Document Type

Theses, Masters

Rights

This item is available under a Creative Commons License for non-commercial use only

Publication Details

Thesis submitted to Dublin Institute of Technology in part fulfilment of the requirements for the award of Masters (M.A.) in Higher Education, November 2015.

Abstract

Computer programming is an essential skill that all computing students must master and is increasingly important in many diverse disciplines. It is also difficult to learn. One of the many challenges novice programmers face from the start are notoriously cryptic compiler error messages. These report details on errors made by students and are essential as the primary source of information used to rectify those errors. However these difficult to understand messages are often a barrier to progress and a source of discouragement. A high number of student errors, and in particular a high frequency of repeated errors – when a student makes the same error consecutively – have been shown to be indicators of students who are struggling with learning to program. This instrumental case study research investigates the student experience with, and the effects of, software that has been specifically written to help students overcome their challenges with compiler error messages. This software provides help by enhancing error messages, presenting them in a straightforward, informative manner. Two cohorts of first year computing students at an Irish higher education institution participated over two academic years; a control group in 2014-15 that did not experience enhanced error messages, and an intervention group in 2013-14 that did.

This thesis lays out a comprehensive view of the student experience starting with a quantitative analysis of the student errors themselves. It then views the students as groups, revealing interesting differences in error profiles. Following this, some individual student profiles and behaviours are investigated. Finally, the student experience is discovered through their own words and opinions by means of a survey that incorporated closed and open-ended questions. In addition to reductions in errors overall, errors per student, and the key metric of repeated error frequency, the intervention group is shown to behave more cohesively with fewer indications of struggling students. A positive learning experience using the software is reported by the students and the lecturer. These results are of interest to educators who have witnessed students struggle with learning to program, and who are looking to help remove the barrier presented by compiler error messages. This work is important for two reasons. First, the effects of error message enhancement have been debated in the literature – this work provides evidence that there can be positive effects. Second, these results should be generalisable at least in part, to other languages, students and institutions.

Share

COinS