Identifying latent patterns in undergraduate Students’ programming profiles

摘要

Abstract This study aims to explore and reveal profiling patterns in the measurement of cognitive and noncognitivecharacteristics of undergraduate students’ programming performances. Spatial skills, workingmemory, perceived programming self-efficacy, mathematics scores, and academic grade point averagescores were taken indicative variables to be explored. Participants of the study are 100 undergraduatestudents registered to the Programming-I course at two different universities. The data were analyzedthrough multi-dimensional profile analysis. The result of the multidimensional scaling analysis indicated twodifferent profiles for the two groups: high and low programming performance groups. For both groups,relationship between the most similar variables was found to be verbal memory, mathematics achievementand perceived programming self-efficacy. The results indicated that there was a relatively similarrelationship between visual-spatial memory and spatial orientation skills in the low-performance group,while mental rotation skill was significantly different than the other variables. It was noted that two profiles forhigh- and low-performance groups were quite different in terms of mental rotation skill. It was also found that spatial orientation, visual-spatial memory and mental rotation performances were all different from eachother, and from the other three variables in the group with high programming performance. The mostdefinitive variables for low- and high-performance groups were self-efficacy, verbal memory andmathematics achievement. This study revealed that only verbal memory was the determinant variable inboth groups for working memory.