Article
Analysis of the effects of cognitive and non-cognitive predictors on college performance: an innovative application of decision tree and association rules
The Journal of Computing Sciences in Colleges
(2006)
Abstract
An innovative application of Decision Tree and Association Rules has been presented. When conventional statistical techniques (e.g. t-test, simple regression, multiple regression, discriminant analysis etc.) concluded that high school grade point average and standard test scores are the indicators of a typical college student's academic performance, then the above two data mining techniques ruled out these predictors showing that student's age has significant effects on non-cognitive variables, which influence his/her college performance and the number of credit hours enrolled predicts withdrawal trends. Empirical evaluation also shows the data mining techniques outperform and deserve better suitability since they neither require prior knowledge about outcome nor pre-assumptions about predictor variables' degree of redundancy.
Keywords
- College Performance
Publication Date
2006
Citation Information
Mohammed Ali. "Analysis of the effects of cognitive and non-cognitive predictors on college performance: an innovative application of decision tree and association rules" The Journal of Computing Sciences in Colleges (2006) Available at: http://works.bepress.com/mohammedali/45/