Perceptions of Computer Science at a South African university

Abstract

First year students at the University of the Witwatersrand in Johannesburg, South Africa, were surveyed about their perceptions of Computer Science before and towards the end of their first year courses. The aim of this research was to investigate how the students’ attitudes changed during these courses and to assess the impact of the innovative breadth-first curriculum that has been developed in the School of Computer Science which emphasises the fundamentals of the discipline and the mathematical nature of Computer Science. The results show that most perceptions did not change much or that there were changes in both directions. More students, and particularly female students, were positive about their own understanding of the nature of Computer Science after the course than before. However, when asking specifically about jobs and course content, there was not a substantially deeper understanding at the end of the course of what content they would expect to encounter in a Computer Science course or working as a Computer Scientist. Fewer students, particularly male students, thought that Computer Science and mathematics were closely related after the course than before and this was an unexpected result, which may be the result of discrete mathematics topics being taught in courses separate from those in which continuous mathematics topics are taught. Students became less positive about working with computers after the course, a result which supports prior research, but is an issue for concern as computers will play some role in their future careers. The students found the courses challenging and different from their expectations with few students finding the courses unenjoyable.

Introduction

This paper describes research into changes in first year Computer Science students’ perceptions of and attitudes to Computer Science at a South African university, the University of the Witwatersrand (Wits), located in Johannesburg. Our experience and prior research (Herbert, 2000, Sanders and Galpin, 1994) have shown that students often only have a narrow understanding of the nature of Computer Science, even when they have chosen to study it, and that they often make a strong link between Computer Science and programming, to the exclusion of other aspects of the discipline. This is most likely due to lack of career counselling, and an introduction to computing at school level that focuses on applications, such as word processing, and programming. Another possible explanation for lack of understanding is the diversity of the discipline, since it contains aspects of science, engineering and mathematics (Denning et al., 1989) and because it is not well-defined (Nielsen, von Hellens, Greenhill, & Pringle, 1998).

Other research has also shown incorrect perceptions such as that computing careers only involve programming, there are limited career opportunities, there is little interaction with people, and that computing careers are linked to administration or secretarial work (Clarke and Teague, 1996, Craig, 1997, Durndell and Thomson, 1997, Greening, 1998, Ryan, 1994, Symonds, 2000, Teague and Clarke, 1991).

The research was also motivated by our innovative curriculum (Sanders & Mueller, 2000) which was developed to cope with the fact that some of our students have good school-leaving Mathematics marks, but no or little exposure to computers due to the legacies of apartheid. The curriculum positions Computer Science within the Mathematical Sciences and has a focus on fundamentals of the discipline, as opposed to specific programming languages or computing technology. The aim of the curriculum is to broaden students’ understanding of the discipline and to use the mathematical aspects of the discipline as an entry point to the discipline.

The research investigated students’ attitudes at the beginning and towards the end of our first year courses. The methodology used in this research involved comparing responses from questionnaires completed at the start and near the end of the academic year. For educational research, it is often difficult or impossible to sufficiently isolate various variables sufficiently to show causality. McMillan and Schumacher (2001) suggest that when extraneous variation cannot be controlled, then this needs to be taken into account in the interpretation of the results. In our case, the students are exposed to many influences during the year that may affect their perceptions, including our course. Since it is not possible for pragmatic and ethical reasons to present two different courses to them, or to remove factors other than the course, we have to consider the students within their environment, and be cautious in our interpretations.

The research shows that our students became more confident about their understanding of the nature of Computer Science, although when questioned about what content could be expected in Computer Science, did not show much change. The students changed their opinion about the relatedness of Computer Science and Mathematics, becoming less convinced of a link between the two. They also became more negative about working with computers, a concern since their future careers will most likely involve computers to a substantial extent. These results, as well as the fact that there was little change for many attitudes, raise questions about what influences students the most – what is said in lectures or what they spend time on in the laboratories, or external factors not related to the course. Additionally, the location of material may affect their perceptions – if discrete mathematics topics are only covered in Computer Science, then the link between this course, and mathematics courses based on continuous mathematics may be opaque to students.

In the rest of the document, the phrase “understanding of Computer Science” will be used to mean “understanding of the nature of Computer Science” as opposed to meaning success in learning the material in our courses. Similarly, “understanding the content of Computer Science” will be used to mean being able to identify different aspects of Computer Science and it will not mean having an understanding of the course material.

The structure of the document is as follows: the next section considers related research into attitudes, as well as giving details of the structure of the South African education system and our curriculum. Then we present the research methodology, the results and finally provide an interpretation and discussion of the results. The questionnaire used in this research can be found in a technical report (Galpin & Sanders, 2005) which presents a preliminary analysis of the data gathered by this questionnaire including some analysis not directly relevant to this paper.