Survey of programming previous knowledge in university level

Abstract

In recent years, a preponderant role has been given to the teaching of programming in the compulsory levels of education. Numerous alternatives have also emerged to learn to program in a self-taught way. As a result of these processes, the public entering the university may have heterogeneous programming prior knowledge. Computer science university degrees usually assume zero programming knowledge, but that is increasingly far from reality. In this work we intend to survey the population of university students who are studying the subject Introduction to Programming at the National University of Quilmes, a subject that was originally designed to teach basic concepts of programming to students without any previous knowledge in the area. We developed a diagnostic test to investigate previous experiences in programming, as well as to evaluate some concepts regarding syntactic constructions of programming languages. Our findings show the enormous heterogeneity of the public, having students who learned programming concepts at different stages (at school, with online courses, self-taught), with different programming languages ​​(both in text and in blocks) and also, with very different perceptions of how much they really know.