You are on page 1of 20

Journal of Experimental Child Psychology 110 (2011) 373392

Contents lists available at ScienceDirect

Journal of Experimental Child


Psychology
journal homepage: www.elsevier.com/locate/jecp

Predictors of childrens prosocial lie-telling: Motivation,


socialization variables, and moral understanding
Mina Popliger a,, Victoria Talwar a, Angela Crossman b
a
b

Department of Educational and Counseling Psychology, McGill University, Montreal, Quebec, Canada H3A 1Y2
Department of Psychology, John Jay College, City University of New York (CUNY), New York, NY 10019, USA

a r t i c l e

i n f o

Article history:
Received 6 January 2011
Revised 9 May 2011
Available online 12 June 2011
Keywords:
Prosocial lying
Motivation
Socialization variables
Moral judgments
Concealment ability
Politeness

a b s t r a c t
Children tell prosocial lies for self- and other-oriented reasons.
However, it is unclear how motivational and socialization factors
affect their lying. Furthermore, it is unclear whether childrens
moral understanding and evaluations of prosocial lie scenarios
(including perceptions of vignette characters feelings) predict
their actual prosocial behaviors. These were explored in two
studies. In Study 1, 72 children (36 second graders and 36 fourth
graders) participated in a disappointing gift paradigm in either a
high-cost condition (lost a good gift for a disappointing one) or a
low-cost condition (received a disappointing gift). More children
lied in the low-cost condition (94%) than in the high-cost condition
(72%), with no age difference. In Study 2, 117 children (42 preschoolers, 41 early elementary school age, and 34 late elementary
school age) participated in either a high- or low-cost disappointing
gift paradigm and responded to prosocial vignette scenarios. Parents reported on their parenting practices and family emotional
expressivity. Again, more children lied in the low-cost condition
(68%) than in the high-cost condition (40%); however, there was
an age effect among children in the high-cost condition. Preschoolers were less likely than older children to lie when there was a high
personal cost. In addition, compared with truth-tellers, prosocial
liars had parents who were more authoritative but expressed less
positive emotion within the family. Finally, there was an interaction between childrens prosocial lie-telling behavior and their
evaluations of the protagonists and recipients feelings. Findings
contribute to understanding the trajectory of childrens prosocial

Corresponding author. Fax: +1 514 398 6968.


E-mail address: mina.popliger@mail.mcgill.ca (M. Popliger).
0022-0965/$ - see front matter 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jecp.2011.05.003

374

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

lie-telling, their reasons for telling such lies, and their knowledge
about interpersonal communication.
2011 Elsevier Inc. All rights reserved.

Introduction
From an early age, children are socialized to be truthful in most social contexts. However, they are
also taught, explicitly or implicitly, not to tell the blunt truth in some social situations where the truth
may be trivial or hurtful to the recipient (Sweetser, 1987). Learning to tell lies for the benet of others
(i.e., prosocial lies) is arguably important if children are to be perceived as polite and considerate. Prosocial lies are motivated by the desire to make others feel good or to spare the feelings of the recipient
and foster amicable social relationships (DePaulo & Bell, 1996; DePaulo & Kashy, 1998). Unlike antisocial lies, which are told solely for personal benet (e.g., to escape punishment, for material gain),
prosocial lies have some benet for the lie recipient (e.g., to be polite, to make another feel better)
and are not intended to cause harm to another individual. These lies are rated less negatively and considered more socially acceptable (e.g., Bussey, 1999; Walper & Valtin, 1992). For adults, prosocial lies
are common, told daily (DePaulo & Kashy, 1998), and a signicant part of maintaining social relationships (DePaulo & Jordan, 1982). Less is known about childrens prosocial lie-telling behavior.
To date, the majority of research on the development of childrens lying has focused on whether
children lie to conceal transgressions (e.g., Lewis, Stanger, & Sullivan, 1989; Polak & Harris, 1999; Talwar & Lee, 2002b), on when children begin to lie intentionally to trick others (Peskin, 1992), and on
how successful children are in telling lies of trickery (Feldman, Jenkins, & Popoola, 1979). Findings
suggest that these lies emerge during the preschool years and that childrens ability to maintain their
lies improves with age. However, it remains unclear how motivational and socialization factors affect
childrens prosocial lie-telling and whether childrens moral understanding and evaluations of prosocial lies predict their prosocial behaviors. These issues were explored in the current studies.
The development of prosocial lying has important implications for understanding social development. It provides a window to view the process by which children learn social skills necessary to communicate with others and form social relationships. Prosocial lies are considered a form of speech act
that both violates and upholds the basic rules of interpersonal communication. On the one hand, Grice
(1980) suggested that one of the most fundamental conventions governing interpersonal communication is the maxim of quality. This maxim requires speakers to be truthful to their communicative partners. Prosocial lies violate this maxim. On the other hand, Lakoff (1973) and Sweetser (1987)
suggested that there are just as many, if not more, fundamental communication rules that require
speakers to be amicable and to help, not harm, their communicative partners. Prosocial lies adhere
more to this set of rules. Brown and Levinson (1987) stressed that there is always tension between
satisfying fundamental conventions of communication (e.g., maxim of quality) and maintaining face
toward others. Although in most situations these two considerations promote consistent behaviors to
achieve a common communicative goal, in politeness situations they often collide. Such circumstances
might require a strategic trade-off between the two goals. Thus, childrens developing ability to tell
prosocial lies provides a unique opportunity to examine their developing knowledge about rules governing interpersonal communication. By examining childrens prosocial lying, we can explore whether
children are capable of reconciling seemingly contradictory rules of communication and using them
adaptively across social situations.
Childrens moral evaluations of prosocial lies
Several studies have addressed childrens perceptions of prosocial lies, focusing on the development of childrens conceptual understanding and moral evaluations of prosocial lying. Bussey
(1999) reported that children (411 years of age) labeled all untrue statements as lies regardless of
whether they were antisocial or prosocial in origin. On the other hand, Lee and Ross (1997) found that
adolescents (1217 years of age) and college students were less likely to identify false statements told

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

375

to protect another (i.e., prosocial lies) as lies than statements intended to harm another (i.e., antisocial
lies). Thus, in contrast to children, many adolescents and adults do not identify untruthful statements
told to help another or to be polite as lies (Sweetser, 1987).
When asked to evaluate prosocial versus antisocial lies, young children respond somewhat differently. In Busseys (1999) study of 4- to 11-year-olds, young children rated all lies negatively; however,
they rated prosocial lies as less negative compared with antisocial or pretense lies. Walper and Valtin
(1992) found that toward the end of elementary school, children provided negative ratings for antisocial lies but positive ratings for prosocial lies. Similarly, Keltikangas-Jaervinen and Lindeman (1997)
reported that 11- to 17-year-olds regarded lying to hurt a friend or for personal gain as very bad,
whereas lies told under duress or for a positive motive (e.g., to save or help a friend) were evaluated
more moderately. In fact, 4- to 9-year-olds in one study stated that a story character should lie about
liking a disappointing gift, and when asked about the feelings of the gift-giver, children thought that
the gift-giver would be happy to hear the gift was liked and unhappy if told the gift was not liked
(Broomeld, Robinson, & Robinson, 2002). Finally, Crossman et al. (2010) found that even the youngest children in their study (36 years) perceived prosocial lies as more acceptable than self-serving
lies.
Taken together, these studies suggest that from an early age, children appreciate the differences
among various types of lies and evaluate prosocial lies less harshly than other types of lies. Furthermore, childrens understanding and moral evaluations of prosocial lying change with age, perhaps
as they develop the ability to consider the contradictory rules evoked by the politeness situation
(i.e., the need to be truthful vs. the need to be polite and avoid harm).
Childrens prosocial lie-telling behavior
To date, there are only a handful of studies that have examined childrens prosocial lie-telling
behaviors (Talwar & Lee, 2002b; Talwar, Murphy, & Lee, 2007; Xu, Boa, Fu, Talwar, & Lee, 2010). Talwar
and Lee (2002b) directly examined the verbal and nonverbal behaviors of childrens prosocial lie-telling using a modied version of the classic Rouge Task (Gallup, 1970; Lewis & Brooks-Gunn, 1979).
Children (37 years of age) participated in this Reverse Rouge Task with an experimenter. Prior to
meeting the children, the experimenter placed a mark of lipstick on her own nose. She asked the children whether she looked okay before she had her picture taken and then left the room. Thus, the children needed to decide whether to tell a prosocial lie when explicitly asked about the appearance of the
experimenter. A confederate then entered and asked the children whether the other experimenter had
looked okay before she had her picture taken (to conrm childrens true beliefs). Of the 65 children, 55
said that the experimenter looked okay for the picture; thus, they told a prosocial lie. Later, adult
detectors who viewed video clips of childrens responses were unable to distinguish between the children who told prosocial lies and those who did not. Thus, children as young as 3 years were able to tell
prosocial lies successfully and avoid detection by adults. Interestingly, only 11% of children explained
that they had lied because they wanted to avoid causing embarrassment. Therefore, children may have
lied for a mixture of both self-oriented motivations (e.g., to avoid the potential negative consequences
of telling the truth) and other-oriented motivations (e.g., to be polite, to avoid embarrassing the lie
recipient).
In another study, Talwar, Murphy et al. (2007) used an undesirable gift paradigm (adapted from
Saarni, 1984) to examine whether 3- to 11-year-olds would tell prosocial lies to a gift-giver after receiving a disappointing gift. Children were given a gift (i.e., a bar of soap) after playing a game and were left
alone in the room to open it. On the experimenters return, children were asked whether they liked the
gift. Children could either tell a prosocial lie about liking the gift or tell the blunt truth and confess their
disappointment about the gift. Overall, the majority of children (77%) told the gift-giver that they liked
the gift while confessing to their parents that they did not like the gift. Older children (911 years) were
more likely (84%) to tell a prosocial lie than were preschoolers (35 years, 72%).
Finally, Xu and colleagues (2010) examined Chinese childrens prosocial lie- or truth-telling behaviors and their moral knowledge about truths and lies. They found that among 7-, 9-, and 11-year-olds,
all were able to correctly classify untruthful statements (including prosocial lies) as lies and truthful
statements as truths. However, it was only as age increased that children were also more inclined to

376

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

tell prosocial lies themselves (50% of 9-year-olds and 60% of 11-year-olds). Furthermore, when asked
why they lied, most younger lie-tellers said they were motivated to lie for self-oriented reasons (i.e., to
avoid negative consequences for themselves), whereas the majority of older lie-tellers said they were
motivated to lie for other-oriented reasons (i.e., to avoid hurting the feelings of the gift-giver). These
results indicate that childrens ability to tell prosocial lies increases with age. In addition, childrens
motives for telling prosocial lies also appear to change with age, moving from self-serving motives
at younger ages to more prosocial motives later on. Thus, the signicant age effect found by Xu and
colleagues (2010) and Talwar, Murphy et al. (2007) with regard to childrens lie-telling may reect
a developmental change in childrens tendency to tell a prosocially motivated lie. Although these studies provide evidence that children can tell prosocial lies in politeness situations, there is much that
remains unknown regarding prosocial lie-telling and its development. Specically, because children
tell prosocial lies for both self- and other-oriented motivations, it is important to manipulate the consequences of prosocial lie-telling to understand how motivational factors may affect childrens prosocial lie-telling.
Motivation
The studies described above (Talwar & Lee, 2002b; Talwar, Murphy et al., 2007; Xu et al., 2010) suggest that childrens prosocial lie-telling may reect both self- and other-oriented motivations. However, these studies did not explicitly examine the role of motivational context in childrens lietelling behavior given that the consequences of lying were not manipulated and there was little at
stake if children chose to tell a prosocial lie. Children may decide to lie based on an assessment of
the consequences to themselves versus the benet to others. For instance, Talwar, Lee, Bala, and Lindsay (2004) found that children were less likely to conceal a parents transgression if there were potential negative consequences to themselves (high-cost), but if children could not be blamed for the
transgression (low-cost), they were more willing to lie for the parent. Thus, when telling a lie to benet someone else, children may be less likely to lie if they perceive a high cost to themselves. Because
childrens prosocial lie-telling has been largely examined in low-cost situations where the consequences of lying are negligible (i.e., nothing lost), research is needed to explore the role of motivation
on childrens prosocial lying when the motivational context for telling a prosocial lie is manipulated.
Therefore, a primary objective of the current studies was to examine the impact of motivational context on childrens prosocial lie-telling. Here children were placed in either a low-cost condition, where
there were negligible consequences of lying, or a high-cost condition, where there were negative consequences to their own self-interests (i.e., they lost a desirable gift) if they told a lie. Prosocial lies in
the high-cost condition were lies told for truly prosocial motivations (i.e., prosocial other-oriented
lies), whereby telling such lies was for the benet of others and there was no gain to the self for lying.
In contrast, lies told in the low-cost condition were prosocial self-oriented lies, whereby telling lies in
this situation was also self-serving. Even though the prosocial outcome remained the same, the motivation to do so was not necessarily entirely seless (i.e., children could still keep their gift).
The current studies
The primary objective of this research was to examine how differences in motivational context
might relate to childrens prosocial lie-telling behavior. To examine childrens prosocial lie-telling
behavior, children were placed in a real-life politeness situation where they needed to decide whether
to tell the truth or lie. A modied disappointing gift paradigm (Cole, 1986; Saarni, 1984; Talwar, Murphy et al., 2007) was used to create this situation. Children were given an undesirable gift and then
asked by the gift-giver whether they liked it. The undesirable gift situation was chosen because it is
not uncommon in childrens lives (e.g., receiving clothes as a birthday gift), and young children are
socialized to dissemble their true feelings of dislike in such situations. From 4 years of age onward,
children are already able to dissemble their nonverbal behaviors (Cole, 1986; Saarni, 1984) and verbal
behaviors (Talwar, Murphy et al., 2007) to appear pleased about receiving an undesirable gift. However, the current studies examined the inuence of motivational contexts on childrens prosocial
lie-telling as well.

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

377

Study 1
Children from the second and fourth grades were randomly assigned to either a high- or low-cost
condition. These two age groups were selected for theoretical reasons. Specically, it is in this age
range that there is an overall increase in childrens prosocial lie-telling (Talwar, Lee, Bala, & Lindsay,
2002). Furthermore, at some point during the early and middle elementary school years, children acquire a greater social understanding of their relations with peers and others, self-presentation becomes more important for children, and there are greater social opportunities during which
prosocial lie-telling may be more likely to maintain amicable relations. It follows, then, that fourth
graders would be more likely than second graders to tell prosocial lies. In the high-cost condition, if
children lied to spare the feelings of a disappointing gift-giver, the perceived consequence was high
because children lost a gift they liked. In contrast, if children told the blunt truth, they would not need
to keep the undesirable gift. In the low-cost condition, children were able to keep the desirable gift in
addition to receiving another (undesirable) gift. The perceived cost to children was low if they lied because they had little to lose by telling the gift-giver they liked the gift. It was expected that more children would lie in the low-cost condition and that older children would conceal their lies better during
follow-up questioning.
Method
Participants
Participants were 36 second graders (17 girls and 19 boys, mean age = 7.44 years, SD = 0.65) and 36
fourth graders (17 girls and 19 boys, mean age = 9.31 years, SD = 0.47). Children, who were primarily
from middle-class and Caucasian families, were recruited from schools in an urban North American
city. All participants were seen at their schools if parents consented to their participation. Half of
the children in each grade were randomly assigned to either the high- or low-cost condition.
Materials
Two games were played. The rst was a guessing game where children needed to identify different
objects based on verbal clues. The second was a memory game where children needed to remember
the location of matching cards. Children received prizes for winning these games from a gift basket full
of toys. The desirable gift given to children was a colorful Slinky. The undesirable gift was a pair of
knitted socks (desirable and undesirable gifts were chosen based on pilot data). A small video camera
was concealed in a bag on the opposite side of the room to capture childrens responses.
Procedure
Children participated individually with a female researcher (E1). A hidden camera was set up in a
designated room within each participating school to record childrens behaviors. Children played different games and were told that they could win a prize from the gift basket. Children rst played a
guessing game with E1 to provide an opportunity for children to win a gift. At the end of the game,
E1 picked a toy, wrapped in a brown bag, from the gift basket and gave it to the children. All children
received the desirable gift. Children were left alone for a minute to examine the toy and then were
asked by E1, on her return, whether they liked their gift and what they planned to do with it. All children responded that they liked the gift.
Next, children played a game with a second female researcher (E2) while E1 left to get something
from the classroom. E2 played a memory game with the children. After children had won the memory
game, E2 informed them that they had earned another gift. E2 told children that she had made some of
the gifts herself and wanted to give them one of the gifts she had made because she was really proud
of them. E2 explained to the children, I want to give you a special gift. I made it myself! I hope you
like it! Children were randomly assigned to one of two conditions that manipulated the perceived
consequences of telling a lie (i.e., high-cost to self vs. low-cost to self). In the high-cost condition, children were told by E2 that they could keep only one gift. If they wanted the second gift, they needed to
trade their rst gift to keep the second gift. In the low-cost condition, children were told by E2 that

378

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

they could keep both gifts (i.e., they did not need to trade their gift to keep the second one). In both
conditions, E2 then gave children the disappointing gift wrapped in a brown bag. Children were left
alone to open the gift. E2 returned to the room and asked children how they liked the gift. To examine
childrens ability to maintain their lies and give plausible explanations, E2 then asked children what
they liked about the gift and what they planned to do with it. Finally, E2 left the room and E1 returned
to escort the children back to their classrooms. Before leaving, E1 noticed the second gift and asked
about it. E1 asked the children if they liked the gift and whether they would prefer to have another
gift instead. All children said they preferred to have another gift and chose a second gift from the gift
basket. All children were allowed to keep the rst gift as well.
Children were classied as prosocial lie-tellers if they told E2 that they liked the disappointing gift
but later indicated to E1 that they did not like the gift. If children told both experimenters that they
did not like the gift, they were coded as blunt truth-tellers. In addition, childrens answers to E2s follow-up questions were coded as being plausible or implausible explanations. An example of a plausible answer demonstrating a childs ability to maintain his or her lie was I like how soft they look and
can wear them on a cold day, whereas an example of an implausible explanation demonstrating difculty in maintaining the lie was I dont know [what I like about them] or I would give them away.
Results and discussion
Overall, 83% of children told a prosocial lie about liking the undesirable gift. A binary logistic
regression was conducted to determine which factors might predict childrens prosocial lie-telling.
For this and all subsequent logistic regression analyses, the predictor variables were childrens age
group (second grade or fourth grade), sex, and condition (high-cost or low-cost). Signicance was assessed by a block v2 test (also known as the v2 difference test). In this test, the retention of the interaction term in a model must increase the variability accounted for in order to justify using a more
complex model (Menard, 2002).
The logistic regression analyses with childrens prosocial lie-telling behavior as the predicted variable revealed that the best t model included age group, sex, and condition, without any interaction
terms, as signicant predictors of childrens prosocial lie-telling behavior, v2(3, N = 72) = 8.37, p = .039,
Nagelkerkes R2 = .19, with 83% of cases correctly classied by this model. Only condition was a unique
signicant predictor of childrens prosocial lie-telling behavior, b = 1.89, Wald = 5.25, p = .02, odds
ratio = 1.51. More children lied in the low-cost condition (94%) than in the high-cost condition
(72%) (see Fig. 1). Thus, when telling a lie was costly to themselves, children were more likely to tell
the blunt truth. However, when their personal interests did not conict with the desire to be polite,
nearly all children lied and said they liked the disappointing gift. It might have been easier for children
to tell a lie when there was a self-oriented motivation for lying rather than for truly prosocial otheroriented purposes. There were no signicant age differences in childrens prosocial lying under these
varying conditions (Fig. 1).
Next, childrens ability to conceal their lies by maintaining the lies during E2s follow-up questioning was examined. The logistic regression analysis with childrens explanations (plausible or implausible explanation) as the predicted variable revealed that the best t model included age group, sex,
and condition, without any interaction terms, as signicant predictors of childrens explanations,
v2(3, N = 61) = 12.37, p = .01, Nagelkerkes R2 = .26, with 71% of cases correctly classied by this model.
Age group was a unique signicant predictor of childrens explanations, indicating that as age increased, participants were more likely to maintain and conceal their lies, b = 1.36, Wald = 4.40,
p = .04, odds ratio = 1.033. Specically, fourth graders concealed their lies to E2 more effectively
(79%) than second graders (56%) by providing a plausible explanation during follow-up questioning.
This was consistent with previous studies (e.g., Talwar, Murphy et al., 2007), where older children
were better at maintaining their prosocial lies than younger children. Sex was also found to be a signicant predictor of their explanations, with girls concealing more effectively (83%) than boys (52%),
b = 1.75, Wald = 7.13, p < .001, odds ratio = 1.033. Condition, however, was not a unique predictor of
childrens explanations.
Given the ndings regarding the inuence of motivational context on the development of childrens
lie-telling, a second study was conducted with a different sample of children from a broader range of

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

379

Fig. 1. Percentages of prosocial lie-tellers among second and fourth graders in high- and low-cost conditions in Study 1.

ages (412 years). Because previous studies have found developmental differences in childrens prosocial lie-telling behavior between preschoolers and older (school-age) children (e.g., Talwar, Murphy
et al., 2007; Xu et al., 2010), when the motivational context was not manipulated, replicating the ndings of Study 1 using a wider age range might provide more insight into this behavior. Specically, because it was found that more fourth graders told prosocial other-oriented lies than did second graders,
it might be that preschoolers indeed tell prosocial lies but that if they do they are more likely to be
prosocial self-oriented lies than other-oriented lies. Given that antisocial lies (i.e., self-gain lies told
to conceal a transgression and avoid punishment) have also been shown to emerge earlier than prosocial lies (e.g., Talwar & Lee, 2002a, 2002b), the likelihood that preschoolers would tell more self-oriented prosocial lies seems reasonable. In addition, given that the ndings of Study 1 differentiate
between two different types of prosocial lies depending on the motivational context (i.e., prosocial
other- and self-oriented lies), it is important to extend these ndings and to examine social and cognitive correlates of prosocial lie-telling in different motivational contexts. In particular, the inuence
of socialization variables (i.e., parenting practices and emotional expressivity within the family) on
childrens prosocial lie-telling behavior was examined as potential predictors of lying under different
motivational contexts. Finally, the relationship between childrens moral understanding and their prosocial lie-telling behaviors was also explored in relation to childrens prosocial lie-telling across motivational contexts.
Study 2
Prosocial lie-telling among 4- to 12-year-olds was examined using the disappointing gift paradigm.
To replicate and extend the ndings from Study 1, childrens prosocial lie-telling was examined in
three groups of children (preschoolers, early elementary school age, and late elementary school
age) in Study 2. Similar to Study 1, half of the children in each age group were randomly divided into
either a high- or low-cost condition. It was expected that preschoolers would be least likely to tell a
prosocial lie, especially if they were in the high-cost condition. It was also expected that childrens
ability to conceal their lies during follow-up questioning about their gift would increase with age
regardless of whether they were motivated to tell prosocial lies for self- or other-oriented purposes.
In addition, because prosocial lie-telling is a form of social communication used to be polite in some
situations, it is valuable to investigate socialization variables in relation to such lies to better understand childrens motivations for telling prosocial lies.

380

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

Social factors
Most studies on childrens lie-telling have examined cognitive factors related to the development
of lying (e.g., Chandler, Fritz, & Hala, 1989; Polak & Harris, 1999; Talwar, Gordon, & Lee, 2007; Talwar
& Lee, 2008). However, lying is an interpersonal exercise developed through social inuences as well.
Social domain theory (Turiel, 1983) suggests that childrens social knowledge and moral reasoning
(e.g., about lying) distinguish moral, socialconventional, and personal acts and their consequences,
and that this knowledge is constructed through childrens experiences (Smetana, 1999). Specically,
although antisocial and prosocial lies violate moral rules, telling the truth in prosocial contexts, such
as politeness situations, violates socialconventional rules. Through interactions with their environments, typically parents, children develop the ability to differentiate and apply these rules over time
(Smetana, 1999). Although it has yet to be examined empirically, there is some research to suggest
that different types of lie-telling might be inuenced by different aspects of parentchild interactions.
From a young age, children are socialized to be honest and not to tell antisocial lies, although their
prosocial lie-telling behavior may be overlooked or even encouraged and modeled by parents. Children may be taught, either explicitly or implicitly, to tell prosocial lies in order to be polite or protect
the feelings of others. Such lies may be seen as preferable to telling the truth in some social contexts
(Cole & Mitchell, 1998; Talwar, Gordon et al., 2007). Therefore, parents play a role in transmitting messages to their children that some lies are acceptable, whereas other lies are not.
There has been no direct examination of the inuence of family or social factors on childrens lietelling behavior. Yet, it has been suggested that in response to socialconventional transgressions (e.g.,
blunt truth-telling in politeness situations), authoritative mothers tend to emphasize social order, but
for moral transgressions (e.g., antisocial lie-telling), they emphasize consequences to the welfare or
rights of victims (Smetana, 1999). As children begin to experiment with lying, authoritative parents
might respond to antisocial lies as moral violations and focus on harm. On the other hand, children
who bluntly tell the truth in politeness situations may elicit instruction on social mores from parents
such as If you cant say anything nice, dont say anything at all. Hence, authoritative parents may
have children with better social skills who are more likely and better able to tell prosocial lies that
are other-focused rather than self-focused in appropriate situations (Talwar & Lee, 2008). However,
parents who are not authoritative do not tend to show consistency in their responses to social transgressions. Such parents might give mixed messages, blurring childrens social knowledge distinctions.
Furthermore, because prosocial lying is related to the emotional and social understanding of others,
parenting practices that facilitate the appropriate expression of emotions within the family context
(e.g., Halberstadt, Cassidy, Stifter, Parke, & Fox, 1995) may also inuence childrens prosocial lie-telling. That is, discussing and expressing both positive and negative emotions openly and freely within
the family may make it more salient for children to learn to express their emotions while in situations
that require them to tell prosocial lies. For example, Talwar, Murphy et al. (2007) found that when parents discussed the feelings of the gift-giver, children were more likely to tell prosocial lies and were
more convincing. In addition, family expressiveness has been found to be related to ones social skills
such as childrens understanding about emotions and social values (Denham & Grout, 1992). It may be
that positive emotional expressiveness promotes prosocial behavior by modeling positive display
rules in a disappointing gift situation, whereas negative family expressiveness models negative reactions in such situations. Conversely, negative family expressiveness may promote prosocial lying in
children as a strategy to avoid negativity in their social partner.
Thus, the current study examined parenting practices and emotional expressivity within the family
in relation to childrens prosocial lie-telling in different motivational contexts. Childrens parents completed questionnaires assessing their parenting styles and the emotional expressiveness within the
family. It was expected that for children in the high-cost condition, prosocial liars, compared with
truth-tellers, would have parents who scored higher on their authoritative parenting style rather than
the authoritarian and permissive indexes. In addition, they were expected to have overall high levels
of emotional expressivity, especially for positive emotions expressed. For children in the low-cost condition, prosocial liars, compared with truth-tellers, were expected to have parents who scored similarly on all indexes of parenting styles assessed, and there was not expected to be a strong
relationship with either positive or negative emotions expressed within the family.

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

381

Behavior and understanding


It is unclear whether childrens prosocial lie-telling behaviors are related to their conceptual moral
understanding of lies in politeness situations or evaluations of others feelings. Although a number of
studies have examined childrens moral and conceptual knowledge of lies (e.g., Broomeld et al., 2002;
Bussey, 1999; Lee & Ross, 1997) and a few studies have examined childrens lie-telling behavior (e.g., Talwar & Lee, 2002b; Talwar, Murphy et al., 2007), little research has examined their relation. Yet, the relationship between childrens social and moral conceptions of lying and their actual behaviors is important
because the purpose of socialization is to ensure not only that children know morally what is right or
wrong, and conventionally what is appropriate or inappropriate, but also that they act accordingly.
The few studies that have examined the relation between childrens moral understanding and their
lie-telling behavior have found mixed results. In one study, Talwar and colleagues (2002) found that
childrens understanding of lies was unrelated to childrens lie-telling to conceal their transgressions.
However, Talwar and colleagues (2004) found a modest correlation between childrens understanding
of lies and their lie-telling for another. Talwar and Lee (2008) also reported a signicant relation between childrens lying and their moral evaluations when a more comprehensive measure of childrens
understanding was used. Specically, children who gave higher ratings to the protagonists truthful
behaviors in hypothetical scenarios were more likely to confess their own wrongdoing when asked.
To date, only one study has examined the nature of the relationship between behavior and understanding in childrens prosocial lie-telling. In a study with Chinese children, Xu and colleagues
(2010) found that prosocial liars were more likely to rate prosocial lies positively and that this tendency increased with age. However, they examined childrens prosocial lie-telling behavior only in
relation to their moral judgments of vignette scenarios (i.e., evaluating the goodness or badness of
the lie situation) and did not address childrens evaluations of the protagonists or recipients feelings.
Broomeld and colleagues (2002) suggested that there were developmental differences in childrens
consideration of others feelings when evaluating prosocial stories, and Heyman, Sweet, and Lee
(2009) reported that older children do indeed evaluate lie-telling more positively and truth-telling
more negatively in politeness situations when considering the implications for others. However, neither study examined this in relation to actual lie-telling behaviors.
The relationship among childrens moral evaluations of lies, their understanding of the protagonists and recipients feelings, and their actual prosocial lie-telling behaviors in a disappointing gift situation was explored in Study 2. Children answered questions about their understanding and
evaluations of prosocial lie- and truth-telling scenarios. Specically, for children in the high-cost condition, prosocial liars were expected to evaluate the prosocial lie scenarios most positively when compared with truth-tellers. For children in the low-cost condition, no differences were expected. Because
liars in the high-cost condition were telling prosocial lies for other-oriented reasons with a cost to
their own self-interests, it was expected that they would have a more mature understanding of the
conventional and moral behavior in politeness situations and, thus, there would be a greater relationship between their behavior and evaluations. Finally, prosocial liars in the high-cost condition were
expected to evaluate the protagonist as doing something good and the recipient as being pleased
about the lie, whereas truth-tellers were expected to evaluate the protagonist as doing something less
good and the recipient as being upset about the lie. No differences were expected between the prosocial liars and truth-tellers in the low-cost condition in terms of their evaluations of the protagonists
and recipients feelings.
Method
Participants
Children from 4 to 12 years of age (N = 117, 64 boys and 53 girls) participated. There were 42 preschoolers (46 years, 23 boys and 19 girls, mean age = 60.10 months, SD = 8.60), 41 early elementary
school children (79 years, 21 boys and 20 girls, mean age = 97.22 months, SD = 7.50), and 34 late elementary school children (1012 years, 20 boys and 14 girls, mean age = 132.00 months, SD = 8.20).
Half of the children in each age group were randomly assigned to either the high- or low-cost condition. Children were recruited via advertisements in a free local newspaper targeted at families from an

382

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

urban North American city. Families, who were primarily middle-class and Caucasian, were seen at a
university research laboratory.
Measures
Child measures. Similar to Study 1, two games were played with each child: a guessing game and the
memory game. Children received prizes for winning these games. The desirable gift was either a colorful Slinky or Silly Putty; the undesirable gift was either a pair of knitted socks or a bar of soap (desirable and undesirable gifts were chosen based on pilot data). A video camera was concealed in a bag on
the opposite side of the room to capture a frontal view of childrens faces.
Moral vignettes. Four prosocial vignettes adapted from Bussey (1999) were used to examine childrens
moral evaluations. There were two prosocial true stories and two prosocial lie stories. The vignette
protagonists sex corresponded to the sex of the given child. A sample prosocial true story was as follows: Belindas mother baked some cookies that Belinda thought tasted awful. Belindas mother
asked Belinda if she liked the cookies. Belinda said No, they do not taste very nice. After children
were read each scenario, they were asked a series of questions. First, children were asked whether
they believed that the vignettes character told the truth or a lie. Next, children were asked to provide
a moral evaluation by determining how good or bad it was for the protagonist to have lied or told the
truth. Children were shown a moral response card with the following 6-point Likert scale to assist in
their responses: very very bad (1, three black Xs), very bad (2, two black Xs), bad (3, one black X), good
(4, one gold star), very good (5, two gold stars), and very very good (6, three gold stars). Children used
the same moral response card to rate how they thought the protagonist would feel for having lied or
told the truth and how the lie-or truth-recipient would feel after having been told a lie or the truth.
Parent measures
Parenting styles. The Parenting Styles and Dimensions QuestionnaireShort Form (PSDQ-SF) (Robinson, Mandleco, Olsen, & Hart, 1995) was used to measure parenting styles. The PSDQ-SF is a 32-item
self-rated inventory devised to assess authoritative, authoritarian, and permissive parenting styles, as
well as specic parenting practices within each of these styles. The questionnaire is intended for use
by parents of preschoolers and older (school-age) children. Items are rated on a 5-point Likert scale
ranging from 1 (never) to 5 (always), with scores based on the primary subscales and their respective
parenting factors. Some items are reversed scored.
The Authoritative style factor consists of 15 items with Cronbachs a = .91. Sample items for this
dimension include explains consequences of childs behavior and takes into account childs preferences in making family plans. The Authoritarian style factor consists of 12 items with Cronbachs
a = .86. Sample items of this dimension include explodes in anger toward child and uses physical
punishment as a way of disciplining child. The Permissive style factor consists of 5 items with Cronbachs a = .75. Sample statements of this dimension include states punishments to child and does not
actually do them and appears unsure on how to solve childs misbehavior.
Family expressiveness. The Self-Expressiveness in the Family Questionnaire (SEFQ) (Halberstadt et al.,
1995) measures parenting practices that foster emotional and social understanding. The SEFQ is a 40item scale designed to examine the frequency of emotional expressiveness within the family context.
Items are scored on a 9-point Likert scale, with scores ranging from 1 (not at all frequently) to 9 (very
frequently). Both positive and negative expressiveness items are represented on the scale. There are 22
positive expressiveness items (e.g., praising someone for good work,telling someone how nice they
look) and 18 negative expressiveness items (e.g., blaming one another for family troubles,quarreling with a family member). Scores are obtained by averaging responses for each subscale, with higher
scores representing greater levels of expressiveness. Cronbachs a values ranging from .85 to .94 have
been demonstrated for subscales of positive and negative expressiveness and the combined scale.
Procedure
Two sessions took place at a university research laboratory at least a week apart. Children participated individually in both sessions with trained researchers. During the rst session, the same

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

383

procedures described in Study 1 were conducted with regard to the disappointing gift paradigm. During the second session, children were read the moral vignettes and answered questions. Parents lled
out questionnaires as they waited for their children during the rst session.
Results and discussion
Lie behavior
Overall, 52% of children told a prosocial lie about liking the undesirable gift received from E2. A binary logistic regression was conducted to determine which factors might predict childrens prosocial
lie-telling. For this and all subsequent logistic regression analyses, the predictor variables were childrens age group (preschool, early elementary, or late elementary), sex, and condition (high-cost or
low-cost). Signicance was assessed by a block v2 test (also known as the v2 difference test).
The logistic regression analysis with childrens prosocial lie-telling behavior as the predicted variable revealed that the best t model included age group, sex, condition, and the Condition  Age
Group interaction as signicant predictors of childrens prosocial lie-telling behavior, Nagelkerkes
R2 = .09, v2(4, N = 117) = 15.58, p = .004, with 70% of cases correctly classied by this model. Condition
was a unique signicant predictor of childrens prosocial lie-telling behavior, b = 3.98, Wald = 12.21,
p = .001, odds ratio = 0.71. More children lied in the low-cost condition (68%) compared with children
in the high-cost condition (40%). This was the same pattern of results that was found in Study 1. Thus,
children were more likely to tell a lie about a disappointing gift to the gift-giver when they had little to
lose by telling such a lie.
There was also a signicant interaction between condition and age group, b = 1.41, Wald = 7.16,
p = .007, odds ratio = 3.53 (see Fig. 2). Although the majority of children lied in the low-cost condition
regardless of age, younger children were less likely to tell a prosocial lie in the high-cost condition.
Specically, whereas only 20% of preschoolers told a lie in this condition, 40% of early elementary
school children and 65% of late elementary school children lied. This was consistent with what was
expected given that older children were more likely to lie in the high-cost situation. Similarly, Xu
and colleagues (2010) found that older Chinese children were more likely than younger children to
tell prosocial lies for other-oriented reasons. Together, these results suggest that with increasing
age, children are more likely to tell prosocial lies motivated by other-oriented outcomes than by
self-oriented outcomes.
Next, the additional value of socialization variables for predicting childrens prosocial lie-telling
was examined using a binary logistic regression analysis. The predictor variables were entered on

Fig. 2. Percentages of prosocial lie-tellers in high- and low-cost conditions among preschoolers, early elementary school
children, and late elementary school children in Study 2.

384

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

the rst step, and the social variables (PSDQ-SF Authoritarian mean score, PSDQ-SF Authoritative
mean score, PSDQ-SF Permissive mean score, SEFQ Positive mean score, and SEFQ Negative mean
score) were entered on the second step. The rst step was signicant, Nagelkerkes R2 = .15, v2(4,
N = 111) = 13.65, p < .05. After controlling for age group, sex, condition, and the Age Group  Condition
interaction, the second step was also signicant, Nagelkerkes R2 = .39, v2(5, N = 117) = 17.56, p < .01,
with 72% of cases correctly classied by this model. PSDQ-SF Authoritative mean scores were a significant predictor of lying, b = 1.50, Wald = 6.95, p < .01, odds ratio = 4.49, whereby prosocial liars had
parents who scored higher on this measure (M = 4.25, SD = 0.47) than did nonliars (M = 4.01,
SD = 0.42), consistent with expectations (see Table 1). In addition, SEFQ Positive mean scores were a
signicant predictor of lying, b = 0.03, Wald = 6.46, p = .01, odds ratio = 0.97, whereby prosocial liars
had parents who scored lower on their positive emotional expressiveness (M = 7.57, SD = 0.86) than
did nonliars (M = 7.67, SD = 0.79) (Table 1). Although these effects were relatively small, perhaps children with lesser amounts of positive expressiveness seek such feedback through prosocial lying. These
results demonstrate that prosocial lie-telling is related to social variables. Finally, the interaction
terms among the PSDQ scores, SEFQ scores, and motivation condition were entered on the third step.
The third step was not signicant, Nagelkerkes R2 = .33, v2(5, N = 117) = 10.18, ns.
Childrens ability to maintain their lies during follow-up questioning
Childrens responses to follow-up questions after receiving the disappointing gift were examined to
determine whether they provided plausible or implausible explanations. To determine how good children were at maintaining their lies, a binary logistic regression was conducted on childrens explanations (plausible or implausible) as the dependent variable. The overall regression model only
approached signicance, Nagelkerkes R2 = .15, v2(3, N = 61) = 7.29, p = .063, with 72% of cases correctly classied by this model. Age group was a signicant predictor of childrens ability to maintain
their lies, b = 0.89, Wald = 6.37, p = .01, odds ratio = 0.41. Preschoolers were less likely to conceal
their disappointment based on their verbal responses (M = 1.19, SD = 0.40) than were early and late
elementary school children (M = 1.31, SD = 0.48, and M = 1.22, SD = 0.43, respectively). This nding
was consistent with previous studies (e.g., Talwar & Lee, 2002a; Xu et al., 2010), which also reported
an increased ability for older children to tell and subsequently maintain their prosocial lies.
Next, the additional value of socialization variables for predicting childrens explanations was
examined using a binary logistic regression analysis. The predictor variables were entered on the rst
step, and the social variables (PSDQ-SF Authoritarian mean score, PSDQ-SF Authoritative mean score,
PSDQ-SF Permissive mean score, SEFQ Positive mean score, and SEFQ Negative mean score) were entered on the second step. Refer to Table 2 for means and standard deviations of the social variables
according to the plausibility of childrens answers. After controlling for age group, sex, and condition,
the second step was not signicant, Nagelkerkes R2 = .25, v2(5, N = 61) = 4.89, ns. Thus, the socialization variables assessed did not predict childrens ability to maintain their prosocial lie during followup questioning.
Moral understanding
Childrens classications and evaluations of statements made by story characters as lies or truths
were analyzed using 3 (Age Group: preschool, early elementary, or late elementary)  4 (Type of

Table 1
Means (and standard deviations) of socialization variables by type of child.
Type of child

PSDQ-SFa

SEFQb
c

Lie-teller
Truth-teller
a
b
c

Authoritative

Authoritarian

Permissive

Positivec

Negative

4.25 (0.46)
4.00 (0.42)

1.77 (0.35)
1.79 (0.47)

2.06 (0.59)
2.05 (0.62)

7.57 (0.86)
7.67 (0.79)

4.35 (1.09)
4.32 (1.14)

Parenting Styles and Dimensions QuestionnaireShort Form.


Self-Expressiveness in the Family Questionnaire.
Means in this column differ signicantly between lie- and truth-telling children (p < .05).

385

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392


Table 2
Means (and standard deviations) of socialization variables by plausibility of childrens answers to follow-up questions.
Type of answer

Plausible
Not plausible
a
b

PSDQ-SFa

SEFQb

Authoritative

Authoritarian

Permissive

Positive

Negative

4.19 (0.49)
4.47 (0.29)

1.77 (0.35)
1.68 (0.32)

2.13 (0.62)
1.80 (0.38)

7.57 (0.79)
7.56 (0.86)

4.35 (1.13)
4.32 (1.26)

Parenting Styles and Dimensions QuestionnaireShort Form.


Self-Expressiveness in the Family Questionnaire.

Child: high-cost condition prosocial liar, low-cost condition prosocial liar, high-cost condition blunt
truth-teller, or low-cost condition blunt truth-teller)  2 (Sex)  2 (Character Veracity: lie or truth)
analyses of variance (ANOVAs), with veracity as a repeated measure.
First, childrens accuracy at identifying prosocial lies and truths was examined as the dependent
variable. Analyses revealed a signicant effect for veracity, F(1, 80) = 10.78, p = .002, partial g = .12,
and Veracity  Age Group interaction, F(2, 80) = 5.08, p = .004, partial g = .11. Overall, of a possible
score of 2.00, children were better able to identify prosocial lies (M = 1.82, SD = 0.62) than prosocial
truths (M = 1.39, SD = 0.82). Preschoolers lower accuracy was due to their difculties in classifying
prosocial truths (M = 0.60, SD = 0.85) compared with prosocial lies (M = 1.61, SD = 0.76), whereas older
children were able to classify prosocial truths (early elementary: M = 1.81, SD = 0.54; late elementary:
M = 1.75, SD = 0.51) and prosocial lies (early elementary: M = 1.84, SD = 0.63; late elementary:
M = 2.00, SD = 0.00) accurately. Xu and colleagues (2010) also reported that nearly all children in their
sample between 7 and 11 years of age were able to identify lies and truths correctly. Similar to Bussey
(1999), the familiarity with prosocial lie-telling scenarios in ones everyday lives may have contributed to the overall better identication of lies by children in all age groups, even more so for the early
and late elementary school children.
Another repeated-measures ANOVA was conducted with childrens evaluations of the prosocial
scenarios as the dependent variable. There was a signicant main effect for veracity, F(1, 80) = 9.40,
p = .003, partial g = .105, and for age, F(2, 80) = 8.69, p < .001, partial g = .18. Overall, with the most positive evaluation being 6.00, prosocial truths were evaluated as being better (M = 3.45, SD = 1.35) than
prosocial lies (M = 2.81, SD = 1.20). Preschoolers tended to be harsher in their evaluations (M = 2.50,
SD = 1.34) than early and late elementary school children (M = 3.27, SD = 1.19, and M = 3.05,
SD = 1.25, respectively). This nding was again consistent with Xu and colleagues (2010), who also reported that prosocial lies became less negative between the preschool and elementary school years.
Next, childrens ratings of the vignette protagonists feelings in the prosocial lie and truth scenarios
were compared. There was a signicant main effect of veracity, F(1, 80) = 6.69, p = .017, partial g = .87,
with children indicating that the protagonist was more likely to feel better when he or she told the
truth (M = 4.88, SD = 1.71) than when telling a lie (M = 4.14, SD = 1.91). There was also a signicant
Veracity  Type of Child interaction, F(3, 80) = 5.32, p = .004, partial g = .07. Post hoc analyses revealed
that in the high-cost condition, blunt truth-tellers thought that the protagonist would feel better when
he or she told the truth than when telling a lie, p < .01, whereas prosocial lie-tellers thought the protagonist would feel good regardless of whether he or she told the truth or a lie (see Fig. 3). Similarly,
for children in the low-cost condition, blunt truth-tellers thought that the protagonist would feel better when he or she told the truth than when telling a lie, p < .05, whereas prosocial lie-tellers thought
that the protagonist would feel good regardless of whether he or she told the truth or a lie (Fig. 3).
However, prosocial lie-tellers in the low-cost condition rated the protagonists feelings less positively
than prosocial lie-tellers in the high-cost condition, p < .05. Therefore, although both groups of children followed the same pattern of results for their evaluations of the protagonists feelings, children
in the low-cost condition were less positive about the protagonists lying behavior compared with
their high-cost counterparts.
A nal repeated-measures ANOVA was conducted to compare childrens ratings of the vignette recipients feelings in the prosocial lie and truth scenarios. There was a signicant Veracity  Type of
Child interaction, F(1, 80) = 9.10, p = .017, partial g = .15. Whereas blunt truth-tellers in the high-cost

386

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

Fig. 3. Child ratings of the protagonists feelings about telling the truth or telling a lie, ranging from 1 (very very bad) to 6 (very
very good): Veracity  Type of Child interaction.

condition thought that the recipient would feel the same if told the truth (M = 4.38, SD = 2.32) or a lie
(M = 4.48, SD = 2.19), prosocial lie-tellers thought that the recipient would feel better if told a prosocial
lie (M = 5.68, SD = 1.76) than if told the truth (M = 3.52, SD = 1.96), p < .01. In the low-cost condition,
blunt truth-tellers thought that the recipient would feel better if told the truth (M = 5.15, SD = 2.68)
than if told a lie (M = 4.37, SD = 2.65), p < .05, whereas prosocial lie-tellers thought that the recipient
would feel better if told a prosocial lie (M = 4.97, SD = 2.48) than if told the truth (M = 4.03, SD = 2.24),
p < .05 (see Fig. 4). Hence, childrens evaluations of the recipients feelings differed depending on
whether children were assigned to the high- or low-cost condition and whether they themselves told
a prosocial truth or a lie. The prosocial lie-tellers in both conditions rated the lie scenarios as more

Fig. 4. Child ratings of the recipients feelings about being told the truth or being told a lie, ranging from 1 (very very bad) to 6
(very very good): Veracity  Type of Child interaction.

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

387

positive than the truth scenarios; however, the difference was even greater for children in the highcost condition. On the other hand, it was only the blunt truth-tellers in the low-cost condition who
rated the truth scenarios as more positive than the lie scenarios from the recipients perspective.
General discussion
The current studies investigated childrens prosocial lie-telling behaviors and variables that might
increase the likelihood of predicting whether children would tell a prosocial lie in a politeness situation. Overall, 68% of child participants, ranging in age from preschoolers to late elementary school children, lied and said that they liked a disappointing gift given by another. This rate was similar to
previous research ndings on childrens prosocial lie-telling behaviors (e.g., Talwar & Lee, 2002b;
Talwar, Murphy et al., 2007; Xu et al., 2010). Across the two studies, the motivational contexts of
childrens prosocial lying were manipulated, revealing that children were more likely to lie (81%)
when they perceived a low personal cost. In addition, socialization variables, namely parenting
practices and family emotional expressivity, were related to childrens prosocial lie-telling behavior.
Finally, childrens understanding and evaluations of different moral scenarios, the protagonists
feelings, and the recipients feelings were examined in relation to childrens prosocial lie-telling
behavior. Children better identied prosocial lies as compared with prosocial truths, yet they
evaluated the truths as being more acceptable. Children who told prosocial lies evaluated the moral
behavior of others and the recipients feelings differently from children who told prosocial truths.
These ndings are discussed in detail below.
Motivation
To determine whether children were motivated to tell prosocial lies for self-oriented outcomes (i.e.,
to avoid negative consequences) or other-oriented outcomes (i.e., to be polite), the motivational contexts of a politeness situation were manipulated. Across two studies, childrens lie-telling behavior
varied in these different motivational contexts. For children in both studies, when there was a high
personal cost, children were less inclined to tell a lie. However, when there was little cost to themselves, children were more likely to lie to the gift-giver and tell her that they liked the gift. Furthermore, this tendency to lie generally increased with age. As expected, older children were more
likely than preschoolers to tell a prosocial lie that was other-oriented in the high-cost condition
(Fig. 2). Therefore, with increasing age, there was a developmental change in childrens tendency to
tell a truly prosocial lie. Nevertheless, it is important to note that there were still quite a few children
who lied in the high-cost condition in both studies. This indicates that even though young children
may be more likely to tell a prosocial lie for self-oriented purposes, there are still at least some
who choose to behave in a prosocial way (and lie for other-oriented purposes). Childrens prosocial
lying even when there is a cost to themselves may reect the inuence of socialization, whereby children are exposed to others telling lies in politeness situations from an early age and are taught about
the importance of being polite in such situations in order to spare anothers feelings (Talwar, Murphy
et al., 2007). Furthermore, the fact that at least some young children and more older children lied in
the high-cost situation may also be related to childrens developing perspective taking and empathy
abilities. These skills are necessary when deciding whether or not to tell a prosocial lie in a politeness
situation, and they both are more developed in older children compared with younger children (e.g.,
Findlay, Girardi, & Coplan, 2006; Newcombe & Huttenlocher, 1992).
Xu and colleagues (2010) also found that older children were more likely to tell a lie for prosocial
reasons, as did Talwar, Murphy et al. (2007). Taken together with the current results, these ndings
suggest that childrens tendency to lie in politeness situations is inuenced at least in part by motivational factors and the ability to reconcile acting for self- or other-oriented purposes. Whereas younger
children were motivated to lie in such situations for self-oriented outcomes and to avoid negative consequences to themselves (i.e., losing out on a good gift), older children were increasingly motivated to
tell lies for prosocial other-oriented outcomes even when contrary to their self-interests. Because a
signicant age difference was found in Study 2, which included a group of young preschoolers, and

388

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

not in Study 1, which included only elementary school children, the argument that preschoolers have
difculties in reconciling the contradictory rules of interpersonal communication in politeness situations is further supported. By early elementary school age, children seem better able to understand
and reconcile the rules governing interpersonal communication and, furthermore, to use these rules
adaptively in social situations to maintain amicable relations with others.
Childrens ability to conceal their lies
Based on previous studies, it was hypothesized that older children would be better able to maintain
their lies during follow-up questioning, demonstrating an understanding that they must be aware of
their verbal statements following an initial lie to avoid detection. In both studies, childrens age was
signicantly associated with their ability to provide plausible explanations for their lies regardless of
motivational context. Consistent with previous studies (Polak & Harris, 1999; Talwar & Lee, 2002a;
Talwar, Murphy et al., 2007), among prosocial liars, older children were better able to adequately conceal their lies and conceal information that would expose the truth. It may be that the cognitive skills
(i.e., memory capacity, inhibitory control, planning) necessary to maintain an initial lie statement during follow-up questioning have not yet been fully developed in young children (Talwar & Lee, 2008).
As such, young childrens true thoughts or feelings are more easily revealed and, thus, their lies are
more often detected.
In Study 1, girls were better at concealing their lies when compared with boys. However, no sex
differences were found in Study 2. These ndings are consistent with the mixed results in the literature concerning sex differences in childrens ability to use verbal display rules in politeness situations
(e.g., Cole, 1986; Davis, 1995; Garner & Power, 1996; Saarni, 1984). It could be that sex role socialization factors were inuencing their behaviors (Saarni, 1984). Girls may receive greater socialization to
be polite and to inhibit negative expressions, and they may be overall more concerned about hurting
anothers feelings, as compared with boys (e.g., Fabes et al., 1994; Fuchs & Thelen, 1988; Garner &
Power, 1996). Therefore, girls may learn earlier than boys that they need to follow up their lies with
other congruent statements. It may also be that there are nonverbal differences in girls and boys ability to maintain their lies (e.g., Davis, 1995; Fabes et al., 1994). Another possibility may be that girls are
better than boys at concealing their lies due to better developed inhibitory control abilities (e.g.,
Bjorklund & Kipp, 1996). Further research is needed to examine the differential impacts of socialization factors on girls and boys lie-telling behavior.
Social variables
Although previous studies have suggested a need to examine social variables and their relation to
childrens lie-telling behavior (e.g., Lee & Ross, 1997; Talwar & Lee, 2008), no study has empirically
examined this relationship. The current studies found that two social variables (i.e., parenting styles
and family emotional expressivity) were related to childrens prosocial lie-telling behavior regardless
of motivational context. Children who told a prosocial lie had parents who were more authoritative in
their parenting styles than did blunt truth-tellers. This nding was consistent with Robinson and colleagues (1995), who showed that parents using an authoritative parenting style helped their children
to develop instrumental competence that in turn fosters effective social skill development. Being able
to politely say to another that a gift he or she gave you was something you liked when in fact you did
not like it (and, thus, tell a prosocial lie) is an important social skill necessary to foster amicable relations with others.
The current studies also found that how frequently emotions are expressed within the family was
related to childrens prosocial lie-telling behavior. Specically, when compared with children who told
the truth about their dislike of the disappointing gift, prosocial lie-tellers had families who expressed
positive emotions less frequently within the family. This nding was somewhat unexpected given that
it was the reverse of the predicted relationship and did not differ with regard to childrens motivational context. Talwar, Murphy et al. (2007) reported that when parents discussed the feelings of others with their children following their receipt of a disappointing gift, children subsequently were more
likely to tell a convincing prosocial lie when asked about the gift by the gift-giver. Talking about

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

389

feelings and expressing emotions authentically would correspond to high scores on the family expressiveness scale. However, when telling prosocial lies, one must refrain and dissemble the expression of
authentic emotion; thus, families that foster appropriate use of prosocial lie-telling for other-oriented
motives and model this behavior for their children might score lower on the family expressiveness
scale. Similarly, as Jones, Abbey, and Cumberland (1998) pointed out, it may be that the relationship
between positive emotional expressiveness and prosocial display rule use is not as strong as negative
types of emotional expressiveness within the family and general display rule use.
The current studies ndings suggest that more research is needed to better understand how social
factors inuence the development of childrens prosocial lying in different motivational contexts. As of
this time, this is an area that has received little attention. Employing different methodological techniques to examine socialization variables (e.g., using a longitudinal design, interviewing parents) related to childrens prosocial behaviors throughout childhood may provide additional insights into
the developmental paths of prosocial lie-telling and the role of social variables in childrens lie-telling
behavior.
Moral understanding and evaluations
In Study 2, childrens moral understandings and evaluations of prosocial scenarios and their actual
prosocial lie-telling behaviors were examined. Consistent with previous studies (Bussey, 1992, 1999;
Talwar et al., 2002), children were more accurate at identifying prosocial lie scenarios as compared
with prosocial truth scenarios, with the youngest children having the most difculty in identifying
the prosocial truth scenarios as such. It may be that children are better able to accurately identify lies
rather than truths in prosocial situations because prosocial lies (which are socially acceptable) are
more frequently observed by children in daily life compared with blunt truth-telling in such situations. That is, most children learn early in their development that telling the blunt truth in certain social situations (e.g., when receiving a gift they do not like) might be perceived as being rude rather
than appropriate, and could ultimately jeopardize any amicable relation between themselves and
the gift-giver.
In the current studies, children evaluated prosocial truths as more acceptable than prosocial lies.
Older children gave higher ratings of acceptability for prosocial lies compared with preschool children,
who generally appeared to have difculty in understanding the nuances of the prosocial scenarios and
rated them more harshly. Bussey (1999) also reported that the younger children in her sample (4 years
of age) evaluated prosocial scenarios more negatively compared with older children. Furthermore,
Walper and Valtin (1992) reported that it was only toward the end of elementary school that children
were best able to identify and evaluate prosocial lie scenarios as less negative than other types of lie
scenarios (i.e., antisocial lie scenarios). Thus, with increasing age, children become better able to identify and evaluate prosocial truth and lie scenarios in a way that is comparable to adults.
The current studies also examined childrens evaluations of the protagonists and recipients feelings in prosocial scenarios in relation to their actual behavior in a politeness situation. In general, most
children reported that the protagonist would feel better if he or she told the truth rather than lying in
the prosocial scenarios, contrary to the results reported by Heyman and colleagues (2009). In addition,
children in the current studies generally reported that the recipient would feel okay regardless of
whether they were told the truth or a lie. Signicant interactions between childrens moral evaluations of the protagonists and recipients feelings and a childs prosocial lying were also found. Specifically, truth-tellers (in both motivational contexts) differentiated the scenarios in a manner consistent
with their own behaviors; truth-telling was evaluated as better than lie-telling in the prosocial scenarios. Lie-tellers were more equivocal; whereas those from the high-cost condition rated the protagonist
in the lie scenario as feeling better than did liars from the low-cost condition, there was little difference in lie-tellers ratings of the lie and truth scenarios overall.
Results differed slightly for perceptions of the lie recipient. Lie-tellers indicated that the recipient
would feel better if told a lie rather than the truth, with lie-tellers in the high-cost condition rating lies
more positively than those in the low-cost condition. Here truth-tellers were more equivocal; whereas
truth-tellers in the high-cost condition did not rate the recipient as feeling better or worse if told a
truth or a lie, those in the low-cost condition indicated that the recipient would feel better about being

390

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

told the truth rather than a lie. Overall, lie-tellers rated the outcome of lies told to another as more
positive than did truth-tellers. Thus, childrens perceptions of appropriate social behavior in politeness
contexts seem, in some instances, to comport with their actual lie-telling (or truth-telling) decisions.
The current studies examination of childrens lie-telling and moral evaluations is unique in several
ways compared with previous studies (e.g., Broomeld et al., 2002; Bussey, 1999). Most studies examining childrens lie-telling behavior and their moral understanding of lies have focused on childrens
lies to conceal a transgression, either for themselves or for another (Talwar & Lee, 2008; Talwar
et al., 2002, 2004). These studies have found a limited relation between childrens lie-telling behavior
and their moral evaluations. The current studies specically examined childrens prosocial lie-telling
behavior and evaluations of prosocial lies and truths. Knowing that lies told to conceal a transgression
and those told for prosocial purposes differ in their rate of occurrence and development, it is not surprising that the relation between childrens actual behavior and their understanding of different lie
scenarios differs as well. Only one previous study has specically examined the relation between prosocial lies and moral evaluations. Similar to the current studies, Xu and colleagues (2010) reported
nding that Chinese lie-tellers moral knowledge and action were related. However, unlike Xu and colleagues, the current studies also examined childrens prosocial lie-telling behavior in relation to their
evaluations of the protagonists and recipients feelings in prosocial vignette scenarios. Considering
the evaluative reactions of others when both telling and being told a prosocial truth or lie is important
because it reveals childrens awareness of moral standards in interpersonal situations from the perspective of another (Bussey, 1999). If children anticipate self-disapproval when telling the truth and
self-approval when telling a lie in prosocial scenarios, they are more likely to act in ways that are congruent with such beliefs (Bandura, 1986) because it is arguably the anticipation of such self-evaluative
reactions that ultimately guides ones social behavior.
The current studies provided empirical support for the relation between behavior and moral understanding from the perspectives of others in prosocial situations. Specically, prosocial truth-tellers
provided more positive ratings of the protagonists feelings when he or she told the truth in a vignette
scenario, whereas prosocial lie-tellers rated the recipients feelings as more positive when told a lie in
a vignette scenario. Although it is important to acknowledge that all children partook in the lie scenario at least a week before they evaluated the different vignette scenarios, previous research (e.g.,
Talwar et al., 2002; Xu et al., 2010) has found that childrens evaluations are not affected by whether
they are rst presented with the vignettes or the lie scenario. Moreover, all children in Study 2, regardless of their behavior in the disappointing gift paradigm, evaluated truth-telling in the prosocial vignettes as more acceptable than lie-telling. Nevertheless, because all children participated in the
behavioral paradigm prior to the conceptual questioning, the direction of the relationship between
behavior and moral evaluations needs to be explored further in future studies. In the current studies,
the aim was to capture childrens spontaneous natural prosocial lie-telling behavior. To do this, any
prompting or cueing to children about the motive of the research needed to be minimized or eliminated. If the conceptual questioning had come prior to the behavior, children might have been more
attuned to lie-telling behavior in general and possibly altered their natural tendencies. Nevertheless,
these ndings suggest that more research is needed to better understand the relationship between
moral action and moral understanding.

Conclusions
Children in the current studies were most likely to tell prosocial lies when there was a low perceived personal cost. However, the tendency to tell prosocial lies that were other-oriented in their outcome increased with age, especially for children in the high-cost condition where the stakes of telling
lies were greater and children had something to lose by lying. Indeed, for children in the low-cost condition, no age difference was found across the age range, yet more preschoolers told prosocial lies that
had a self-oriented outcome. Therefore, it seems that only among older children was prosocial lie-telling truly prosocially motivated for other-oriented purposes. Moreover, when children told prosocial
lies, older children were better able to maintain their lies during follow-up questioning compared with
younger children, indicating greater maturity in their mastery of deception. In addition, compared

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

391

with truth-tellers, prosocial liars were more likely to have parents who had an authoritative parenting
style yet reported less frequent expressions of positive emotions within the family regardless of motivational context. Finally, childrens prosocial lying or blunt truth-telling (in the different motivational
contexts) was related differentially to their moral understanding and evaluations of vignette scenarios
and to their perceptions of each story protagonists and recipients likely feelings. This research is a
rst step toward clarifying the mechanisms underlying the development of childrens prosocial lietelling abilities and suggests that further investigation into the role of social and cognitive factors is
needed. The current ndings begin to provide a better picture of the developmental trajectory of childrens prosocial lie-telling, their motivations for telling such lies, and their knowledge about rules governing interpersonal communication.
References
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Bjorklund, D. F., & Kipp, K. (1996). Parental investment theory and gender differences in the evolution of inhibition mechanisms.
Psychological Bulletin, 120, 163188.
Broomeld, K. A., Robinson, E. J., & Robinson, W. P. (2002). Childrens understanding about white lies. British Journal of
Developmental Psychology, 20, 4765.
Brown, P., & Levinson, S. C. (1987). Politeness: Some universals in language usage. New York: Cambridge University Press.
Bussey, K. (1992). Lying and truthfulness: Childrens denitions, standards, and evaluative reactions. Child Development, 63,
129137.
Bussey, K. (1999). Childrens categorization and evaluation of different types of lies and truths. Child Development, 70,
13381347.
Chandler, M., Fritz, A. S., & Hala, S. (1989). Small-scale deceit: Deception as a marker of two-, three-, and four-year-olds early
theories of mind. Child Development, 60, 12631277.
Cole, K., & Mitchell, P. (1998). Family background in relation to deceptive ability and understanding of the mind. Social
Development, 7, 181197.
Cole, P. (1986). Childrens spontaneous control of facial expression. Child Development, 57, 13091321.
Crossman, A. M., Talwar, V., Arruda, C., Brunet, M., Buonaugurio, A., & Runo, K. (2010). Whats in a lie? Perceptions of lying
among children and adults. City University of New York. Unpublished manuscript.
Davis, T. L. (1995). Gender differences in masking negative emotions: Ability or motivation? Developmental Psychology, 31,
660667.
Denham, S. A., & Grout, L. (1992). Mothers emotional expressiveness and coping: Relations with preschoolers social-emotional
competence. Genetic, Social, and General Psychology Monographs, 118, 75101.
DePaulo, B. M., & Bell, K. L. (1996). Truth and investment: Lies are told to those who care. Journal of Personality and Social
Psychology, 71, 703716.
DePaulo, B. M., & Jordan, A. (1982). Age changes in deceiving and detecting deceit. In R. S. Feldman (Ed.), Development of nonverbal behavior in children (pp. 151180). New York: Springer-Verlag.
DePaulo, B. M., & Kashy, D. A. (1998). Everyday lies in close and casual relations. Journal of Personality and Social Psychology, 74,
6379.
Fabes, R. A., Eisenberg, N., Karbon, M., Bernsweig, J., Speer, A. L., & Carlo, G. (1994). Socialization of childrens vicarious emotional
responding and prosocial behavior: Relations with mothers perceptions of childrens emotional reactivity. Developmental
Psychology, 30, 4455.
Feldman, R. S., Jenkins, L., & Popoola, O. (1979). Detection of deception in adults and children via facial expressions. Child
Development, 50, 350355.
Findlay, L. C., Girardi, A., & Coplan, R. J. (2006). Links between empathy, social behavior, and social understanding in early
childhood. Early Childhood Research Quarterly, 21, 347359.
Fuchs, D., & Thelen, M. H. (1988). Childrens expected interpersonal consequences of communicating their affective state and
reported likelihood of expression. Child Development, 59, 13141322.
Gallup, G. Jr., (1970). Chimpanzees self recognition. Science, 167, 8687.
Garner, P. W., & Power, T. G. (1996). Preschoolers emotional control in the disappointment paradigm and its relation to
temperament, emotional knowledge, and family expressiveness. Child Development, 67, 14061419.
Grice, H. P. (1980). Studies in the way of words. Cambridge, MA: Harvard University Press.
Halberstadt, A. G., Cassidy, J., Stifter, C. A., Parke, R. D., & Fox, N. A. (1995). Self-expressiveness within the family context:
Psychometric support for a new measure. Psychological Assessment, 7, 93103.
Heyman, G. D., Sweet, M. A., & Lee, K. (2009). Childrens reasoning about lie-telling and truth-telling in politeness contexts.
Social Development, 18, 728746.
Jones, D. C., Abbey, B. B., & Cumberland, A. (1998). The development of display rule knowledge: Linkages with family
expressiveness and social competence. Child Development, 69, 12091222.
Keltikangas-Jaervinen, L., & Lindeman, M. (1997). Evaluation of theft, lying, and ghting in adolescence. Journal of Youth and
Adolescence, 26, 467483.
Lakoff, R. (1973). The logic of politeness: Or minding your Ps and Qs. In Papers presented at the ninth regional meeting of the
Chicago Linguistic Society (pp. 292305). Chicago: Chicago Linguistic Society.
Lee, K., & Ross, H. J. (1997). The concept of lying in adolescents and young adults: Testing Sweetsers folkloristic model. MerrillPalmer Quarterly, 43, 255270.
Lewis, M., & Brooks-Gunn, J. (1979). Social cognition and the acquisition of self. New York: Plenum.

392

M. Popliger et al. / Journal of Experimental Child Psychology 110 (2011) 373392

Lewis, M., Stanger, C., & Sullivan, M. W. (1989). Deception in 3 year olds. Developmental Psychology, 25, 439443.
Menard, S. (2002). Applied logistic regression analysis (2nd ed.), Quantitative Applications in the Social Sciences, No. 106).
Thousand Oaks, CA: Sage.
Newcombe, N., & Huttenlocher, J. (1992). Childrens early ability to solve perspective-taking problems. Developmental
Psychology, 28, 635643.
Peskin, J. (1992). Ruse and representations: On childrens ability to conceal information. Developmental Psychology, 28, 8489.
Polak, A., & Harris, P. L. (1999). Deception by young children following noncompliance. Developmental Psychology, 54, 561568.
Robinson, C. C., Mandleco, B., Olsen, S. F., & Hart, C. H. (1995). Authoritative, authoritarian, and permissive parenting practices:
Development of a new measure. Psychological Reports, 77, 819830.
Saarni, C. (1984). An observational study of childrens attempts to monitor their expressive behavior. Child Development, 55,
15041513.
Smetana, J. G. (1999). The role of parents in moral development: A social domain analysis. Journal of Moral Education, 28,
311321.
Sweetser, E. E. (1987). The denition of lie: An examination of the folk models underlying a semantic prototype. In D. Holland &
N. Quinn (Eds.), Cultural models in language and thought (pp. 4366). New York: Cambridge University Press.
Talwar, V., Gordon, H., & Lee, K. (2007). Lying in elementary school: Verbal deception and its relation to second-order belief
understanding. Developmental Psychology, 43, 804810.
Talwar, V., & Lee, K. (2002a). Development of lying to conceal a transgression: Childrens control of expressive behaviour during
verbal deception. International Journal of Behavioral Development, 26, 436444.
Talwar, V., & Lee, K. (2002b). Emergence of white lie-telling in children between 3 and 7 years of age. Merrill-Palmer Quarterly,
48, 160181.
Talwar, V., & Lee, K. (2008). Social and cognitive correlates of childrens lying behavior. Child Development, 79, 866881.
Talwar, V., Lee, K., Bala, N., & Lindsay, R. C. L. (2002). Childrens conceptual knowledge of lying and its relation to their actual
behaviors: Implications for court competence examinations. Law and Human Behavior, 26, 395415.
Talwar, V., Lee, K., Bala, N., & Lindsay, R. C. L. (2004). Childrens lie-telling to conceal a parents transgression: Legal implications.
Law and Human Behavior, 28, 411435.
Talwar, V., Murphy, S., & Lee, K. (2007). White lie-telling in children for politeness purposes. International Journal of Behavioral
Development, 31, 111.
Turiel, E. (1983). The development of social knowledge: Morality and convention. Cambridge, UK: Cambridge University Press.
Walper, S., & Valtin, R. (1992). Childrens understanding of white lies. In R. J. Watts, S. Ide, & K. Ehlich (Eds.), Politeness in
language: Studies in history, theory, and practice (Vol. 59, pp. 231251). Trends in Linguistics: Studies and Monographs.
Berlin: Mouton de Gruyter.
Xu, F., Boa, X., Fu, G., Talwar, V., & Lee, K. (2010). Lying and truth-telling in children: From concept to action. Child Development,
81, 581596.