∂A/∂C + ∂A/∂T
C
*∂T
C
/∂C] – [∂A/∂S +
∂A/∂T
S
*∂T
S
/∂S].
This model can be extended to account for different assumptions about the allocation of
classroom time. First, computers may increase the total amount of instruction a student receives
if teachers must divide their time between group and individual instruction. In this scenario,
some traditional class time,
T
S
, is wasted for students and CAI can fill in these down periods.
This should cause increased investment in ICT, and CAI in particular, to be more likely to have a
positive effect on educational outcomes. Alternatively, students may use computers for non-
instructional activities that offset instructional time. Furthermore, mechanical problems with
technology could create instructional downtime. That is, some computer-based instructional
time,
T
C
, may be wasted and thus crowd out more productive instruction. This should cause ICT
investment to be more likely to have a negative effect. We discuss each of these adjustments to
the model and the implications for interpreting estimates in the literature.
Barrow, Markman, and Rouse (2009) propose a model to argue that CAI may increase
total instructional time during a class period or school day. They assume that a teacher
j divides
class time between providing group instruction,
T
j
G
, and individualized instruction for each
student
i, T
ij
. Each student receives group instruction and his or her share of individual
11
instruction. Computer instruction,
T
i
C
, provides supplemental instruction during periods when the
teacher is giving individual instruction to other students. This model differs from the baseline
model presented above in that CAI replaces down time rather than traditional instruction. The
revised constraints make these trade-offs clear.
(2.2)
T
jt
G
+ T
ijt
+ T
it
C
≤ T and T
jt
G
+ ∑T
ijt
≤ T
j
The return to computer-based instruction,
∂A/∂T
C
, is not offset by a reduction in traditional
instruction,
∂A/∂T
S
. Modeled in this way, CAI will improve academic outcomes if it provides
any academic benefit:
f(X
it
,A
it-1
,T
it
,T
t
G
,T
it
C
) ≥ f(X
it
,A
it-1
,T
it
,T
t
G
, 0).
10
Belo, Ferreira, and Telang (2014) model a case in which time spent using computers is
not necessarily productive. For example, students may use computers to watch videos or engage
in social networking activities that do not improve traditional academic outcomes. In this case,
computer time
Do'stlaringiz bilan baham: |