| This study examines the nature of the process of comprehending quantitative data displayed graphically. The Pinker-Kosslyn model of graph comprehension shows that both perceptual and categorization processes are important to graph comprehension. These processes are examined in terms of mathematical experience and precollege mathematics achievement, measured by Math ACT. Three different levels of college mathematics students were examined (Intermediate Algebra, Calculus II and an upper division mathematical Applied Statistics course). The research questions were (1) whether there were differences between levels of mathematical experience in graph reading skills, (2) whether there were differences in views of graph typicality for different levels of mathematical experience, and (3) whether different graph types were read differently by subjects of different levels of mathematical experience. The first result showed that there are differences between groups in pattern description skills while there were no differences in value reading skills. It also showed that there was no significant correlation between Math ACT and either graph reading skill. The second result, analyzed the typicality data using multidimensional scaling and cluster analysis showed that there were differences in views of typicality for the groups of mathematical experience examined, with the group with the most experience having the most complex configurations and clusters and the group with the least experience having much more simple configurations and clusters. These clusters showed an increasing attention to orientation of the graph as a factor in typicality rating with increasing mathematical experience. The third result showed that there were differences in graph reading ability for different types of graphs. In fact, the value reading skill results show that the graphs that were associated together in the clusters from the typicality data are read equally well. However, these graphs were not grouped in terms of perceptual complexity. Thus categorization or schema selection is seen to play an important role in graph comprehension. However, perceptual complexity is shown not to have as important a role in graph comprehension as the preliminary study suggested (McKnight and Fisher, 1991a, 1991b). |
