Harms, K. J., Rowlett, N., & Kelleher, C. (2015). Enabling independent learning of programming concepts through programming completion puzzles. In Visual Languages and Human-Centric Computing (VL/HCC), 2015 IEEE Symposium on Visual Language and Human-Centered Computing (VL/HCC) (pp. 271-279). New York: IEEE.


Type: Empirical

Purpose: The purpose of this article was present programming completion puzzles as independent learning alternatives to tutorials.

Findings: The puzzle group participants to 23% less time between conditions, and required 50% more mental effort than tutorial group participants. MANOVA analysis reveal significant differences on training task time for sequential execution, parallel execution and repeated & parallel tasks. The puzzle group had to constantly check the output to validate the work. The tutorial group did not have to do this because this group was led through the correct procedures of the program. The puzzle group rated the challenging tasks as requiring more mental effort than the tutorial group. The self-developed task survey and the TEQ found no statistically significant differences between puzzle and tutorial groups. Puzzle group and tutorial group are similar in terms of motivating for students. The puzzle group performed 26% better on transfer tasks. The puzzle group performed statistically better than tutorial group on two transfer tasks: repeated and parallel repeated tasks.



Sample Size: 27

Participant Type: Twelve females and fifteen boys were recruited from a mailing list at the Academy of Science of St. Louis