Different Performance Of Shape Bias Between Adults And Children:Evidence From Hypothesis Testing Task

Shape bias refers to the phenomenon that children generalize a novel word to the objects that are similar in shape rather than color, size, or texture in the process of word learning. Shape bias was first found in the novel name extension task. Recently, researchers found that children would also show a shape bias in classification and generalization task. Not only children, but also adults attended to shape bias. And it was found that with the growth of age, shape bias manifests robustly.Regarding the phenomenon of shape bias, a debate has arisen that whether it results from perceptual categories or conceptual knowledge. Attentional Learning Account claims that children could learn the correlations between objects and words according to their experience. So they tend to generalize the novel name to objects in similar shape. Others stress that concept knowledge plays an important role in the process of word learning and induction. People not also depend on perceptual features, but also concept knowledge. And shape is an available cue to decide the category of objects, that why children attend more to shape than to color and texture.Recently, growing research suggest that the phenomenon of shape bias is affected by many factors, such as solidity, domain, animation, set size of stimuli, as well as the social effect. Until, the effect of pattern on shape bias is still unclear. A pattern is an arrangement of lines or shapes, especially a design in which the same shape is repeated at regular intervals over a surface. The feature of pattern just like shape and color is perceived visually. There were various patterns on the surface of objects, including stripes, little repeated shapes and so on. Hence pattern is more complicated than shape and color, because it contains global and local properties. Some researches discussed the importance of texture in wording learning, which is similar to shape in a way. In those researches, some found that children still attended to shape, which others found that there was no difference between shape and texture, in which food were used as stimuli. It suggests that children would depend on both intrinsic and extrinsic properties in the process of generalization for food. Most researches have proved that participants show a shape bias when a test object shared the same shape with the target whereas another test object shared the same color or texture. However, in real life, the test objects have varying degrees of similarity to the target. In this situation, does shape bias still exist or diminish?In the present study, we explore the effect of improved similarity between target object and testing object on adults’and children’s induction. Moreover, the dimension of texture which is popular used in previous studies was replaced by pattern in present study to investigate whether the shape bias would diminish.The participators of Experiment1and Experiment2were adults. The materials and procedure of Experiment1and Experiment2was the same, except the limited chance to test in Experiment2and the manner of recording participants answers. In Experiment1, participants had many times to the test the testing objects and if they found the key feature, they could report orally, while in Experiment2, the testing times were limited and computer could record adults’answers automatically. Experimenter presented the targets which contain three dimensions (shape, color and pattern) to the participants firstly. The targets were defined as an edible candy, and its edibility was associated with one of three features (the key feature). The task of participants was to discover the key feature. In order to complete the task successfully, participated needed to test the six testing objects. The testing objects could be classified into two kinds——high similarity (shared two features with the target) and low similarity (shared one feature with the target). There were three high similar items which are in the same shape, color and stripe separately with the target and the left two features were different with the target, while three low similar items are in the different shape, color and stripe separately with the target and the left two features were the same with the target. The results of two experiments show that adults preferred high similarity to low similarity and they attended more to shape than to color and pattern with the limited testing times, which suggest that shape bias is affected by time pressure.The participators of Experiment land Experiment2were6years old. The materials used in children’s experiments were the same with those of adults’ experiments. In children’s experiments, four pictures of protagonists (one girl and three boys) were presented to children to attract their attention. The target was described as a biscuit or an edible candy that the girl found, and the task of children was to found biscuits or edible candies from testing objects for three boys according to the key feature of the target. The procedure of Experiment3and Experiment4was the same, except the definition of targets. In Experiment3, the stimuli were defined as biscuits or wood brick to avoid children doubt that why candies were classified into two kinds (edible or inedible). The results of two experiments reveal that children prefer high similar objects to low similar objects, which was inconsistent with the results of adults’experiments. However, it was found that children attended more to pattern than to shape and color, which was not found in adults’experiments. We doubt the divergent was resulted from the difference of description of stimuli. In order to test the possibility, the objects were described as candies just like adults’experiments. And the results show that children still showed pattern preference. It suggests that the pattern of children was not caused by instruction and the mechanism of shape bias between adults and children was different. In the process of hypothesis testing, shape is an available cue to generalize the properties of food, while for six-year-olds, pattern is more important than shape.The results of the present study suggest that the similarity between target items and testing items plays an important role in induction for adults and children. Furthermore, it was found that the performances of shape bias between adults and children were different. In hypothesis testing task, adults showed a shape bias, but children attended to pattern, which was not found in previous study. The results of the present study were novel and enriched the relevant fields.