The Motivational Aspect of Children’s Delayed Gratification: Values and Decision Making in Middle Childhood

Delayed gratification is the ability to postpone an immediate gain in favor of greater and later reward. Although delayed gratification has been studied extensively, little is known about the motivation behind children’s decisions. Since values are cognitive representations of individuals’ motivations, which serve to guide behavior, we studied the relationship between children’s values and delayed gratification. Two main distinct motivations overlapping with values may underlie this decision: conservation - the desire to reduce uncertainty and preserve the status quo, and self-enhancement – the desire to maximize resources and profit for the self. Accordingly, we hypothesized that conservation values would relate to children’s preference to hold on to what is given as soon as possible, and that self-enhancement values would relate to children’s preference for delaying gratification. Seven-year old children (N = 205) ranked their values with the Picture-Based Values Survey (Döring et al., 2010) as part of the Longitudinal Israeli Study of Twins (LIST) (Avinun and Knafo, 2013). The children also played a decision-making animation game that included delayed gratification decisions. In support of our hypotheses, greater delayed gratification related negatively to conservation values, specifically to security and tradition, and related positively to self-enhancement values, especially power and achievement. This is one of the first demonstrations that children’s values relate meaningfully to their behaviors.


INTRODUCTION
One of the most important challenges we face in life is the need to delay gratification. The ability to either forgo immediate temptation or to persist in an undesirable activity, in order to reach a later goal, is a key component to success in many life tasks, such as preparing for exams (Bembenutty, 2007;Zhang et al., 2011), losing weight (Davis et al., 2004;Nederkoorn et al., 2006), quitting smoking (Bickel et al., 1999;Mitchell, 1999), and saving for retirement (O'Donoghue and Rabin, 2000;Wulfert et al., 2002). Delay of gratification is not a capability we are born with, but one achieved throughout development (Mittal et al., 2013). Therefore, it is an important aim to investigate delayed gratification in childhood and understand its underlying motivation. The current study aims to achieve this goal using data from a sample of children who reported their values and made economic decisions in which they were required to decide whether to delay or not to delay gratification.

Delayed Gratification
Delayed gratification is the ability to postpone an immediate reward for the sake of more distant long-term gains. Generally, delayed gratification is assessed in tasks requiring individuals to forgo a smaller but immediate reward for the sake of receiving a larger reward in the future (Mischel and Ebbesen, 1970;Carducci, 2009).
The ability to delay gratification is known to increase with age (Mischel and Metzner, 1962;Yates et al., 1981;Thompson et al., 1997;Lemmon and Moore, 2007;Atance and Jackson, 2009). During the fourth year of life, children acquire the ability to deal with future-oriented situations (Thompson et al., 1997), and throughout the fifth-year, children exhibit cognitive strategies needed for delaying gratification (Mischel et al., 1989). Thus, by the age of five, children can opt to delay gratification (Moore and Macgillivray, 2004). Nevertheless, there are meaningful withinage individual differences. For example, in Carlson (2005) more than 30% of five-years-old children preferred NOT to delay gratification. Most of the extant research attributes this heterogeneity to individual differences in cognitive skills enabling delay, such as executive function (Baumeister and Vohs, 2003;Happaney et al., 2004;Zelazo and Carlson, 2012), and to variation in morphology and activation of brain regions sub-serving these capacities, such as the prefrontal cortex (Mobini et al., 2002;McClure et al., 2004;Casey et al., 2011).
Although delay of gratification has been studied extensively, little is known about the motivational features underpinning children's preference to delay gratification (Tobin and Graziano, 2009). Studies which deal with motivation typically relate to the situational constraints of the experiment. For example, children show greater motivation to refrain from eating a single cracker, if the delayed reward they can expect is larger (Carducci, 2009). Studies show that unreliable environments reduce children's preference to wait for greater rewards. Delayed gratification requires trust in people who give the reward to deliver it in the future as promised, and if there is no trust, the child will prefer the certain option, an immediate if smaller gain (Kidd et al., 2013;McGuire and Kable, 2013;Michaelson et al., 2013).
Focusing on individual differences in the motivational aspects of delayed gratification, research has primarily focused on differences in personality traits underlying the successful implementation of self-control. For example, whether they have strong ego resiliency or strong willpower. Indeed, individuals who score higher on these traits tend to have more success in delaying gratification (Metcalfe and Mischel, 1999;Mischel and Ayduk, 2004). However, the content of this motivation, i.e., the personal motivational goals behind the desire to delay gratification, has not been studied. The motivation of some children for preferring the smaller reward has received even less attention. This leads to the question at the heart of the current investigation. What provides the motivation behind individuals' decision to delay or not to delay gratification?
And especially, what motivates some children to prefer a smaller reward?
To shed light on the motivational factors involved in delayed gratification, we propose to study the role of values, because values are cognitive representations of individuals' motivations (Kluckhohn, 1951;Rokeach, 1973;Schwartz, 1992Schwartz, , 2012Fischer and Boer, 2016). Values represent what is important or desirable, and thus suggest which behaviors should be performed (Miles, 2015). Indeed, values are systematically predictive of relevant behaviors (Sagiv and Roccas, 2017). Therefore, we investigate the relationship between children's values and decision making involving delayed gratification.

Values
Values are defined as desirable, abstract goals, varying in importance across individuals and groups, and considered to be stable over time and across situations (Schwartz, 1992). Values serve as guiding principles for behavior, as principles to justify one's behavior, and as guides toward the evaluation of other people and of the self (Rokeach, 1973;Schwartz, 1992;Schwartz et al., 2012). Thus, one of the reasons for values' importance is their ability to influence and predict the individual's perceptions, behavior and decision-making (Feather, 1995;Rohan, 2000;Roccas and Sagiv, 2010).
The most prominent framework for studying personal values in psychology is Schwartz's theory of human values (Davidov et al., 2008). Schwartz (1992) identified ten universal personal value types, which differ in the type of motivational goals they express, specifically: universalism, benevolence, conformity, tradition, security, power, achievement, hedonism, stimulation, and self-direction.
According to Schwartz's theory, there is a structure of dynamic relationships among values, representing the compatibilities and conflicts among them. The ten basic values are organized as a circular continuum according to the motivations they express, with adjacent values sharing compatible motivations and conflicting motivations reflected in opposite values (Davidov et al., 2008;Maio et al., 2009). The ten values can be also organized in four higher order values across two bipolar dimensions. The first dimension contrasts openness to change values (stimulation and self-direction) with conservation values (tradition, conformity, and security). The second represents self-transcendence values (universalism and benevolence) versus self-enhancement values (power and achievement). Hedonism shares element with openness to change and self-enhancement (Schwartz, 1992(Schwartz, , 2012. Recent research demonstrates that this structure is replicated, using age-appropriate methods, already in middle childhood (Abramson et al., 2017;Collins et al., 2017).
Following up on value-behavior associations found in adults and adolescents (Roccas and Sagiv, 2010;Pulfrey and Butera, 2013;Benish-Weisman, 2015), a few recent studies conducted among children found value-behavior correlations.
Children's values have been shown to relate meaningfully to behaviors reflective of the motivational aspect of each value (Vecchione et al., 2016). For example, children's conservation values were reflected in their tendency to avoid dangers, while self-enhancement values related to behaviors such as doing one's best to win competitions (Vecchione et al., 2016). In an experimental setting, children's decision to donate candy to other children correlated positively with their relative preference for self-transcendence over self-enhancement values (Abramson et al., 2017). The ability of children to distinguish among values based on their motivational structure, and the relevance of values to children's behavior, suggest the importance of values to children's decisions. Below, we propose specific associations between values and delayed gratification.

The Current Study
Values have been linked, theoretically and empirically, to motivational brain systems of approach and avoidance (Fischer and Boer, 2016;Fischer, 2017). Our hypotheses reflect such links between motivational systems and the values that promote them.
Gray's reinforcement sensitivity theory (RST; Gray, 1970;Gray and McNaughton, 2000) a leading neurobiological theory of personality seeking to account for the underlying biological and psychological dynamics of individual differences (Fischer, 2017), focuses on two personality dimensions, anxiety and impulsivity. Each of these dimensions represents a different neurobiological system which responds uniquely to the environment. The first system, Behavioral Inhibition System (BIS) seeks to inhibit behavior which might have negative or painful outcomes. In other words, it is underlain by the motivation to prevent aversive stimuli. A person who has a sensitive BIS tends to anxiety reactions, avoidance and desire for security. Concerned with avoidance of novel and uncertain situations or stimuli, the BIS can be seen as represented by Conservation values (Fischer, 2017).
Another system described by RST is the Behavioral Activation System (BAS), responsible for motivation to reach pleasant stimuli. BIS and BAS are sensitive to different kind of reinforcements (Torrubia et al., 2001). While BIS is sensitive to negative reinforcement, BAS is sensitive to positive reinforcement (Carver and White, 1994). When a person has to make a decision to delay or not to delay gratification, a person with a sensitive BIS might want to avoid the risk of loss, leading to preference for a smaller and immediate gain. In contrast, a person who has a sensitive BAS would focus on the opportunity to get a larger gain and would prefer to delay gratification. The BAS, with its approach motivation which enables or promotes seeking opportunities and engaging in novel situations, can be seen as overlapping with Self-enhancement values (Fischer, 2017).
Thus, when children, or adults, deliberate whether to prefer a smaller but immediate reward or a larger reward in the future, two different motivations, overlapping with the core dimensions of values, may underlie their decision. The first dimension of values (Schwartz, 1992) contrasts conservation with openness to change. Conservation values, focusing on reducing uncertainty and preserving the status quo (Schwartz, 2012), were hypothesized to relate to children's preference to hold on to what is given as soon as possible, because conservatives prefer the expected and known and avoid unpredictable situations (Schwartz et al., 2014), which is immanent in the choice to delay gratification. In contrast, openness to change values characterize individuals motivated to seek and accept change, innovation and adventure (Schwartz, 2012). We therefore expected children high in conservation values to be more likely to prefer a safe reward, if smaller, over an uncertain, larger reward.
The second dimension of values contrasts self-transcendence values of caring for others, with self-enhancement values, focusing on power and achievement. A focus on selfenhancement values is compatible with a motivation to attain resources for the self (Fischer, 2017), and we thus expected self-enhancement to relate positively to children's preference for maximization of profit via delaying gratification. Based on the motivational structure of values (Schwartz, 2007) decisions correlating positively with one end of a value dimension would correlate negatively with the other end of the value dimension. This would mean a negative correlation between self-transcendence values and delayed gratification. On the other hand, because the cover story of the decisions children engaged in was to obtain rewards for a later prosocial behavior (as detailed in the method section), self-transcendence could also relate positively to delayed gratification in the current case. We therefore do not propose a hypothesis in the case of self-transcendence.

Participants
The sample includes 205 seven-year old children (mean age = 90.60 months, SD = 3.46), of whom 109 (53.1%) are boys. 91% of the families lived in two-parent households, and the total number of children averaged 3.69 (SD = 1.59). The participants were part of a larger sample, participating in the Longitudinal Israeli Study of Twins (LIST), in which all Hebrewspeaking families of twins born in Israel during 2004 -2005 were invited to participate (Avinun and Knafo, 2013). When children reached the age of seven, the families living in Greater Jerusalem were invited to take part in an experimental session in the laboratory. Current participants are a subsample of the children who participated in a study of the genetics of values (Uzefovsky et al., 2016). The protocol for the experiment was approved by the Hebrew University Social Sciences research ethics committee, and written informed consent was obtained from participants' parents.

Measures
Values Values were assessed with the Picture-based Value Survey for Children (Döring et al., 2010), which was specifically designed to study values in childhood. The tool has been translated to Hebrew and slightly adjusted to Israeli culture, as described by Abramson et al. (2017), and Uzefovsky et al. (2016).
The experimenter asked the children "How would you like to be in your future life." Then, children were shown 20 cartoonlike pictures, two pictures for each value type. The pictures were printed on removable stickers. Each cartoon describes the same character performing a value-relevant action, accompanied by a brief caption. For example, the picture of a child wearing a helmet while riding the bicycle, and the text "to be safe" represent security values; an image of a winner on a podium accompanied by the caption "to be best" represent achievement values, and so on. An experimenter presented the child with all of the items and read each caption out loud. Children were then asked to sort the items according to five levels of importance. First, the child was asked to choose two "Very important" items, then two "Not at all important" items. Then, the child was asked to choose four items as "important" and four as "not important." Finally, the remaining eight items were automatically ranked in the intermediate level. Thus, the score in the scale ranges from 1 (not important at all) to 5 (very important). Structural analyses (reported by Uzefovsky et al., 2016) indicated that children's value structure closely resembled the prototypical Schwartz (1992) value structure found with adults and adolescents.
The ranking scores for each pair of items measuring the same value were averaged to obtain a score on each of the 10 values. In addition, to reduce the number of analyses, values were aggregated to form higher-order value scores of selftranscendence (universalism and benevolence), conservation (conformity, tradition, and security), self-enhancement (power and achievement), and openness to change (stimulation, selfdirection, and hedonism).

Delayed Gratification
Delayed Gratification was measured by an animation game, which was created especially for the current research. In the game, designed to engage children, the child was asked to help free the city by beating the evil wizard, who took all the colors from the city before the festival, and made everybody sad. To achieve this goal and free the city the child was required to collect as many gems as possible. Collecting and using gems along the way required making 23 decisions which simulated situations of economic gain or loss. During the game, among other decisions, there were three decisions that measured delayed gratification. In these decisions a witch, a lion, or a turtle offered the child the choice between a small, immediate reward, and a later but double in size, reward (5 or 10, 10 or 20, and 10 or 20, respectively, for the three trials). The time when the later reward would be given was unknown. See Figure 1 for an example.
The child's choices (0 = immediate and small reward, 1 = delayed and larger reward) were summed into one variable labeled delayed gratification. The scale ranged from 0 to 3, based on the count of times in which a delayed reward was chosen.

Socioeconomic Status
As families' socioeconomic status (SES) has been linked to both children's values (Uzefovsky et al., 2016) and their delayed gratification (Green et al., 1996), we included SES in our design. SES was indicated in the current sample (Avinun and Knafo-Noam, 2017), by three mother-rated variables: household income (rated as much below, below, similar, above, or much above a given national average income), mother's years of education, and the ratio of household number of rooms/residents. These components were standardized and an average was calculated to create the SES score.

RESULTS
About half (53.2%) of the children were willing to delay gratification in all three opportunities. A substantial proportion of children delayed only once (13.2%) or twice (20%), while 13.7% of the children did not delay gratification in any of the three opportunities.
Preliminary analyses showed that delayed gratification did not relate to age variation (in months), or twins' zygosity. Delayed gratification showed an insignificant relationship with the SES composite score and (beta = −0.02, ns), as well as its three components (household income, beta = −0.08, mother's years of education, beta = 0.07, and number of rooms/number of resident ratio, beta = −0.06, all ns). Finally, there were small sex differences in delayed gratification, such that boys delayed gratification slightly more times (M = 2.31, SD = 0.99) than girls (M = 1.91, SD = 1.16), t = 2.61, p = 0.01.
Bivariate correlations between values and delayed gratification were computed using the TYPE = COMPLEX option in Mplus (Muthén and Muthén, 2010), which accounts for the non-independence of observations for individuals within the same family. In line with our hypothesis, delayed gratification related negatively to conservation values, r = −0.15, p = 0.018. Follow-up analysis showed that delayed gratification related specifically to security, and to a lesser extent tradition (see Figure 2). The motivationally opposed values of openness to change showed a modest, positive correlation with delayed gratification, r = 0.11, n.s., reflecting a positive correlation with hedonism (Figure 2).
Moving on to the other major value dimension, delayed gratification related positively to self-enhancement values, r = 0.13, p = 0.035. We did not propose a directional hypothesis for the association between delayed gratification and the motivationally opposed values of self-transcendence. These values showed a negative correlation with delayed gratification, r = −0.11, n.s.
Because consistent (though small) sex differences have been observed in self-enhancement and self-transcendence values (Schwartz and Rubel, 2005;Knafo and Spinath, 2011), including in the current sample (Uzefovsky et al., 2016), we tested if sex accounted for the association between delayed gratification and self-enhancement. The association between self-enhancement values and delayed gratification was slightly reduced when sex was entered into the analysis as well, from β = 0.13 to β = 0.11, p = 0.06. Further analyses showed no interaction between sex and values in predicting delayed gratification.
To get a fuller understanding of our findings, we were interested in the role of specific combinations of values in children's delayed gratification. Recent work has shown the advantage of using value profiles to understand their associations with other variables (Ungvary et al., 2018). We therefore assessed the association between delayed gratification and value profiles based on a cluster analysis, to examine how values work together FIGURE 1 | An example of one of the delayed gratification decisions. Children were asked to choose between receiving 4 gems immediately and receiving eight gems later.

FIGURE 2 | Correlations between delayed gratification and values.
beyond each value independently. In addition, we asked whether values related to the frequency of low delayed gratification or to that of high levels of delayed gratification. A 2-step cluster analysis was conducted, using data on all four higher-order values (an analysis using only self-enhancement and conservation yielded similar findings). The analysis identified three clusters, illustrated in Figure 3. The first cluster (29.8% of all participants) included children high on self-enhancement values and low on conservation values and self-transcendence. We refer to children in this cluster as "Self-enhancers" for simplicity.
The second cluster (33.7%) included children who were on average close to the mean on all four values, though higher on openness to change and self-transcendence than on conservation and self-enhancement. This distinction represents one organizing principle of the value system, contrasting values reflecting anxiety (conservation and self-enhancement, particularly power) with values reflecting growth (openness and self-transcendence, Schwartz, 2015). Therefore, we refer to this cluster as "Growth focus." The third cluster (36.6%) was characterized by high degrees of conservation and low degrees of openness to change. We will refer to these children as "Conservatives." As would be expected from the correlation analyses, self-enhancers showed the highest levels of delayed gratification (M = 2.33, SD = 0.81), and children with growth focus values were not very different on average (M = 2.17, SD = 1.14), while conservative children delayed gratification on average 1.92 times (SD = 1.23).
To understand whether values related to high or low levels of delayed gratification we next created two new measures for delay of gratification. The first distinguished children who chose a larger and later reward, in all three opportunities, from children who at least once chose not to delay gratification. The proportion of children choosing to delay gratification in all  three opportunities did not differ across the three value clusters, χ 2 (DF = 2) = 2.70, ns.
A different picture emerged when we compared the three value clusters in the prevalence of children who never chose to delay gratification. While almost a quarter (22.7%) of conservatives never delayed gratification, only two children (3.3%) in the selfenhancers group were not willing to delay gratification at least once, χ 2 (DF = 2) = 10.76, p = 0.005, Cramer's V = 0.1619, a medium-size effect (see Figure 4).

DISCUSSION
Children's tendency to delay gratification has been linked to long-term positive outcomes, including higher intelligence (Bembenutty and Karabenick, 2004;Duckworth and Seligman, 2005), social responsibility and social competence (Mischel et al., 1989), and improved academic performance (Mischel et al., 1988;Wulfert et al., 2002). Moreover, the ability to delay gratification has been identified as a protective factor against serious psychological and physical health problems, such as conduct disorders, antisocial behavior, hyperactivity, addictive problems (Mischel et al., 1989;Moffitt et al., 2011;Paulus et al., 2015) and obesity (Schlam et al., 2013). In addition, delayed gratification relates negatively to being convicted of a crime and having economic problems (Moffitt et al., 2011).
Despite the importance of research on delayed gratification, a gap regarding the content of the motivation underlying the decision to delay gratification exists. We addressed this question focusing on the relationship between 7-year-old children's values and their decision-making involving delayed gratification. To the best of our knowledge, this is the first empirical demonstration of the association between delay of gratification and values. In support of our hypotheses, conservation values, especially security and tradition, correlated negatively, and selfenhancement values positively, with delayed gratification.
Children valuing conservation values prefer an immediate reward, even if it is smaller. As with the proverbial "bird in the hand, " individuals with high conservation values are more likely than individuals high on openness values to prefer the known and safe even if changes have the potential to lead to better outcomes (Gable et al., 2003;Janoff-Bulman, 2009). This is in line with main motivation underlying conservation values, especially security and tradition: preserving the status quo and avoiding unexpected situations (Schwartz, 2012;Schwartz et al., 2012). Interestingly, conformity values did not show the same association with delayed gratification as other conservation values did. In our previous work, in contexts involving trust in authorities, conformity did not relate to other variables in a way similar to the other conservation values (Nir and Knafo, 2009;Daniel et al., 2013). Possibly conformity values, which promote relying on social norms and authorities, are related to children's likelihood of trusting the experiment and foregoing a current reward for a later reward, an effect opposed to the overall tendency of conservation values to avoid uncertainty.
In contrast, children focusing on self-enhancement prefer to wait and maximize rewards. This reflects the motivation underlying self-enhancement values, that is the desire to have more resources available to the individual and to increase one's power and influence (Schwartz, 2012;Schwartz et al., 2012). The current work may be compared with work on Trinidadian adolescents (11-14-year-old), where individuals with a high need for achievement tended to prefer to wait until they achieved a higher reward (Mischel, 1961). While need for achievement, defined as "competition with a standard of excellence" (McClelland et al., 1953), is also conceptually related to self-direction values, it is possible that the selfenhancement aspect of achievement is responsible for the associations reported by Tobin and Graziano (2009). Future work should seek to investigate whether achievement values are involved in the longitudinal relationships between children's delay of gratification and their future scholastic achievements (Mischel et al., 1989;Wulfert et al., 2002).
However, it is important to note, that the effect sizes of most findings are modest. All the betas we observed were between 0.13 and 0.19. Moreover, although this study refers to values as motivational factors for delaying or not delaying gratification, it is limited in its ability to detect the causal influence of values on behaviors. Experimental work suggest that values can have a causal influence on behavior (Maio et al., 2009;Sagiv and Roccas, 2017). Nevertheless, because values can also be used to justify one's behavior (Schwartz, 1992) causal effects from behavior to values are also possible. We are aware of only one longitudinal study on children's values and behaviors. In that study, values and behaviors showed reciprocal longitudinal effects (Vecchione et al., 2016). Future research would benefit from longitudinal designs which would allow for greater inference regarding the direction of causality between values and delay of gratification.

Developmental Perspective
Until about a decade ago, there was a theoretical assumption that the development of values occurs in adolescence, with the formation of identity. However, with the recent development of research tools enabling the measurement of values among children (Döring et al., 2010;Collins et al., 2017), studies indicated meaningful value systems are present already in middle childhood, replicating the Schwartz (1992) value structure found in adolescents and adults (Cieciuch et al., 2016;Collins et al., 2017), even among children as young as 7 (Uzefovsky et al., 2016) and 5 years old (Abramson et al., 2017). This line of research opened ground for studies of the relations between values and behavior in children.
This is the first study of the relation between values and delayed gratification in childhood, which has the advantage of being a more informative period for assessing delay of gratification behavior. Additionally, this study is one of the first studies to show a relation between children's values and behavior. The current findings are consistent with what has been found among adults and adolescents, that individuals' values have a relationship with their behavior (Roccas and Sagiv, 2010;Pulfrey and Butera, 2013;Benish-Weisman, 2015).
Children's preferred choice in their decision making in our study involved delay of gratification. Half of the children opted to wait for the larger reward and most of them decided to delay gratification at least one time (of 3 times). Based on past research showing that delayed gratification increases from early to middle childhood (Mischel and Metzner, 1962;Yates et al., 1981;Thompson et al., 1997), younger children are expected to show lower levels of delayed gratification, and larger variability. This larger variability, on the one hand, might enable finding stronger associations with values in younger children. On the other hand, there is some (limited) evidence that value-behavior relationships become stronger with age during childhood (Henshel, 1971;Abramson et al., 2017). Research on children's values and behaviors from early to middle childhood is sorely needed.

Limitations
Although correlations between delay of gratification and values were observed, the effects were small in magnitude and further research is needed to understand what moderates the association between values and delayed gratification. One potential reason for the low correlations is the low variability in delayed gratification, which could result from the age of the sample, as noted. Another possibility is that children were more likely to forego current rewards when they were presented in the context of a computerized game, than they would if presented with reallife rewards such as a marshmallow candy. Replication of the current findings with such rewards is an important direction for future research.
Moreover, due to the relatively small number of decision trials, our delay of gratification task limits our ability to compare our findings more generally to delay discounting tasks commonly used in experimental economics (Green et al., 1996;Green and Myerson, 2004;Odum, 2011). Such tasks typically present subjects with a series of choices in which they must indicate a preference between smaller and more immediate rewards (e.g., $10 now), vs. larger and more distant rewards (e.g., $100 in 1 year). Immediate or distant rewards are varied systematically (e.g., $20 now vs. $100 in 1 year) until subjects reach an indifference point; whereby the immediate and delayed quantities are of equivalent subjective value. These tasks allow for a more finely tuned parameterization of the features characterizing delay of gratification and have the potential to contribute to our understanding of the association between values and delay of gratification. In contrast, our study is different from past work as the participants were not told when they could expect the future reward, which may have introduced an additional element of uncertainty into participants' decisions. We chose this more limited design in order to maximize 7 year old children's engagement with the task. Future work on children may develop measures for assessing delayed discounting which will then be associated with children's values.
Another potential limitation of our study is the composition of the sample: 7-year-old twins from the Jewish-Israeli population. Twins have not been shown to have substantially different personalities when compared to singletons (Johnson et al., 2002), but extending the findings with children from other cultures and religious backgrounds can elucidate important information about how the link between values and behavior may vary across different cultures. Finally, our twin sample was not large enough for complex genetic analyses. In a follow-up study with an expanded sample we intend to study the genetic and environmental contributions to children's decision making in the context of delayed gratification, as well as the role of genes and the environment in the association of delayed gratification with values.

CONCLUSION
The results support our hypothesized associations between children's high self-enhancement and low conservation values and their tendency to delay gratification. Accordingly, the results suggest that potential gain and distrust in the ability to retain a future reward, respectively, promote or hinder delayed gratification. Following the same children through adolescence in additional developmental research, this study will have implications for understanding how the motivation to delay gratification develops, contributing to children's and adolescents' functioning.

ETHICS STATEMENT
This study was approved by the Hebrew University Social Sciences Research Ethics Committee.

AUTHOR CONTRIBUTIONS
SI, IS, and AK-N designed the research. LT wrote the first draft of the manuscript. LT and AK-N were responsible for the analyses. All authors made contributions to the conception of the manuscript and participated in the writing process by adding substantively relevant content.

FUNDING
This research was supported by a grant from the Science of Generosity Initiative, Center for the Study of Religion and Society, University of Notre Dame, funded by the John Templeton Foundation, to AK-N.