Unplugged or plugged-in programming learning: A comparative experimental study

In recent years, computer programming has reappeared in school curricula with the aim of transmitting knowledge and skills beyond the simple ability to code. However, there are different ways of teaching this subject and very few experimental studies compare plugged-in and unplugged programming learning. The purpose of this study is to highlight the impact of plugged-in or unplugged learning on students' performance and subjective experience. To this end, we designed an experimental study with 217 primary school students divided into two groups and we measured their knowledge of computational concepts, ability to solve algorithmic problem, motivation toward the instruction, self-belief and attitude toward science. The programming sessions were designed to be similar between the two conditions, only the tools were different. Computers and Scratch software were used in the plugged-in group while the unplugged group used paper instructions, pictures, figurines and body movements instead. The results show better learning performance in the plugged-in group. Furthermore, although motivation dropped slightly in both groups, this drop was only significant in the unplugged condition. Gender also seems to be an important factor, as girls exhibit a lower post-test motivation and a lower willingness to pursue their practice in programming outside the school context. However, this effect on motivation was only observable in the plugged-in group which suggests that educational programming software may have a positive but gendered motivational impact.