Brief Report
The influence of model status on the tendency of young children to over-imitate

https://doi.org/10.1016/j.jecp.2013.05.004Get rights and content

Abstract

The current study aimed to integrate the trust and over-imitation literatures by allowing groups of 5-year-old children to view one of four adult models, differing in their level of status (high or low), retrieve a reward from inside a transparent puzzle box. Each of the models performed a sequence of tool actions on the box before retrieving the reward. These actions varied according to their causal necessity, with some of the actions being causally necessary for reward retrieval and others being causally irrelevant. The results suggest that young children are selective copiers, reproducing the irrelevant tool actions most frequently after having viewed the high-status models. It is suggested that this bias toward rank-ordered copying is likely to be a strategy favored by natural selection because the behaviors displayed by high-status individuals are often the behaviors that are locally adaptive and may, by extension, provide copiers with a selective advantage within their environment.

Introduction

The aim of the current study was to integrate two research strands currently stimulating a great deal of interest within the developmental literature: trust and over-imitation. With respect to trust, recent studies have shown that children as young as 4 years do not trust other individuals indiscriminately; instead, they operate selectively, choosing to rely on information provided by (a) individuals who are familiar caregivers (Corriveau & Harris, 2009), (b) individuals who are culturally prototypical (e.g., by sharing the same accent as the children) (Kinzler, Corriveau, & Harris, 2011), and (c) individuals who have demonstrated themselves to be accurate in a domain (e.g., by naming a familiar object correctly) as opposed to individuals who have shown themselves to be less trustworthy in the same domain (Birch et al., 2008, Clément et al., 2004, Harris, 2007, Harris and Corriveau, 2011, Koenig et al., 2004).

The selectivity witnessed in the trust literature appears, at first glance at least, to be at odds with preschool children’s extreme lack of selectivity in other spheres, most notably that of tool/object use. Over the past 8 years, there has been a great deal of interest in a curious phenomenon that has become known as over-imitation (e.g., Horner and Whiten, 2005, Kenward, 2012, Kenward et al., 2011, Lyons et al., 2011, Lyons et al., 2007, McGuigan et al., 2007, Nielsen, 2006, Nielsen and Blank, 2011, Nielsen and Tomaselli, 2010). The typical scenario in these studies is that observers watch an adult model demonstrate a task using tool actions that are superfluous (e.g., performing causally unnecessary taps before retrieving a reward from a box) to successful task completion before being allowed to attempt the task themselves. The usual outcome in such studies, sometimes in the face of strong demands to do otherwise (e.g., Lyons et al., 2007), is that the observers (irrespective of whether they are adults or children) copy the unnecessary actions with an extremely high level of fidelity (McGuigan, 2012, McGuigan et al., 2012).

However, a recent study, which provided a tentative first step in linking over-imitation and trust, found that 3- and 5-year-olds were significantly less likely to perform the unnecessary tool actions performed by a 5-year-old child model than those performed by an adult model (McGuigan, Makinson, & Whiten, 2011). Similarly, causally irrelevant tool actions did not spread from child to child when the same task was presented in a linear diffusion format (McGuigan & Graham, 2010). One possible reason for this differential performance is that the observers deemed the adult models to be more experienced, and subsequently more knowledgeable, in the domain of tool use than the child models. This “pickiness” suggests that children are not necessarily blanket imitators; rather, they perform redundant tool actions selectively by judging the relative expertise of the model, a strategy that is likely highly adaptive in the context of cultural learning (Richerson & Boyd, 2005).

Of interest to the current study was just how fine-grained this capacity for selective imitation can be. The results outlined above show that children are clearly capable of discriminating between individual models based on their perceived expertise (at least when the models are from extreme age groups). However, we know very little about children’s ability to make finer discriminations between the behavior of adult models who are all of a similar age but vary according to a particular attribute. One model attribute that may have a particularly powerful influence on copying behavior is the model’s status in relation to the observers. It is likely that adult models are high status relative to children generally; however, in everyday life, children interact with a variety of adults, some of whom have higher status (e.g., parents, teachers) than others. If children do not view all adults as equal, then we may see differences in the extent to which tool behaviors are copied after observing adults of differing status. However, we currently know very little about over-imitation in this context, making any link between status and copying behavior somewhat speculative at this stage.

There is good reason to suspect that these two areas are closely intertwined because within human societies high-status or “prestigious” models generally have a powerful influence on the behavior of other group members through displays of their superior skills and knowledge (Barkow, 1975, Henrich and Gil-White, 2001). This deference to status has also been found within groups of children, with young children being more likely to imitate dominant same-age peers over nondominant peers during free play (Abramovitch and Grusec, 1978, Grusec and Abramovitch, 1982) as well as in more structured picture preference tasks (Harvey & Rutherford, 1960). Similarly, if we take age as a proxy of prestige, children are more likely to copy older peer models (Brody & Stoneman, 1985) and to copy adult models over child models (Bandura & Kupers, 1964). Children have also been shown to be capable of discriminating between adults of differing levels of authority by imitating aggressive models more readily than nonaggressive models (Bandura, Ross, & Ross, 1961) and by copying the choices made by a dominant parent more readily than those made by a passive parent in a picture preference task (Hetherington, 1965). Taken together, these studies show that children are sensitive to status hierarchies from a very early age, a sensitivity that we aimed to capitalize on in the current study.

More specifically, the aim of the current study was to explore the link between trust in high-status individuals and over-imitation while keeping the age, sex, verbal/nonverbal communication and task demonstration performed by each model consistent. To achieve this, four groups of 5-year-old children were presented with a task demonstration by one of four adult models who varied according to the level of status held in relation to the children. We hoped to provide a naturalistic display of status by choosing adult models with whom the participants interacted in their school environment. Two of the models were selected to be high status: the participants’ class teacher and head teacher (principal). The remaining two models were selected to be of lower status than the teachers: one who was unfamiliar to the children and one who was familiar to the children but held no particular authority over them. The unfamiliar model was included to provide a baseline measure of the tendency to copy adults generally, whereas the familiar model was included to ensure that any differences witnessed between the high-status models and the unknown model were not simply due to the fact that the teachers were familiar to the participants. It was predicted that the children would display selective copying based on the status of the models, with more irrelevant actions being performed following task demonstration by the high-status models than by the low-status models and with the demonstration by the head teacher eliciting the most over-imitation.

Section snippets

Participants

A total of 41 participants (20 boys and 21 girls, mean age = 65 months, SD = 3, range = 60–73) took part in the study. A further 9 participants (6 in the low-status conditions and 3 in the high-status conditions) were excluded from the study because they either never touched the box (n = 4) or they performed only exploratory actions on the box (n = 5) during their trials. The participants were recruited from local schools in Central Scotland.

Apparatus

The apparatus used was a transparent box (20 cm3) that was

Results

To explore the influence of model status on the level of imitation witnessed, the data for both types of action (irrelevant and relevant) were initially analyzed using a between-participants analysis of variance (ANOVA) with the data from each of the four model conditions collapsed into two status dimensions (high status: head teacher/class teacher; low status: familiar/unfamiliar baseline). On completion of the initial analyses, performance across the four individual model conditions was

Discussion

The extent to which young children displayed irrelevancies in their reproductions appeared to be heavily influenced by the status of the model, with high-status models, particularly the head teacher (principal), being copied with higher levels of fidelity than the lower status models. Nevertheless, some evidence of copying of the irrelevant tool actions was evident across all model conditions. When taken together, this pattern of performance suggests two things. First, it suggests that young

References (32)

  • J.E. Grusec et al.

    Imitation of peers and adults in a natural setting: A functional analysis

    Child Development

    (1982)
  • P.L. Harris

    Trust

    Developmental Science

    (2007)
  • P.L. Harris et al.

    Young children’s selective trust in informants

    Philosophical Transactions of the Royal Society B

    (2011)
  • O.J. Harvey et al.

    Status in the informal group: Influence and influencibility at differing age levels

    Child Development

    (1960)
  • M.E. Hetherington

    A developmental study of the effects of sex of the dominant parent on sex-role preference, identification, and imitation in children

    Journal of Personality and Social Psychology

    (1965)
  • V. Horner et al.

    Causal knowledge and imitation/emulation switching in chimpanzees (Pan troglodytes) and children (Homo sapiens)

    Animal Cognition

    (2005)
  • Cited by (0)

    View full text