Information Technology Reference
In-Depth Information
being used by the peer-student in the video. The student is also provided with a tran-
script of the video to refer back to when deciding which strategy or strategies were
being used. The purpose of this phase is to provide the student with practice self-
explaining and with a model of how the text might have been explained by another
student. This provides the student with more examples of the strategies being used,
and induces the student to reflect on what types of strategies can be used, and when.
Evaluation of One-on-One SERT
McNamara (2004b) examined the effects of SERT on comprehension and explana-
tion quality in a study with 42 college students. Half of the participants received
SERT and the remaining participants read aloud the four science texts (control con-
dition). During the training phase, self-explanation, as compared to reading aloud,
only improved comprehension for the most difficult of the four training texts. After
training, all of the participants (SERT and control) were asked to self-explain a dif-
ficult text about cell mitosis. Those students who were prompted to self-explain
(as in Chi et al., 1994) were compared to those who were provided with train-
ing to self-explain using the reading strategies (i.e., SERT). Those who received
the additional training on reading strategies (i.e., SERT) showed significantly better
comprehension than those who were merely prompted to self-explain.
Analyses of the self-explanations produced by the participants after training
indicated that SERT's primary role was in helping the low-knowledge readers to
use logic, common sense, or general knowledge to self-explain the text. Thus, the
results showed that SERT helped the low-knowledge students to more effectively
self-explain the text (using more effective strategies) and as a consequence they
showed considerably better comprehension than the low-knowledge participants in
the control condition who had not received training. Importantly, the low-knowledge
participants who received SERT showed comprehension performance comparable to
the high-knowledge participants.
Notably, these benefits only emerged on the text-based questions. The low-
knowledge readers did not have sufficient domain knowledge to generate inferences
to support a coherent situation model. Nonetheless, the use of paraphrasing along
with the generation of inferences based on logic and general knowledge helped the
readers to understand the basic ideas in the text and form a more coherent textbase
level understanding. This is important educationally because a coherent textbase is
a prerequisite to building knowledge.
McNamara (2000a) also reported on a study with middle school students (grades
6-8) who participated in a laboratory study with one-on-one SERT. These findings
were never published in a refereed journal because the training did not yield a pos-
itive effect of SERT. There was no hint of a benefit of self-explanation and reading
strategy training for these younger children. It seemed that they did not have suf-
ficient basic skills in order to gain from the training. In particular, it seemed that
they did not have sufficient paraphrasing skills and they did not have the ability to
explain the text using knowledge. Rather than explaining the text and making con-
nections between ideas in the text, their explanations tended to take them on tangents
