|
Sample Size Issues
Research studies with too few students, classrooms, or schools
may lead to inappropriate conclusions about the effectiveness of
an intervention.
The size of the sample affects:
- the sensitivity of a study to find meaningful changes
- the reliability, or stability, of the results
If a researcher found technology use had no effect on student performance,
it might be because there were too few students in the sample to
provide sufficient sensitivity, or statistical
power, to detect meaningful changes in student performance even
when they occur. As the sample size becomes larger, the study's
power increases, and researchers will be more likely to find a meaningful
and statistically
significant difference in test scores if a difference occurs.
The use of small sample sizes may also lead to unreliable or unstable
findings. In order to be able to apply a study's findings to a larger
population, researchers must be certain that if the study were to
be repeated with the same number of, but different, students they
would get the same or similar results. When small sample sizes are
used -- fewer than 30 students -- there is a high probability that
the results of a study might change dramatically if a different
sample of students were selected. As a general rule, larger numbers
improve accuracy.
|
