How prior experience and task presentation modulate innovation in 6-year-old-children

Highlights

• Six-year-old children displayed robust low innovation rates in the floating peanut task.

• Enhanced salience of the tool and an action demonstration increased performance to some extent.

• Performance was not influenced by prior experience with the tool (i.e., no functional fixedness).

Abstract

Low innovation rates have been found with children until 6–8 years of age in tasks that required them to make a tool. Little is known about how prior experience and task presentation influence innovation rates. In the current study, we investigated these aspects in the floating peanut task (FPT), which required children to pour water into a vertical tube to retrieve a peanut. In three experiments, we varied the amount of plants that 6-year-olds (N = 256) watered prior to the task (zero, one, or five plants), who watered the plants (child or experimenter), and the distance and salience of the water source. We expected that prior experience with the water would modulate task performance by either boosting innovation rates (facilitation effect) or reducing them given that children would possibly learn that the water was for watering plants (functional fixedness effect). Our results indicate robustly low innovation rates in 6-year-olds. However, children’s performance improved to some extent with increased salience of the water source as well as with an experimenter-given hint. Due to the low innovation rates in this age group, we investigated whether watering plants prior to the FPT would influence innovation rates in 7- and 8-year-olds (N = 33), for which we did not find evidence. We conclude that 6-year-olds struggle with innovation but that they are more likely to innovate if crucial aspects of the task are made more salient. Thus, although 6-year-olds can innovate, they require more physical and social scaffolding than older children and adults.

Introduction

Problem solving is defined as a process in which individuals evaluate and select appropriate actions to overcome obstacles to fulfill a desired goal (e.g., Deloache, Miller, & Pierroutsakos, 1998). Prior experience with parts of the problem may have positive or negative effects. For example, experts restructure problems faster than novices (e.g., in chess: Reingold et al., 2001, Sheridan and Reingold, 2014). Nonetheless, people often experience difficulties in coming up with novel solutions for familiar problems, and they struggle when using familiar objects in unfamiliar functional ways (mental set and functional fixedness effect; e.g., Bilalić et al., 2008b, Duncker, 1945). A subset of problems may be described as innovation tasks that require a creative solution, for example, to manufacture a tool that you have never made or even used before (e.g., Beck, Apperly, Chappell, Guthrie, & Cutting, 2011). Such problems cannot be solved by analytical reasoning alone but rather require a creative process because the precise path from the starting point to the target state is unspecified (sometimes referred to as “ill-structured problems”; e.g., Cutting et al., 2014, Jonassen, 1997).

Recently, Beck and colleagues published a series of studies that required children to make a hook out of a straight pipe cleaner wire to retrieve a bucket with a sticker located at the bottom of a vertical tube (e.g., Beck et al., 2011, Chappell et al., 2013, Cutting et al., 2014; see also Weir, Chappell, & Kacelnik, 2002). Children performed rather poorly in the hook task but showed consistent improvement with age, with 6-year-olds showing an innovation rate of about 40% and 8-year-olds reaching about 60% (Beck et al., 2011; see also Chappell et al., 2013). Children aged 4–7 years preferentially selected the bent pipe cleaner when choosing between a straight pipe cleaner and a bent one, indicating at least an understanding of the tool affordances (Beck et al., 2011). Even telling 4- to 7-year-olds to produce something out of the given materials or encouraging them to try something did not improve their performance (Chappell et al., 2013, Cutting et al., 2011). This suggests that children’s failure was not caused by fear to bend the pipe cleaner or by perseverance with one solution strategy. Interestingly, another study showed that success rates were comparable between 3- to 5-year-old South African Bushmen and Western children, indicating that this finding is robust across cultures (Nielsen, Tomaselli, Mushin, & Whiten, 2014; see also Neldner, Mushin, & Nielsen, 2017).

Similar age-dependent innovation rates have been found with the floating peanut task (FPT; Hanus et al., 2011, Mendes et al., 2007). This task requires participants to pour water into a vertical tube to obtain a peanut resting on the bottom of the tube and was invented to study nonhuman great apes’ problem-solving (Hanus et al., 2011, Mendes et al., 2007, Tennie et al., 2010). Whereas children used a pitcher, bottle, or cup to transport and pour the water into the tube in previous studies (Hanus et al., 2011, Nielsen, 2013), great apes transported the water in their mouths and spat it into the tube (Hanus et al., 2011, Mendes et al., 2007, Tennie et al., 2010). Hanus et al. (2011) presented 4-, 6-, and 8-year-old children with the FPT with either a dry or wet (i.e., quarter-filled) tube. The probability of solving the task steadily increased as a function of age and the tube condition (dry/wet), reaching an innovation rate of about 50% in 6-year-olds and 75% in 8-year-olds with the wet tube. When Nielsen (2013) tested the FPT with 4-year-olds, he found comparable results, with nearly none of the children solving the task spontaneously. The hook task and the FPT both require using a familiar object or a liquid in a novel way, and consequently they represent cases of complex behavioral innovation. The hook task further requires tool manufacture (making a hook), something that is not required in the FPT. On the other hand, the solution in the FPT involved an additional action planning step given that it was solely the water source (drinker), not the water itself, that was visually available for the participants while facing the problem.

Prior experience with parts of the problem, such as a tool or a specific manipulation, can influence task performance in various kinds of problems (e.g., Birch and Rabinowitz, 1951, Flavell et al., 1958, Yonge, 1966), and although some prior experience may lead to a fixation effect, too much experience can cause a reversed pattern. For example, experts in a given field might flexibly choose from different solution strategies because of their great experience (e.g., Bilalić et al., 2008a, Flavell et al., 1958, Star and Seifert, 2006). Previous studies suggest that the functional fixedness effect (Duncker, 1945), which entails a fixation on the function of an object, seems to develop at around 6 years of age (e.g., Defeyter and German, 2003, German and Defeyter, 2000). German and Defeyter (2000) presented 5- to 7-year-old children with a task that required them to use a box that contained several objects in a novel functional way, namely as a support for stacking cuboids. Only 6- and 7-year-olds exhibited a functional fixedness effect, whereas 5-year-olds did not (German & Defeyter, 2000). Cutting et al. (2011) presented 4- to 7-year-olds with two tasks, counterbalanced for order across participants. Whereas the hook task required them to bend a pipe cleaner to produce a hook, the unbending task required them to unbend a U-shaped pipe cleaner into a straight wire to poke out a ball from a tube (Cutting et al., 2011). Only few children solved both tasks, and success in the first task did not predict success in the second task. This study suggests that prior experience with bending or unbending the tool neither facilitated nor hindered children’s ability to produce the respective contrasting solution. Chappell et al. (2013) gave 4- to 7-year-olds the opportunity to explore the materials prior to the hook task to ensure that they had experienced the objects’ features. Prior exposure did not have an effect on children’s performance, indicating no facilitation effect of prior experience with parts of the problem (Chappell et al., 2013). However, a combination of pieces of information seemed to help children in another study. When Cutting et al. (2014) let 5- and 6-year-olds explore the materials prior to the test and showed them a template hook, innovation rates increased substantially. Interestingly, this effect was not significant for 4- and 5-year-olds and not for the older children when they only explored the materials but did not see the target tool. Hanus et al. (2011) asked 4- to 8-year-olds to water plants prior to the FPT to familiarize them with the water. This prior experience may have potentially influenced children’s performance in the task, although its direction is unclear. On the one hand, drawing children’s attention to the water could have facilitated the solution. On the other hand, watering plants prior to the task might have blocked the idea of using the water in a different functional context, that is, showing a functional fixedness effect.

One aspect that has received little attention regarding functional fixedness is the role of self-experience versus other-experience. In other words, is it necessary for an individual to experience the function herself or himself, or is it enough to observe the function being used by another person? From the teleological–intentional perspective, one would expect that observing the function is enough to establish the idea of what the object is for, and indeed some findings suggest that this is the case in children as young as 2½ years (e.g., Casler and Kelemen, 2005, Defeyter et al., 2009, German and Johnson, 2002, Hernik and Csibra, 2009). Whereas previous studies explored whether children assign functions to objects after observing another individual using them, here we focused on whether observing the function would also induce a functional fixedness effect. The FPT seemed like a good task for studying this effect because it has the right level of difficulty that allowed for a two-sided hypothesis.

In general, prior experience with a tool can be gained individually or socially and may be linked to ostensive cues. For example, young children take an actor’s intention into account in a task context, which may facilitate problem solving thereafter (Carpenter et al., 2002, Carr et al., 2015, Huang, 2013). Interestingly, children mainly copy actions that have the desired outcome (Want & Harris, 2001) even though overimitation (i.e., copying nonefficient actions) is a quite robust phenomenon among younger children (Király, Csibra, & Gergely, 2013). Moreover, it seems that children’s innovative abilities are potentially tempered by a bias toward social learning (Csibra and Gergely, 2009, Király et al., 2013), which may explain, at least in part, the relatively low innovation rates found in problem-solving tasks (Beck et al., 2011, Cutting et al., 2011, Hanus et al., 2011, Nielsen, 2013). Thus, it is important for children to differentiate between relevant and irrelevant prior experience gained through their own actions or through observation (Williamson et al., 2008, Yu and Kushnir, 2014).

In the current study, therefore, we explored the effect of watering plants prior to being confronted with the FPT in 6-year-old children and whether it mattered how children experienced this, namely whether they watered the plants themselves or they watched an experimenter doing so. We chose 6-year-olds because they performed at an intermediate level in the FPT in a previous study, allowing us to entertain a two-sided hypothesis (Hanus et al., 2011). Moreover, the functional fixedness effect seems to develop at around 6 years of age (Defeyter and German, 2003, German and Defeyter, 2000). We implemented the FPT in a game in order to induce a positive mood and to decrease social pressure because positive affect seems to facilitate solutions in creative problems (e.g., Lin, Tsai, Lin, & Chen, 2014).

In Experiment 1 (N = 96), we investigated the effect of watering plants prior to the FPT (five, one, or zero plants) and the impact of whether 6-year-old children watered the plants themselves or whether they observed the experimenter doing so (self-experience vs. other-experience). We hypothesized that watering more plants would either have a positive (i.e., facilitating) effect or a negative (i.e., functionally fixating) effect on innovation rates that would be more profound when the experienced with the water had been gained by children themselves. In Experiment 2 (N = 64), we focused on the influence of the distance to the water (close or far) and the condition of the tube (dry or wet, i.e., quarter-filled with water). We hypothesized that innovation rates would increase with water being close and that this effect would be even more pronounced when the tube already contained water. In Experiment 3 (N = 96), we examined the same variables as in Experiment 1 but increased the salience of the water source (bucket close to the tube and transparent). We were again interested in whether watering plants prior to the FPT and the type of experience would have an impact on innovation rates. In Experiment 4 (N = 33), we focused on 7- and 8-year-old children to assess age changes in performance of some selected conditions of the previous experiments. Half of them got an extensive watering experience (five times), whereas the other half did not use the water at all prior to the task. We hypothesized again that watering plants could have either a positive or detrimental effect on innovation rates.