A task-based program comprehension framework for novices / Amal Abdulghany Qassem Al-Shargabi

• Main Author: Al-Shargabi, Amal Abdulghany Qassem
• Format: Thesis
• Language: English
• Published: 2016
• Subjects: Electronic Computers. Computer Science; Programming. Rule-based programming. Backtrack programming; Computer simulation
• Online Access: https://ir.uitm.edu.my/id/eprint/38824/1/38824.pdf; https://ir.uitm.edu.my/id/eprint/38824/

Novices face much difficulty in comprehending even small computer programs. A framework is needed to help novices' attain sufficient program comprehension skills. To develop such framework, this research re-examines the three main factors that influence the novices' program comprehension; the programmer, the program code, and the task. Most studies so far focused on the first two especially programmer, with tasks traditionally applied only as a measure for program comprehension. Very few studies have inverted this supplementary role of task and instead examined the task factor itself. There is a research gap on the effects of tasks on novices' program comprehension. Moreover, current program comprehension mental models (PCMMs) have been mostly constructed for expert programmers, which do not match the novices' mental model. To solve these two problems, the research proposes a task based program comprehension framework for novices. Employing empirical software engineering research design approach, 1) fourteen effective programming tasks for novices have been identified, 2) a novices' PCMM has been constructed, and 3) the effects of the tasks on the PCMM have been investigated. Consequently, the tasks were classified according to the cognitive domain of Revised Bloom's Taxonomy. The tasks in each cognitive category were then ranked based on their effectiveness on novices' program comprehension using a survey of instructors. Subsequently, novices' PCMM with four core abstraction levels, i.e. Statement, Block, Module, and Program, and an extended level, i.e. Domain, was developed, and validated by experts. The effects of eight tasks on the novices' PCMM were investigated through an experiment involving 69 novices, and six tasks were validated through another experiment involving 178 novices in three universities. Both experiments also validated the ranking of these tasks on the novices' PCMM. The first experiment demonstrated that all the tested tasks were able to significantly improve novices' PCMM. The first and second ranked tasks were studied more closely, where the first ranked tasks consistently showed higher improvements than the second ranked tasks in each of the four tested Bloom's categories of Remember, Understand, Analyse, and Create. The second experiment demonstrated that different tasks improved the abstraction levels and the information categories differently. It also indicated that higher cognitive category tasks improve PCMM at higher abstraction levels. The general implication is that the framework can be an effective tool for computing educators to incorporate program comprehension in programming, and thus shift from merely teaching tracing and debugging tasks only. These tasks need to be introduced in stages in the teaching of programming, starting initially from the lower cognitive categories' tasks such as Recall and culminating at the higher cognitive categories' tasks such as Modification in possibly team project assessments. However, these tasks should be applied with taking the consideration of the novices' programming levels and the information categories need to be improved. The key contribution of this thesis is a new developed framework, which includes novices' PCMM, and a set of classified and ranked effective tasks that can improve novices' PCMM.