Information Technology Reference
In-Depth Information
of knowing guides a strategic choice: if they think they have enough relevant knowl-
edge, learners will try to directly retrieve the answer from memory, but if they do
not think they have enough relevant knowledge, they will instead try to compute a
plausible answer from a set of related facts stored in memory (Reder, 1987).
After learners find an answer in their memory or derive a plausible one from rele-
vant knowledge, they control whether to withhold or report the answer. This decision
is greatly influenced by another form of monitoring: their
confidence
in the answer.
A strong correlation has been found between subjective confidence in the correct-
ness of an answer and the willingness to report that answer (Koriat & Goldsmith,
1996). When forced to provide a response for every general knowledge question
posed, learners report more answers but have lower overall accuracy compared to
learners who are allowed to respond with “I don't know.” This shows that under
“free report” circumstances, learners selectively withhold low confidence answers
in order to boost their overall accuracy. Furthermore, learners shift their confidence
criteria for reporting answers in response to external reward structures, suggesting
that learners have great control over which answers they report. Learners are willing
to report lower confidence answers when external incentives reward quantity over
accuracy but withhold these lower confidence answers when the external incentives
reward overall accuracy instead of quantity.
The Study of Metacognitive Control
In laboratory experiments, the relative effectiveness of metacognitive control
is evaluated by comparing memory performance following learner-based versus
experimenter-based control of some aspect of study. The implicit assumption in
such a comparison is that learners seek primarily to maximize performance. It is
worth noting, however, that students and other learners outside the laboratory may
have more complex goals. Such learners have constraints on the time they have to
spend (Son & Metcalfe, 2000) and the effort they are willing to expend and may be
seeking to satisfice rather than optimize (Simon, 1957).
Given this complex interplay of goals and abilities, as well as the high demand
for effective metacognitive monitoring, it is all the more impressive that there is
a wealth of results indicating that metacognitive control is widely used and often
quite effective. The next portion of this chapter will focus on examples of such
metacognitive control.
Effectiveness of Metacognitive Control
Self-Pacing of Study
One way of assessing the value of metacognitive control is to evaluate the efficacy
of self-paced study (or study-time allocation). As discussed earlier, learners usu-
ally devote more time to the items which they judge to be most difficult; however,
