which may appear less suspicious when they come from a neutral proposer compared to a selfish one.

Regarding the cognitive bias, the Outcome Bias and the Hindsight Bias are present in the whole sample, with no relevant age differences. In particular, in the Outcome Bias, the doctor’s decision is better evaluated by young adults compared to young adolescents and adolescents, respectively in the successful and in the unsuccessful conditions. It probably means that young adults identify with the figure of the doctor more than adolescents both when the outcome is good and when the outcome is bad, and that conversely this mental simulation is not equally feasible when the task requires to assume the perspective of the patient, whose expertise leading to the decision is not as scientifically based as the medical one. Probably, when participants are required to assume the perspective of the patient, they are all exposed to the bias, as no one likes to wear the patient’s shoes. In the Hindsight Bias, we observed that the adolescents claim that their first estimation should not be changed, and that their first estimation was a good estimation compared to the other two groups. Notwithstanding this, their bias is similar to the bias of the other two groups. Altogether, these findings may show that in adolescence there is a weaker connection between choices and a reflective stance toward choices. These findings could be explained in the light of the developmental changes affecting social behavior in the transitional phase between childhood and adulthood. In fact, as mentioned in the Introduction section, the understanding of intentionality, strategic thinking, and complex reasoning emerge progressively during adolescence and early adulthood (Güroglu et al., 2009; Huizenga et al., 2007; Steinbeis et al., 2012). Additional interpretative instances can be found in the neuroscientific literature. Indeed functional and structural neuroimaging studies have shown a prolonged development of the Rostrolateral prefrontal cortex (RLPFC) during adolescence and the activity of this area is often associated with the development of cognitive functions including logical and relational reasoning (Dumontheil, 2014). However, it should also be considered that limitations in brain development seem to be poorly linked to impulsivity and poor decision making during adolescence (Romer, 2010).

Overall, our results may represent a further contribute to understanding the effect of a conflict between psychological and material aspects of the decision. The common crucial point between our results in the UG rounds and in the decision biases, especially in the Outcome Bias, is the protagonist of the decision, that is, the “who” of the decisional process. In the first case, it is the type of the profile of the person who makes the offer that matters: such a profile, combined with the entity of the offer, generates a conflict that weakens the salience of the outcome itself. In the second case, it is the person with whom the participant is more likely to identify with, that is, the doctor, which makes people more sensitive to the bias.

Altogether, our findings enrich the debate about human being’s decision making abilities, now assimilated to an extremely rational homo economicus and thus maximizing the result of decisions, now endowed with a bounded rationality and therefore inclined to consider other factors in decision making (e.g., emotions). Future research may profitably investigate also the neural correlates of decisional biases in order to wider the neural comprehension of decision making processes. To account for possible differences in decision making and permeability to the biases depending on age, future investigations should also modify and enrich contents and contexts of decisions themselves.

This study has some limitations that need to be carefully considered for the interpretation of results. We have used a convenient sample that reduces external validity. Race and ethnicity were not controlled. Moreover, although the analyses have been appropriately corrected, sample sizes are not homogeneous between groups. Concerning the psychological description of the UG proposer, the synthetic labels used may not have allowed a realistic contextualization of the relational interaction, reducing the possibility of a real effect. Future research will have to consider the use of more extensive psychological descriptions of the proposer, so as to increase their ecological validity.

Footnotes

Article Notes

References

Baron

Hershey

J. C.

(1988) Outcome bias in decision evaluation. Journal of Personality and Social Psychology 54(4): 569–579. doi:10.1037/0022-3514.54.4.569.

Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity and social history. Games and Economic Behavior, 10, 122–142.

Bernstein

D. M.

Erdfelder

Meltzoff

A. N.

Peria

Loftus

G. R.

(2011) Hindsight bias from 3 to 95 years of age. Journal of Experimental Psychology: Learning, Memory, and Cognition 37(2): 378–391. doi:10.1037/a0021971.

Bicchieri

(2006) The grammar of society: The nature and dynamics of social norms, Cambridge, UK: Cambridge University Press.

Bicchieri

Chavez

(2010) Behaving as expected: Public information and fairness norms. Journal of Behavioral Decision Making 23(2): 161–178. doi:10.1002/bdm.648.

Birch

S. A. J.

Bloom

(2004) Understanding children’s and adults’ limitations in mental state reasoning. Trends in Cognitive Sciences 8(6): 255–260. doi:10.1016/j.tics.2004.04.011.

Birch

S. A. J.

Brosseau-Liard

P. E.

Haddock

Ghrear

S. E.

(2017) A ‘curse of knowledge’ in the absence of knowledge? People misattribute fluency when judging how common knowledge is among their peers. Cognition 166: 447–458. doi:10.1016/j.cognition.2017.04.015.

Burnham, T., McCabe, K., & Smith, V. L. (2000). Friend-Or-Foe Intentionality Priming in an Extensive Form Trust Game. Journal of Economic Behavior and Organization, 43, 57–73.

Camerer

C. F.

(2003) Behavioral game theory: Experiments in strategic interaction, Princeton, NJ: Princeton University Press.

10.

Camerer

C. F.

Fehr

(2006) When does “economic man” dominate social behavior. Science 311(5757): 47–52. doi:10.1126/science.1110600.

11.

Castelli

Massaro

Sanfey

A. G.

Marchetti

(2010) Fairness and intentionality in children’s decision-making. International Review of Economics 57(3): 269–288. doi:10.1007/s12232-010-0101-x.

12.

Castelli

Massaro

Sanfey

A. G.

Marchetti

(2014) “What is fair for you?” Judgments and decisions about fairness and Theory of Mind. European Journal of Developmental Psychology 11(1): 49–62. doi:10.1080/17405629.2013.806264.

13.

Cohen

(1988) Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates.

14.

Crone

E. A.

Dahl

R. E.

(2012) Understanding adolescence as a period of social-affective engagement and goal flexibility. Nature Reviews Neuroscience 13(9): 636–650. doi:10.1038/nrn3313.

15.

Crone

E. A.

van der Molen

M. W.

(2004) Developmental changes in real life decision making: Performance on a gambling task previously shown to depend on the ventromedial prefrontal cortex. Developmental Neuropsychology 25(3): 251–279. doi:10.1207/s15326942dn2503_2.

16.

Crone

E. A.

Bunge

S. A.

Latenstein

van der Molen

M. W.

(2005) Characterization of children’s decision making: Sensitivity to punishment frequency, not task complexity. Child Neuropsychology 11(3): 245–263. doi:10.1080/092970490911261.

17.

Delgado, M. R., Frank, R. H., & Phelps, E. A. (2005). Perceptions of moral character modulate the neural system of reward during the trust game. Nature Neuroscience, 8(11), 611–1618.

18.

Dumontheil

(2014) Development of abstract thinking during childhood and adolescence: The role of rostrolateral prefrontal cortex. Developmental Cognitive Neuroscience 10: 57–76. doi:10.1016/j.dcn.2014.07.009.

19.

Fabio

R. A.

Antonietti

Pravettoni

(2008) Benefici e costi dei processi di automatizzazione dell’attenzione visiva [Benefits and costs of the processes of automation of visual attention]. Ricerche di Psicologia 30(3): 109–120.

20.

Faul

Erdfelder

Lang

A.-G.

Buchner

(2007) G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods 39(2): 175–191. doi:10.3758/BF03193146.

21.

Fehr

Fischbacher

(2003) The nature of human altruism. Nature 425(6960): 785–791. doi:10.1038/nature02043.

22.

Fehr

Schmidt

K. M.

(1999) A theory of fairness, competition, and cooperation. The Quarterly Journal of Economics 114(3): 817–868. doi:10.1162/003355399556151.

23.

Fehr

Bernhard

Rockenbach

(2008) Egalitarianism in young children. Nature 454(7208): 1079–1083. doi:10.1038/nature07155.

24.

Field

(2013) Discovering statistics using IBM SPSS statistics, 4th ed. London: Sage.

25.

Fischhoff

(1975) Hindsight is not equal to foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance 1(3): 288–299. doi:10.1037/0096-1523.1.3.288.

26.

Geraci

Surian

(2011) The developmental roots of fairness: Infants’ reactions to equal and unequal distributions of resources. Developmental Science 14(5): 1012–1020. doi:10.1111/j.1467-7687.2011.01048.x.

27.

Güroglu

van den Bos

Crone

E. A.

(2009) Fairness considerations: Increasing understanding of intentionality during adolescence. Journal of Experimental Child Psychology 104(4): 398–409. doi:10.1016/j.jecp.2009.07.002.

28.

Güroğlu

van den Bos

Rombouts

S. A.

Crone

E. A.

(2010) Unfair? It depends: Neural correlates of fairness in social context. Social Cognitive and Affective Neuroscience 5(4): 414–423. doi:10.1093/scan/nsq013.

29.

Güroğlu

van den Bos

van Dijk

Rombouts

S. A.

Crone

E. A.

(2011) Dissociable brain networks involved in development of fairness considerations: understanding intentionality behind unfairness. Neuroimage 57(2): 634–641. doi:10.1016/j.neuroimage.2011.04.032.

30.

Güth

Schmittberger

Schwarze

(1982) An experimental analysis of ultimatum bargaining. Journal of Economic Behavior & Organization 3(4): 367–388. doi:10.1016/0167-2681(82)90011-7.

31.

Hartley

C. A.

Somerville

L. H.

(2015) The neuroscience of adolescent decision-making. Current Opinion in Behavioral Sciences 5: 108–115. doi:10.1016/j.cobeha.2015.09.004.

32.

Hoffman, E., McCabe, K., Shachat, S., & Smith, V. (1994). Preferences, property rights and anonimity in bairganing games. Games and economic behaviour, 7, 346–380.

33.

Huizenga, H. M., Crone, E. A., & Jansen, B. J. (2007). Decision-making in healthy children, adolescents and adults explained by the use of increasingly complex proportional reasoning rules. Developmental Science, 10(6), 814–825.

34.

Jacobs

J. E.

Klaczynski

P. A.

(2002) The development of judgment and decision making during childhood and adolescence. Current Directions in Psychological Science 11(4): 145–149. doi:10.1111/1467-8721.00188.

35.

Larson

G. E.

Saccuzzo

D. P.

(1986) Gender, neuroticism and speed-accuracy tradeoffs on a choice reaction-time task. Personality and Individual Differences 7(6): 919–921. doi:10.1016/0191-8869(86)90096-6.

36.

Marchetti

Castelli

Harlé

K. M.

Sanfey

A. G.

(2011) Expectations and outcome: The role of proposer features in the Ultimatum Game. Journal of Economic Psychology 32(3): 446–449. doi:10.1016/j.joep.2011.03.009.

37.

Massaro

Castelli

Sanvito

Marchetti

(2014) The ‘I knew it all along’ phenomenon: second-order false belief understanding and the curse of knowledge in primary school children. European Journal of Psychology of Education 29(3): 311–326. doi:10.1007/s10212-013-0200-7.

38.

Moron, M. (1997). Unpublished MA Thesis. Krakow: Jagiellonian University.

39.

Overgaauw

Güroğlu

Crone

E. A.

(2012) Fairness considerations when I know more than you do: Developmental comparisons. Frontiers in Psychology 3: 424. doi:10.3389/fpsyg.2012.00424.

40.

Pohl

R. F.

Bayen

U. J.

Martin

(2010) A multiprocess account of hindsight bias in children. Developmental Psychology 46(5): 1268–1282. doi:10.1037/a0020209.

41.

Rabin

J. S.

(1993) Incorporating fairness into game theory and economics. American Economic Review 83(5): 1281–1302. doi:10.2307/2117561.

42.

Reyna

V. F.

Farley

(2006) Risk and rationality in adolescent decision making: Implications for theory, practice, and public policy. Psychological Science in the Public Interest 7(1): 1–44. doi:10.1111/j.1529-1006.2006.00026.x.

43.

Roese

N. J.

Vohs

K. D.

(2012) Hindsight Bias. Perspectives on Psychological Science 7(5): 411–426. doi:10.1177/1745691612454303.

44.

Romer

(2010) Adolescent risk taking, impulsivity, and brain development: implications for prevention. Developmental Psychobiology 52(3): 263–276. doi:10.1002/dev.20442.

45.

Sanfey

A. G.

(2007) Social decision-making: insights from game theory and neuroscience. Science 318(5850): 598–602. doi:10.1126/science.1142996.

46.

Sanfey

A. G.

Rilling

J. K.

Aronson

J. A.

Nystrom

L. E.

Cohen

J. D.

(2003) The neural basis of economic decision-making in the Ultimatum Game. Science 300: 1755–1758. doi:10.1126/science.1082976.

47.

Schmidt

M. F.

Sommerville

J. A.

(2011) Fairness expectations and altruistic sharing in 15-month-old human infants. PLoS One 6(10): e23223. doi:10.1371/journal.pone.0023223.

48.

Sloane

Baillargeon

Premack

(2012) Do infants have a sense of fairness? Psychological Science 23(2): 196–204. doi:10.1177/0956797611422072.

49.

Solnick, S. J., & Schweitzer, M. E. (1999). The influence of physical attractiveness and gender on ultimatum game decisions. Organizational Behavior and Human Decision Processes, 79(3), 199–215.

50.

Steinbeis

Bernhardt

B. C.

Singer

(2012) Impulse control and underlying functions of the left DLPFC mediate age-related and age-independent individual differences in strategic social behavior. Neuron 73(5): 1040–1051. doi:10.1016/j.neuron.2011.12.027.

51.

Steinberg

Cauffman

(1996) Maturity of judgment in adolescence: Psychosocial factors in adolescent decision making. Law and Human Behavior 20(3): 249–272. doi:10.1007/BF01499023.

52.

Steinberg

Morris

A. S.

(2001) Adolescent development. Journal of Cognitive Education and Psychology 2(1): 55–87. doi:10.1891/194589501787383444.

53.

Sutter

(2007) Outcomes versus intentions: On the nature of fair behavior and its development with age. Journal of Economic Psychology 28(1): 69–78. doi:10.1016/j.joep.2006.09.001.

54.

Szymura

Nęcka

(1998) Visual selective attention and personality: An experimental verification of three models of extra version. Personality and Individual Differences 24(5): 713–729. doi:10.1016/S0191-8869(97)00198-0.

55.

Szymura, B., Slabosz, A., & Orzechowski, J. (2001). Some benefits and costs of the selectivity automatization. Poster prepared for the 12th Conference of the ESCOP, Edinburgh, 5–8 September 2001.

56.

van Duijvenvoorde

A. C.

Peters

Braams

B. R.

Crone

E. A.

(2016) What motivates adolescents? Neural responses to rewards and their influence on adolescents’ risk taking, learning, and cognitive control. Neuroscience & Biobehavioral Reviews 70: 135–147. doi:10.1016/j.neubiorev.2016.06.037.

57.

Will

G. J.

Crone

E. A.

van Lier

P. A.

Güroğlu

(2016) Longitudinal links between childhood peer acceptance and the neural correlates of sharing. Developmental Science 21(1)): doi:10.1111/desc.12489.

58.

Zanolie

de Cremer

Güroğlu

Crone

E. A.

(2015) Rejection in bargaining situations: An event-related potential study in adolescents and adults. PLoS One 10: e0139953. doi:10.1371/journal.pone.0139953.