Studies of cheating

Today, we’re looking at two studies of cheating by young children. First, Zhao, Yan, Harris & Lee (2025) published “Telling Young Children an Adult’s Emotional Reactions to Their Future Honest or Dishonest Behavior Causes Them to Cheat Less” Developmental Psychology. Here are the edited abstract and impact statement:

The current research examined whether telling young children about an adult’s emotional reactions to their future honesty or dishonesty influences their cheating. In five preregistered studies, children aged 3–6 years participated in a challenging test, purportedly to assess their knowledge but actually to measure their honesty (N = 480; 240 boys; all middle-class Han Chinese). Telling 5- to 6-year-olds about a familiar adult’s negative emotional reactions to their future dishonesty significantly reduced subsequent cheating, regardless of whether the adult was their homeroom teacher or their mother. Telling 3- to 4-year-olds about their mother’s positive reactions to honesty or her negative reactions to dishonesty also reduced cheating. Thus, providing information about a familiar adult’s emotional reactions can effectively promote honest behavior among young children. 

Can young children learn how to behave morally based on verbal information about others’ emotional reactions? We found that telling 5- to 6-year-olds about the unhappy reaction of their homeroom teacher and mother to dishonesty made them less inclined to cheat. Similarly, telling 3- to 4-year-olds about their mother’s happy reactions to honesty or unhappy reactions to dishonesty reduced cheating. Thus, evoking a familiar adult’s potential emotional reactions is a feasible and flexible way to promote honesty in early childhood and can be added to the honesty-promoting toolbox that parents and teachers use. 

I thought this one was cool in illustrating the power of a parent or teacher modeling emotional reactions to cheating. The next study looks more closely at the link between moral judgment and cheating. Zhao, Yan, Peng & Harris (2025) published “Moral Judgment and Cheating: Evidence of a knowledge–behavior link in early childhood” in Child Development. Here are edited excerpts:

This research with two studies examined whether young children's moral judgments of honesty and dishonesty predict their actual cheating behavior. Participants were 200 children aged 3–6 years (2021–2022. Study 1: N = 80, Mage = 4.96, 40 girls; Study 2: N = 120, Mage = 4.98, 60 girls; all middle-class Han Chinese). Children completed a temptation resistance paradigm assessing honest or cheating behaviors. They also made moral judgments about story characters who cheated or acted honestly on school tests. Results showed that the more negatively children judged cheating, the less likely they were to cheat. This finding indicates a knowledge–behavior link regarding cheating behavior. It underscores the importance of studying the connection between moral cognition and action in early childhood.

One of the enduring controversies in psychology is whether humans act according to what they know. In many situations, people know one thing but do not act accordingly, generally called the Knowledge–Behavior Gap (e.g., altruism). This Knowledge–Behavior Gap is particularly pronounced and problematic in the moral domain, where people conduct themselves unethically despite knowing it to be morally wrong. Evidence has shown that in addition to mitigating situational factors, people have quite sophisticated cognitive strategies to rationalize their immoral actions (e.g., neutralization, discounting, moral licensing). We hypothesize that such strategies may be responsible for the Moral Knowledge–Behavior Gap in older children and adults but not in early childhood when young children have yet to learn these strategies. In two studies, the current research directly tested this hypothesis. Specifically, we examined whether children between 3 and 6 could make appropriate moral judgments about academic honesty and cheating, and more importantly, whether these judgments could predict children's actual cheating behavior on a test.

We conducted two studies to test our hypothesis that a Knowledge–Behavior Link exists in the context of honesty among young children. That is, children who judge honesty more positively and cheating more negatively are more likely to act honestly themselves. To test the hypothesis, we situated 3–6-year-old children in a simulated test situation where they had to decide whether to cheat or not. We also asked children to make moral judgments about child protagonists who either cheated on various tests or acted honestly. The results supported the hypothesis regarding the relation between children's moral judgments of cheating and their actual cheating behavior. However, the hypothesized relation between children's moral judgments of honesty and their cheating was not supported.

We found that many children between 3 and 6 years of age readily engaged in cheating to perform well. Nevertheless, their rates of cheating were lower than in existing studies involving similarly aged children with the use of similar research paradigms. Typically, in these studies, more than 90% of children cheated. One major difference between the current and previous research is that children were offered physical rewards to perform well in most previous studies, whereas in the current research, rewards were not provided. Taking away the physical rewards appeared to lead some young children to be less motivated to cheat. Nevertheless, the reduced cheating among some children provided sufficient individual variation for the current research to assess whether children's cheating behaviors are related to their moral judgments. This was particularly the case for Study 2 in which only 55.8% cheated.

Regarding moral judgments of honesty and cheating in test situations, we found that children as young as three readily judged cheating negatively and honesty positively. This finding is in line with previous research showing that children can make appropriate judgments about various moral or immoral acts from a very young age (e.g., lying, stealing, and aggression). Nevertheless, the current research is the first to show that young children can make appropriate moral judgments regarding honesty and dishonesty in a test situation.

For 3–4 year olds, we found that moral judgments for honesty were significantly more positive when test rules were explicitly stated compared to when they were not stated. Both 3–4 and 5–6 year old judged cheating more negatively when the rules were conveyed explicitly versus implicitly. These findings generally align with the existing moral judgment literature, suggesting that young children are highly sensitive to rules articulated by authority figures. They tend to judge rule violations more harshly if the violations directly flout the authority figure's explicit instruction (Smetana et al. 2014).  Our findings show that this general finding can be applied to transgressions in test situations.

Regarding the central focus of the present research, we found that children's moral judgments of cheating were significantly associated with their actual behavior. Those who rated cheating more negatively were less inclined to cheat, suggesting a Cheating Knowledge–Behavior Link. This finding is inconsistent with the wide-ranging existing literature on children's moral knowledge and behavior. It is widely accepted that children and adults display a Moral Knowledge–Behavior Gap whereby they typically know an act to be morally wrong but do it anyway. 

As mentioned in the introduction, another explanation for the failure to observe a link between judgment and behavior could be that most children rated lying very negatively in these studies. However, their rate of lying was very high. This low variability in both variables might mask an otherwise significant association between moral judgment and behavior in honesty. Our findings lend support to this explanation. In the current research, we used a behavioral paradigm that produced considerable variability in cheating behavior among children (particularly in Study 2). We also used moral judgment materials that produced sufficient variability (particularly in the conditions without rule reminders in Study 2). As a result, we were able to uncover a significant and robust association between children's moral judgments of cheating and their actual cheating behavior.

One possibility is the well-known negativity bias in early moral and behavioral socialization. It is well established that parents and teachers tend to focus on negative outcomes. In daily life, parents and teachers pay more attention to children's wrongdoings and give harsh criticism or other negative feedback for their actions that violate moral and social norms. By contrast, these caregivers and educators rarely give positive feedback for actions that adhere to moral and social norms. Additionally, evidence from social evaluation in infants suggests that 3-month-old infants are more sensitive to actions that violate moral norms (e.g., hindering) than to those that adhere to moral norms (e.g., helping). Thus, children may be more sensitive to stories concerning child characters behaving negatively than positively. In turn, children's negative judgments of moral violations may be more effective than their positive judgments of moral conformity in guiding their behavior when facing similar situations. In other words, a negativity bias may exist regarding the link between moral knowledge and behavior in school activities. Specifically, our findings suggest that, at least among children as young as 3 to 6, there is a Knowledge–Behavior Link but only regarding negative judgments of cheating, not positive judgments of honesty. This is in line with the notion that young children's moral judgments are heavily shaped by the fear of punishment, leading them to act strategically.

Another explanation could be that the knowledge–behavior link is domain-specific, existing for dishonesty but not necessarily for honesty contexts. This aligns with findings from several studies of older children, which found that children's positive moral judgments about prosocial lies (i.e., lying in order to be polite, modest, or to protect their group) predicted their actual lying behaviors. This domain specificity may also explain why no significant knowledge–behavior link was found for sharing in young children.

Theoretically, our discovery that older preschoolers can accurately judge test cheating and honesty without rule reminders and younger ones can do so with a brief rule reminder provides strong evidence for the early development of rule awareness regarding test-taking. This discovery also supports the broader theoretical account of the early origin of moral and norm awareness in early childhood.

More important, our findings undermine the doctrine of specificity theory and support the idea that children's moral understanding is linked to their honest behavior. They suggest that situational factors do not solely determine children's decision to be honest or cheat. Thus, a theoretical account of the development of honesty must consider not only what is external (i.e., situational factors) but also what is internal (e.g., moral cognition and other social-cognitive factors such as executive functioning and theory of mind). More broadly, the existing research on moral knowledge–behavior gaps mainly involves older children and adults. The present findings suggest that we should also consider the preschool years to better understand the developing link between moral knowledge and behavior. It is possible that at the beginning of moral socialization, children's moral actions are often aligned with their moral knowledge. However, with age, children consider other internal and external factors when deciding whether to act according to their moral knowledge, leading to a moral knowledge–behavior gap.

Methodologically, the existing studies claiming a Knowledge–Behavior Gap tended to have low moral knowledge and behavior variabilities. By contrast, the present research used measures that produced results with considerable individual variability. In this context, we obtained evidence for a Cheating Knowledge–Behavior Link. This finding illustrates the importance of using appropriate measures in future research to assess the existence of knowledge–behavior links or gaps in other moral or non-moral domains.

The current research also has important practical implications. Previous research mainly focused on the role of situational factors in children's cheating behavior, leading to suggestions about creating environments conducive to honesty. Our findings support moral theorists advocating moral education about honesty and cheating as an avenue to reduce early cheating. Further, for younger preschool children, it is also essential to be explicit about the rules so that they can understand the moral implications of honesty and cheating better and, consequently, cheat less.

In conclusion, this research revealed that moral judgments about cheating in children aged 3–6 significantly predicted their actual cheating; the more negative their judgment, the less likely they were to cheat. Thus, the current research confirms the existence of a Cheating Knowledge–Behavior Link in the preschool years. It also suggests that with improved methodologies, we may be able to discover other knowledge–behavior links in other moral domains from early childhood. Such discoveries will ultimately help resolve the enduring theoretical debate about the relation between moral knowledge and behavior and provide evidence-based guidance for promoting morality in childhood and beyond.

I thought it helpful to have detailed information about the knowledge-behavior link. Consistent with the Zhao et al. study, here we see that negative moral judgment predicts less cheating. I also found it helpful to differentiate the younger children from the older ones.