摘要:In computer science education at school, computational thinking has been an emerging
topic over the last decade. Even though, computational thinking is interpreted and integrated in
classrooms in different ways, an identification process about what computational thinking is about
has been in progress among computer science school-teachers and computer science education researchers
since Wing’s initial paper on the characteristics of computational thinking. On the other
hand, the constructionist learning theory by Papert, based on constructivism and Piaget, has a long
tradition in computer science education for describing the students’ learning process by hands-on
activities. Our contribution, in this paper, is to present a new mapping tool which can be used to
review classroom activities in terms of both computational thinking and constructionist learning.
For the tool, we have reused existing definitions of computer science concepts and computational
thinking concepts and combined these with our new constructionism matrix. The matrix’s most
notable feature is its scale of learners’ autonomy. This scale represents the degree of choices
learners have at each stage of development of their artefact. To develop the scale definitions, we
trialed the mapping tool, coding twenty-one popular international computing activities for pupils
aged 5 to 11 (K-5). From our trial, we have shown that we can use the mapping tool, with a
moderate to high degree of reliability across coders, to analyse classroom activities with regard to
computational thinking and constructionism, however, further validation is needed to establish its
usefulness. Despite a small number of activities (n = 21) being analysed with our mapping tool,
our preliminary results showed several interesting findings. Firstly, that learner autonomy was low
for defining the problem and developing their own design. Secondly that the activity type (such
as lesson plan rather than online activity) or artefact created (such as physical artefact rather than
onscreen activity or unplugged activity), rather than the computational thinking or computer science
concept being taught was related to learner autonomy. This provides some tentative evidence,
which may seem obvious, that the learning context rather than the learning content is related to
degree of constructionism of an activity and that computational thinking per se may not be related
to constructionism. However, further work is needed on a larger number of activities to verify and
validate this suggestion.