Our previous research asked children to play a game in which they had to say which of three versions of an animal was correct. Sometimes the two incorrect versions were produced by exchanging a part from another animal. Sometimes the incorrect versions were produced by altering the relative sizes of the correct parts. Though these two games were equally difficult for adults, children found the part changes much easier. They were at the level of adults by age 11 for part changes. However, it was not until aged 15 that they were at the same level as adults for distinguishing between size changes.
In a parallel section of our previous work, the same age progression was found in a task that asked children to learn to recognize novel objects made from geometric shapes and exposed to them from different aspects. Again adolescent children had surprising difficulty in dealing with the relationships between parts.
Our research will arrive at a better understanding of the adolescent's difficulties in these cognitive tasks. It is our belief that we can remember objects stored in our minds in two different ways. One of the ways is as a whole object that could be basically a single part. The second way explicitly states the relationship between the parts (e.g., "beside" or "on top of"). This second way demands more attentional capacity and therefore the child may prefer to remember objects in mental images from the first holistic method.
As well as showing if the difficulties extend to artefacts, the research will see whether it extends to the learning of novel objects. We will also see if instructions to use one or other type of image will change the developmental progression that is observed without instruction.
Developmental differences in object recognition would have implications for the use and design of educational software because in most computer applications so much depends on the speedy apprehension of visual displays. It would also have implications for educational practice. Children approach creative design and architecture as either partist or wholist (Kimbell et al., 1996). However, these findings have not been related to how the child approaches the recognition of real-world objects. Our findings would necessarily inform issues of individual differences and could have implications for the better assessment and prediction of those children best suited for careers in those fields.
The late development of cognitive processes has attracted considerable current research. The majority of research, including that in object recognition, shows late attainment of the adult potential. Thus, the unsurprising clear differences between preschool children and adults in perceptual classification do not quickly disappear in school-age children. We have shown previously in independent lines of research surprising limitations within the further trajectory towards adult performance (Davidoff & Roberson, 2002; Jüttner & Rentschler, 2002; Jüttner, Müller & Rentschler, 2006). Davidoff and Roberson asked about the correct appearance of animals that children as young as 5 could name successfully. It was not until 11 years that children were at adult levels for the correct recognition of a part change and not until 15-16 years for the recognition of the correct size ratio of parts. Jüttner, Rentschler and co-workers found the same inability of younger children to deal with spatial relationships between parts in the rather different task of crossmodal learning with novel objects. The current project examines alternative explanations for these surprising patterns of cognitive abilities in object recognition. In particular, we wish to compare the developmental properties of representations that code relations between parts (structural descriptions) compared to those that code view-centred (holistic or one-part) representations. Such representations suggest two alternative hypotheses for understanding the rather tardy demonstration of adult-like performance in object recognition. The first hypothesis (Hypothesis A) is that younger children simply do not have the necessary object representations (structural descriptions) to perform configural-based recognition. The second hypothesis (Hypothesis B) is that even younger children have object representations that code the relationship between parts but prefer to use other simpler (holistic) representations. The specific aims of the project are:
Browser does not support script.