Class Evaluation

This outline is based on several sources. Click [References] to view sources.

I. Approaches

	Teachers self evaluations correlate positively with students, although teachers tend to be harder on themselves than students are.
	Correlation of students' ratings with faculty peer observers. Increases if peers observe teacher on at least three occasions.
	Multisection approach. meta-analysis results (six meta-analysis to date) r= .40 - .50 range overall

II. Uses and Misuses

Reliability

- Are students' ratings stable over time? Are students better able to judge a teacher's effectiveness after a course is over? r in the .80 .85 range.

- Instructor effects versus course effects

Validity

- Valid for What?

• What criteria should be used for tapping teaching effectiveness?

- Remember, we do not directly measure effectiveness.

• Observed Scores versus latent variable.

Bias factors

- Are student ratings biased, do they compromise validity?

- Common Bias Criticisms (Some examples).

• More productive researchers put less time into teaching, get lower ratings. Wrong.

• Difficult courses receive low ratings. Wrong. If anything, opposite.

• Class size; large courses (esp. 60-90 range), get somewhat lower ratings. Effect is small. Reverse J curve.

• Easy graders get higher evaluations. Effect is small. Overall, little effect.

• Teachers high in showpersonship receive higher ratings. Only to the extent that their students learn more.

Flashy teachers who otherwise do a poor job will get panned by students on the average.

What are evaluations best for?

- Formative evaluation (larger measures)

(mid semester evaluations can be very helpful)

- Summative evaluation

Misuses to Avoid

- Remember: effect size is far from perfect

- Looking at variability: The confidence interval

- Course enrollment effects

III. Proper Uses

	Use student rating information in conjunction with other sources of information. Students are good evaluators of some aspects of teachers' behavior, but not others.
	Interpret student ratings only if a strong majority of students enrolled in a course complete the evaluation form.
	Interpret student ratings in light of what is known from research studies and from the local context.
	Make summative decisions only after a teacher has taught a large number of classes, that is, after patterns in student ratings emerge.
	Open-ended questions should be completed by students along with student rating forms.
	The University of New Hampshire rating form does not address all aspects of teaching that should be evaluated in a particular course. Teachers should develop their own questions when appropriate. There is space on the current rating form for students to rate special items.
	Teachers and administrators should compare individual teachers' ratings with department, college, and university norms.