Information Technology Reference
In-Depth Information
of the script (Dillenbourg & Tchounikine, 2007). Flexibility can be increased by
reducing extrinsic constraints which are implemented for technical reasons (e.g.
data consistency problems if one student drops out) or sometimes simply because
the designer did not consider leaving this choice open. Designing environments
with a clear scenario but which still allow for flexibility remains a challenging
trade-off for which new software architectures would be interesting.
9. The integration of a new learning technology in the classroom legacy (factor 13)
is important. It takes ages to convince a student who is used to chat on-line with
system X to move to system Y. There is a need to design our learning environ-
ments from the very beginning either as very open systems that can interact with
any other web service or simply to consider our learning environment as web ser-
vices. The advent of “cloud computing” may push learning technologies towards
a better technical integration
Implications for Design-Based Research
Design-based research (DBR) (Collins, 1992; Sandoval & Bell, 2004) relies on
prototyping-testing cycles with the participation of all classroom actors. If this
participatory design is carefully conducted, it should produce methods that “work
well.” However, the generalizability of the method and hence of the associated
learning effects is arguable. Despite our conviction that DBR is the best method
for studying classroom orchestration, we must acknowledge that the generaliza-
tion remains a problem. Conjecture maps help the interpretation of results and their
potential generalization. Conjectures embed the relationship between interventions
and expected learning outcomes or expected process changes and are not very dif-
ferent from experimental hypotheses. One solution we have chosen for the logistics
project was to complement the DBR approach by a more formal experiment with
a control group. When a method “works well,” it means that it is compliant with
the many constraints of the context where it has been tested. It's never “it works
well universally” but “it worked well in my class this year.” Understanding which
constraints were satisfied and why the method satisfied them is the condition to gen-
eralize the DBR results. This generalizability is not only for the sake of science but
also because it is our social duty to come up with methods that “work well” beyond
a single context (even if no method will “work well” universally). We elaborate this
in the next section.
Orchestration as Constraints Management
Teachers who have to orchestrate the classroom and its technologies actually face a
multi-constraints management problem.
