Inproceedings,
Solving Code-tracing Problems and its Effect on Code-writing Skills Pertaining to Program Semantics
Proceedings of the 2015 ACM Conference on Innovation and Technology in Computer Science Education - ITiCSE \textquotesingle15, page 314-319. ACM Press, (2015)
DOI: 10.1145/2729094.2742587
Abstract
An earlier study had found that solving code tracing problems helped improve code writing skills of students. But, given the instruments used in the earlier study, the improvement in code writing pertained primarily to language syntax. A follow-up within-subjects controlled study was conducted to investigate whether solving code tracing problems could help improve code writing skills pertaining to the semantics of a program. In the study, students were asked to write code for a control and a test problem both before and after a problem-solving session on code tracing. Increase in the score from pre-quiz to post-quiz was treated as improvement in code writing attributable to code tracing. Repeated measures ANOVA was used to analyze the data collected over four semesters. A statistically significant improvement in code writing skills pertaining to program semantics was observed on the test problems, but not on control concepts in the control problems. The improvement in code-writing skills as they pertain to program semantics accrued to the students who scored 90% or more on code-tracing problems in this study. Finally, the transfer in learning from code-tracing activities to code-writing skills may be near as well as far.
