An extended case study on the introductory teaching of programming

Abstract

Learning to program is a complex and arduous process undertaken by thousands of undergraduates in the UK each year. This study examined the progress of transforming the pedagogical paradigm of an introductory programming unit from a highly controlled, reductionist 'cipher' orientation to one in which students have more freedom to explore aspects of programming more creatively. To facilitate this, certain programming concepts were introduced much earlier that had previously been the case. This was supported by an analysis of the semiotics and symbology of programming languages that showed that there was no intrinsic support for the traditional sequence of introducing programming concepts. A second dimension to the transformation involved doubling the number of assessments to emphasise the benefits of continual engagement with programming. The pedagogical transformation was to have been phased over four successive cohorts, although the fourth phase had to be delayed due to a revalidation that amalgamated three programmes into a framework.

The study was planned during the second phase of the transformation. To ensure that the study did not disrupt the students’ learning experience the main focus of the research was on quantitative analyses of the work submitted by the students as part of the coursework for the unit. This work included programming portfolios and tests. In all, the work of more than 400 students completing more than a thousand portfolios and a thousand tests were analysed, providing a holistic view of waypoints in the learning process.

The analyses showed that the second and third cohorts responded positively to the greater level of freedom, creating more sophisticated applications utilising a wider range of programming constructs. In the latter part of the fourth cohort a more traditional, constrained approach was used by another tutor that resulted in a narrowing of the range of programming concepts developed.

The quantitative instruments were augmented by questionnaires used to gauge the students' previous experience, and initial views. Analyses of these returns showed that there appeared to be a limited relationship between a student's previous experience and the likelihood that he or she would succeed in the unit and be eligible to continue to the next stage of the undergraduate programme.

The original plan was for qualitative instruments to be introduced in the final two cohorts. The re-organisation alluded to earlier restricted qualitative methods to short, semi-structured interviews during the third cohort.
Within the study, certain aspects of the pedagogical transformation were considered in more depth: the development and use of a code generator and
criterion-referenced assessment. These innovations were part of another dimension of the transformation of the unit, emphasising comprehension and modification equally with construction. This dimension reflects the changing nature of programming, incorporating existing code wherever possible. The analyses showed that comprehension skills developed to a greater extent within the unit compared with modification and construction.

The main conclusions of the study were that the pedagogical changes had a beneficial effect on the learning of all students, including those with considerable previous experience, and those who had never written a program before.

More information

Identifiers

Catalogue record

Export record