Expectations about recipients' prosociality and me

文件名称: Expectations about recipients' prosociality and mental time travel relate to。。。。

所属分类: 教育

开发工具:

文件大小: 851kb

下载次数: 0

上传时间: 2019-09-03

提供者: yangta******

下载 (851kb)

不能下载？报告错误

详细说明：Expectations about recipients' prosociality and mental time travel relate to resource allocation in preschoolers.pdfY. Kumaki et aL /Journal of Experimental Child Psychology 167(2018)278-294 Several studies support the idea that preschoolers expect their friends to reciprocate prosociality compared with other peers or strangers; however, no experiment in a controlled setting has been con ducted to verify this Naturalistic observation studies have revealed that reciprocity among friends is stronger than that among non-friends in preschoolers' resource sharing and prosocial behavior Fujisawa, Kutsukake, Hasegawa, 2008; Howes, 1983). Paulus and Moore(2014)showed that chil- dren aged 4 and 5 years tended to expect that people would choose equal allocation more with their friends than with other peers in prosocial and sharing games from a third-person perspective. In thi experiment, children were introduced to three puppets: a protagonist, a friend and a peer. Then, they were asked to state which alternative the protagonist would choose when the recipient was the friend or peer. Children also decided their way of sharing with their own friends and peers; those who expected higher selective allocation from others tended to choose equal allocation more with their friends than with peers in their own allocation. Their findings suggest that preschoolers have the understanding that people are prosocial to their friends and such an understanding may lead to selec tivity based on friendship in their own allocation. In terms of reciprocity, however, children s expec tations of others' allocation to themselves, rather than understanding from the third-person perspective, is more important. According to the theory of reciprocity, children would decide their own way of allocation based on their expectations of recipients'prosociality toward them Given the previous evidence, it is possible that children aged 4.5-6 years selectively allocate to their friends based on the expectation of reciprocal prosociality. However, little is known about when children develop this expectation-based reciprocity in allocation. moreover, reciprocity is not the onl component of friendship. Researchers have proposed other elements of friendship such as mutual preference for interactions and sharing positive emotions(e.g, Howes, 1983). Previous research about selective allocation based on social relationship has also showed that reciprocity is not the only factor. For example, Renno and Shutts(2015) showed that selectivity to the same gender is based on both preference and expectation. Therefore in Study 1 we aimed to investigate the developmental change of selective allocation based on expectations and social relationship We asked children about their expectations of recipients'prosociality and their own allocation. This procedure enabled us to examine the relationship between expectations and children s own allocation directly. moreover, because many previous studies have paid attention to how the allocation of preschoolers differs between costly and non-costly situations(e.g, Burkart Rueth, 2013; Fehr et al, 2008), we conducted both prosocial (non-costly) and sharing(costly) trials. Thus, we also investigated differences in the effects of expectations and social relationship on childrens allocation between costly and non-costly situations If children selectively allocate resources based on the expectation of the recipients' prosociality, it might be considered as future-oriented behavior. Allocation of resources to one who is expected to be prosocial will lead to reciprocal interactions with the recipients in the future. Several studies have shown that preschoolers allocate resources in a manner that increases their future gains from recip rocal interaction Kenward, Hellmer, Winter, and Eriksson(2015 showed that 4-year-olds share their resources strategically to maximize their future gains. Similarly, Xiong, Shi, Wu, and Zhang(2016) showed that 5-year-olds shared more when the recipient had an opportunity to reciprocate than when he or she did not These studies may suggest that preschoolers consider future gains when they allo cate resources If selective allocation is motivated by a future reciprocal relationship with the recipients, the devel- opment of future thinking might be one of the important cognitive factors for the acquisition of this behavior. Future thinking has several aspects such as mental time travel, delay of gratification, plan- ning, and prospective memory(Atance Jackson, 2009). Delay of gratification refers to the ability to inhibit the desire to receive immediate rewards in order to obtain future rewards(mischel, Shoda, Rodriguez, 1989). Reciprocity is a long-term cooperative system that provides individuals with long- term gains through reciprocal interactions given that rewards from reciprocal interaction are not pro vided immediately. Thus, inhibition of immediate desire might be required to obtain long-term gains from the system of reciprocity Several studies have already investigated the relationship between delay of gratification and resource allocation in preschoolers. For example, Sebastian- Enesco and Warneken(2015 showed that 5-year-olds with a higher score on delay of gratification tended to share more resources in the alloca Y. Kumaki et aL /Journal of experimental Child Psychology 167(2018)278-294 tion task. Garon et al.(2011)examined the relationship between delay of gratification and selective sharing to friends in children aged 3-4.5 years. In their experiment, children were asked to share var ious toys with their friends and other peers. the results showed that the tendency to share more attractive toys with their friends was positively correlated with delay of gratification ability. These results suggest that the ability of delay of gratification might relate to selectivity in sharing situations. Mental time travel comprises reconstruction of personal events from the past and construction of possible events in the future apart from the influence of current mental states(Suddendorf Corballis, 1997). Children begin to acquire the ability of mental time travel by 4 years of age (Suddendorf Redshaw, 2013). Mental time travel may be an important factor of selectivity depending on friendship. The mental time travel ability might let children imagine what will happen to the rela tionship with the recipients if they share or do not share the resources little is known about the devel- opmental relationship between mental time travel and social behavior in preschoolers. Thus, in Study 2 we aimed to investigate the developmental relationship between selective allocation based on friend- ship and expectations about recipients'prosociality and future thinking The developmental change was investigated in Study 1, and the cognitive basis, especially future thinking, was investigated in Study 2 Study 1 In Study 1, we tested children aged 3. 5-6 years because several studies have shown the develop mental change of reciprocal allocation based on previous allocation by recipients aged 3 and 4 years and 5 and 6 years(e.g, House et al., 2013). Following the procedure established by moore(2009), we administered the prosocial game and sharing game for three recipients: a friend, another peer, and a stranger. In addition to this, participants were asked to predict the recipients'allocation to themsely before they chose how to allocate their resources. The aim of adding this procedure was to measure children s expectations about recipients'prosociality and examine its relationship with children's own allocation. We hypothesized that(a) children aged 5 and 6 years would tend to choose more equal alternatives in their own allocation when they expect that the recipients would choose the equal alternative rather than the selfish alternative; (b) on the other hand, for children aged 3. 5 and 4 years, expectations about recipients' allocation might not affect their own allocation given that previous studies have shown that allocation based on reciprocity is not observed in that age group(e.g House et al., 2013; Sebastian-Enesco Warneken, 2015) Method Participants The final sample consisted of 48 children aged 3. 5-6 years. They were divided into two groups based on age: 3.5 and 4 years(n=21, 15 boys and 6 girls; M=51.86 months, SD=4.25, range=44- 59)and 5 and 6 years(n=27, 15 boys and 12 girls; M=70.26 months, SD=6.79, range=60-79). A minimum sample size of 20 per group was determined based on a previous study that used a similar procedure (Paulus, 2016). Six children were excluded from the analyses because of errors in recording data (n=1)or refusing or being unable to identify the names of friends and other peers (n=5. All chil dren were recruited from urban kindergartens in Japan, were not suffering from mental or neurolog ical disorders, and had Japanese as their first language. This study was approved by the ethics committee of Kyoto University(No. 26-P-27) Informed written consent was obtained from the par ents of all the participants The experiment was conducted individually in a quiet room in the kindergartens the childrens responses were videotaped and recorded in real time by an experiment assistant. Following an estab lished protocol(Moore, 2009), children were first prompted to think of a child in their class with whom they often played (friend) and a child in their class with whom they rarely played (peer ) chil dren were asked to draw the friend and peer from memory Next, they were shown a photograph of a sex-matched child who was a stranger and were asked to draw a picture of him or her. Each drawing Y. Kumaki et aL /Journal of Experimental Child Psychology 167(2018) 278-294 was made on a different 10 x 14-cm card with color pencils. These drawings were used to indicate the recipients of the allocation games(explained in the following paragraph). Before the games started, the children were asked to identify the person in each drawing by name. all children in the final sam ple answered the questions correctly. In the allocation games, children were asked to choose how to allocate colored stickers based on two alternatives. They were given one white tray to keep on their table to save their stickers, and three trays were placed on the opposite side The pictures of their allocation partners (friend/peer/stranger were attached to each tray, and their stickers were saved in their respective trays Two types of games were administered a prosocial game and a sharing game (fehr et al., 2008; House, Henrich, Brosnan, Silk, 2012). In the prosocial game, children chose their allocation method from two alternatives: 1: 1(1 for themselves and 1 for the partner)and 1: 0(1 for themselves and o for the partner ) In the sharing game, children chose their allocation method from two alternatives: 1: 1 and 2: 0. The alternatives were shown to participants using two laminated cards in which two circles were drawn. In each trial, the children were first asked to state which alternative the partner would choose if he or she played (partner question to make them conscious of recipients' prosociality before their own allocation. In the prosocial game, for example, the experimenter asked, "If [name of friend, peer/stranger] had a chance to choose how to share which alternative would he/she choose? One for himself herself and zero for you or one for himself herself and one for you "The children pointed at the card showing the alternatives to respond. After the partner question, they were asked to choose their own alternative(own behavior ) In the sharing game, for example, the experimenter said, " What would you like to do? Two for you and zero for him her or one for you and one for him/ her "The par ticipants responded by pointing at the card. After that, children distributed the stickers to each tray according to their own choice. Then, the experimenter provided new stickers for the next trial with different colors or figures from the previous trial. Two trials were conducted for each condition The choice test was a 3(Partner Type: friend, peer, or stranger)x 2(Game Type: prosocial or share within-participant design, and every child was presented with 2 trials for each of the six conditions Therefore a total of 12 trials were conducted for each child The responses to the partner question and the own behavior question in all games were recorded and we computed the ratios of choosing 1: 1 out of 2 trials for both the partner question and own behavior question in each condition. The order of presentation of the partner and game type were counterbalanced across participants Statistical analyses All statistical analyses were conducted using R statistical software(Version 3. 2.5: R Core Team, 2016). to determine the factors that predicted childrens expectation in the partner question and choice in their own behavior, the data were analyzed with generalized linear mixed-effect models (GLMMs)using the Ime4 software package(Bates, Maechler, Bolker, Walker, 2015) Results Figs. 1 and 2 show the descriptive data for each game and question We first analyzed the predictive factors that explained childrens expectation in the partner question using GLMM analyses for each game. The GLMMs for each type of game were computed separately. The ratio of expecting 1: 1 out of 2 trials was included in the models as the dependent variable using binomial distribution with logit link. The age group(3.5-4 years or 5-6 years), partner's identity(friend, peer, or stranger), and their interaction were included in the models as independent variables. Participant identification(child iD)was included as a random effect to account for within-participant measures. for the prosocial game only the effect of age group was significant(Estimate =1. 13, SE=0.55, Z=2.05, p=.04), which means that children aged 5 and 6 years were more likely to expect 1: 1 than those aged 3.5 and 4 years. There were no significant effects of partner identity or the interaction between the partner and the age group For the sharing game, none of the factors predicted the children s expectations significantly. Next, GLMMs on the own behavior question were computed to determine the predictive factors of the children s responses in the own behavior The dependent variable was the ratio of choosing 1: 1 out of 2 trials in the own behavior question, and participant identification(child ID)was included as a random effect to account for the within-participant measures As possible predictors, age group partner's iden Y. Kumaki et aL /Journal of Experimental Child Psychology 167(2018)278-294 283 日3.5-and4-year 日5-and6 year-olds 8 Friend Peer Stranger Prosocial game 日3.5-and4-year-olds 口5-and6-year-olds Friend Peer Stranger Sharing game Fig 1. Mean numbers of trials in which children expected 1: 1 in the partner question in the prosocial and sharing games: (A) prosocial game; (B)sharing game. Error bars represent standard errors ity, the response to the partner question for each partner, and all their first-order interactions were included in the analyses. Hence there were many possible variables, and the final model was deter mined based on the forward stepwise model selection based on the akaike information criterion (AIC) Table 1 shows the final model for the prosocial game. The model including the interaction between age and expectation in the partner question improved data fit(AlC=316.)compared with the null model(AlC= 344.5). The main effect of age group(Estimate=1.71, SE=0.57, z=3.01, p=.003)and the interaction between age and expectation (Estimate =-3.63, SE=0.79, z=-4.58, p<.001) signifi cantly predicted the children s own choice. The main effect of expectation in the partner question was not significant The significant effect of age group indicates that children aged 5 and 6 years were more likely to choose 1: 1 than those aged 3. 5 and 4 years. In terms of the interaction between age group and the partner question, we used a generalized linear model (glm) for each age group to examine the difference of the effect of the partner question in the two age groups. The ratio of choos- ing 1: 1 in the own behavior question was the dependent variable using binomial distribution with logit link, and the ratio of expecting 1: 1 in the partner question was the independent variable. In chil dren aged 3. 5 and 4 years, we could not find a clear relationship between the children 's response to 284 Y. Kumaki et al /Journal of Experimental Child Psychology 167 (2018)278-294 ■3.5-and4year-olds 口5-and6 year-olds ① 6 E o Friend Peer Stranger Prosocial game 口3.5-and4-year-olds 日5-and6-year-olds Friend Peer Stranger Sharing game Fig. 2. Mean numbers of trials in which children chose 1: 1 in the own behavior question in the prosocial and sharing games: (A) prosocial game; (B) sharing game. Error bars represent standard errors Table 1 Final model for children's choice in the own behavior question in the prosocial game Facto Estimate z Value Age group 1.71 0.57 03 Expectation in partner question 0.69 0.49 140 Age group x Expectation 3.63 0.79 4.58 <,001 Random effect(child ID) 0.27 0.36 -0.77 44 p p Y. Kumaki et aL /Journal of Experimental Child Psychology 167(2018)278-294 the partner question and their own behavior question(estimate =-0.42, SE=0.41, 2=-1.02, p=31) (see Fig 3A). On the other hand, in children aged 5 and 6 years, those who expected 1: 1 in the partner question tended to choose 1: 1 in the own behavior question(Estimate =2.61, SE=0.50, Z=5.23, p< 001)(Fig.3B The final model for the sharing game is shown in Table 2. In the first step the model that includes the interaction between age group and the partner question improved data fit(alC=197.3) compared with the null model (alc= 204.9). In the second step, including the main effect of the partner's identity further improved the model (alc= 193.0). The final model showed a significant main effect of partner (peer Estimate=-1.48, SE=0.56, Z=-2.65, p=008 )and interaction between age group and expecta- tion(estimate=3.87, SE=1.58, Z=2.44, p=014). The main effect of partner indicates that the ratio of choosing 1: 1 was smaller when the partner was a peer(M= 48, SD=.30 )compared with a friend(m 64, SD=37)or stranger (M=51, SD=35 ) In terms of the interaction between age group and the partner question, GLMs for each age group were conducted separately to examine the difference of the effect of partner question in the two age groups. In children aged 3. 5 and 4 years, there was no significant relationship between the children's response to the partner question and the own behavior friend peer EoEaooo6g ◆ stranger 0.0 02 04 08 Ratio of expecting 1: 1 in partner question 3.5-and4 friend o56 stranger 56 0.0 0.2 04 0.6 081 Ratio of expecting 1: 1 in partner question 5-and 6-year-olds Fig 3. Relationship between responses to the partner question and the own behavior question in the prosocial game: (A)3.5 and 4-year-olds;(B)5-and 6-year-olds Y. Kumaki et aL /Journal of Experimental Child Psychology 167(2018) 278-294 Table 2 Final model for children,s choice in the own behavior question in the sharing game Factor Estimate z Value Age group 0.29 1.20 Partner(peer) 0.56 008 tner(stranger) 0.92 0.50 Expectation in partner question 1.73 1.42 16 Age Group x Expectation 3.87 1.58 2.44 Random effect(child ID) -2.09 0.98 2.14 03 p<.05 p< e friend ◆ stranger 0.002 0.4 06 08 10 Ratio of expecting 1: 1 in partner question 3.5-and 4-year-olds friend 2505305只956o856° ◆ stranger 8 0002 06 08 1.0 Ratio of expecting 1: 1 in partner question 5-and 6-year-olds Fig 4. Relationship between responses to the partner question and the own behavior question in the sharing game: (A)3.5-and 4-year-olds; (B)5-and 6-year-olds question(Estimate=-0.32, SE=0.73, z=0.44, p=.66)(see Fig 4A). In children aged 5 and 6 years, those who expected 1: 1 in the partner question tended to choose 1: 1 in the own behavior question ( Estimate=1.29,SE=0.41,z=3.16,p=.002)(Fig.4B) Y. Kumaki et aL /Journal of Experimental Child Psychology 167(2018)278-294 Discussion Our findings indicated that in the prosocial game the effects of expectations about recipients'allo- cation on children's own allocation were different for the two age groups. Children aged 5 and 6 years tended to choose an equal allocation for individuals who they expected would reciprocate the proso ciality. On the other hand, in children aged 3. 5 and 4 years, expectations about recipients'allocation did not significantly predict their own allocation. The result of the sharing game was similar to that of the prosocial game; hence age influenced expectations about recipients'allocation In children aged 3.5 and 4 years, expectations of the recipients'allocation did not predict their decisions, whereas chil dren aged 5 and 6 years were more likely to choose equal allocation when they expected an equal allo cation by the recipients These results suggest that reciprocal allocation based on expectations about recipients'prosociality might be acquired between 4 and 5 years of age. Studies have shown that reciprocal allocation based on previous allocation by the recipients is acquired at around 5 years( house et al, 2013; Sebastian Enesco Warneken, 2015). Our results suggest that reciprocity based on this expectation might also be acquired during the same period The noticeable difference between the prosocial game and the sharing game was the ratio of choos ing equal allocation. In the prosocial game the ratio of choosing 1: 1 was much larger than in the shar ing game. In the prosocial game, the children did not need to consider their own gains because they could get the same number of stickers regardless of the alternatives they chose. On the other hand, they needed to take their self-interest into account in the sharing game. Previous research has repeat edly revealed that resource allocations of preschoolers were strongly affected by their self-interest e.g., Blake, McAuliffe, Warneken, 2014; Blake Rand, 2010; Kogut, 2012; Smith, Blake, Harris 2013). Children might choose the selfish alternative more often in the sharing game than in the proso cial game for this reason Burkart and Rueth(2013) pointed out that the prosocial game entails higher attentional demands compared with the sharing game and this lowers the performance of young children in the prosocial game. According to these authors, children may choose the way of allocation less carefully in the prosocial game than in the sharing game because their own gains are the same between the two alter natives in the prosocial game. However, the children in Study 1 were more likely to choose equal allo cation in the prosocial game than in the sharing game. This result suggests that children paid attention to both the recipients gains and their own gains when they decided their way of allocation not only in the sharing game but also in the prosocial game. Moreover, the effect of expectation and its develop mental change was consistent between the two allocation tasks. Thus, the prosocial game worked as the sharing game in terms of the effect of expectations in the current study Study 2 The purpose of Study 2 was to investigate the developmental relationship between future thinking and selectivity in resource allocation. We focused on children aged 5 and 6 years because the effect of expectations and social relationship with the recipients on childrens own behavior were observed only in 5-and 6-year-olds in Study 1. Among the facets of future thinking, we focused on two facets delay of gratification and mental time travel Delay of gratification may relate to selective allocation through the process of controlling childrens own behavior for future gains. It includes aspects of both future thinking and inhibition. Previous stud ies have shown that the ability of delay of gratification relates to costly sharing in preschoolers. For example, Moore, Barresi, and Thompson(1998)showed that delay of gratification and performance in a costly allocation task, which is very similar to the sharing game, were positively correlated in 3-and 4.5-year-olds. Moreover, Garon et al. (2011) revealed a positive correlation between delay of gratification and selectivity in costly resource sharing. Costly sharing tasks demand higher ability to inhibit self-interest. Thus, this relationship can be explained by inhibition of self-interest rather than by future thinking. To clarify whether the relationship between delay of gratification and allocation is based on the aspect of future thinking or only on inhibition, it is necessary to examine the relationship

(系统自动生成,下载前可以参看下载内容)