Caveats on Using Classroom Observations for Summative Purposes
Researchers agree that classroom observation is more successful as a strategy
for improvement than for making personnel decisions. If it is used for personnel
decisions, the system must be carefully constructed and carried out.
Some considerations:
- For observations to provide valid, reliable data, at least two different
observers should make two observations each. An effort should be made to
obtain a balanced sample of the instructor's performance. If the course includes
lecture, discussion, and a laboratory component, for example, the observer
should see a sample of each of these activities. Single observations do not
provide an adequate sample and research has shown that ratings from a single
observation have NO correlation to student learning or student evaluation
results.
- Evaluative criteria need to be decided on in advance and shared with the
TA who is being observed. What will you be rating in this observation? The
TA's command of the material? Ability to respond to student questions? Enthusiasm
for the subject matter? Whether or not students are encouraged to participate?
- The observer and instructor should meet before the
observations to discuss the overall goals for the course, the objectives
for this class session, the students, and the teaching strategies chosen
along with a rationale for the choice. This information is necessary to
make judgments about the appropriateness of content selected and teaching
strategies.
