The Role Of Linguistic Labels In 6-9-year-old Children's Specific Inductive Inference

Many studies show that linguistic labels have influence on children's specific inductive inference, but have different ideas about the essential role of linguistic labels because there is no consistent view about the mechanism of children's specific inductive inference. Some researchers regard the perceptual similarity as the base of children's specific inference, and the linguistic labels play a quantitative role as one of the attributes of the stimulus (see Label-as-Attribute Model). Other researchers hold a category-based induction, however, and they think inductive generalization is driven by the category membership rather than by featural overlap, and the linguistic labels are looked on as marks of categories by children, playing a qualitative role across the process of induction. This research adopted pictures of novel creatures with systematic chasge in their perceptual similarity, examined the influence of linguistic labels and perceptual similarity in 6-9-year-old children's specific inductive inference, and discussed the developmental characteristics and mechanism of their specific induction. More ever, this research examined the Label-as-Attribute Model.This research included two experiments. Experiment one examined the influence of ail-labeled condition and non-labeled condition on 6-9-year-old children's specific inductive inference while the perceptual similarity of the stimuli is systematically changed. Besides, it explored the developmental characteristics and mechanism of their specific induction. Based on experiment one, experiment two examined how linguistic labels works on 6-9-year-old children's specific induction when only some of the stimuli are labeled. Individual test was adopted in both experiments with oral reports of the subjects on why they choose like that, and the results were shown as below:1.6-9-year-old children's specific inductive inference is greatly influenced by both perceptual similarity and linguistic labels of the compares entities, and the proportion of label-based induction varied as a function of the overall similarity of compared entities no matter it is under all-labeled condition, partly-labeled condition, or non-labeled condition.2. Linguistic labels have a quantitative influence on 6-9-year-old children's specific inductive inference, especially to 6-7-year-old children. This influence would decrease obviously since 7 years, but would last till adult time. 7-8 years is an important period of time of linguistic labels' influence on children's specific inductive inference.3. The Label-as-Attribute Model can in some degree predict the rule children's specific inductive inference. Children after 7-8 years didn't make induction exclusively rely on linguistic labels, however, and this model can't explain the tendency of linguistic labels' role and it's change hi this process. While information modality theory can explain this phenomenon more rationally, holding that auditory stimuli might have greater attentional weight in younger children's information processing than in older, and this weight will decrease gradually with age till a relatively stable level.4. Specific ordinal relation among the stimulus pattern conditions in 6-9-year-old children's specific inductive inference is as follows: P ( T-00 ) = P (T-11) = P (T-22 ) > P (T-21) = P (T-10) > P (T-20). This result consist with prediction of the Label-as-Attribute Model, proving the number of overlapping attributes among compared entities played an important role in 6-9-year-old children's specific inductive inference.5. When one (stimulus A) of the stimuli is lack of linguistic label, children might still in some way regard it as stimulus A has a different label, but in this condition, the influence of linguistic labels is smaller than when stimulus A really has one different label. This result also supported views of the auditory modality theory.