Here is some of what we know about student rating forms from the research:

• Ratings of overall teaching effectiveness are moderately correlated with independent measures of student learning and achievement. Students of highly rated teachers achieve higher final exam scores, can better apply course material, and are more inclined to pursue the subject subsequently. (Sources: Abrami, Apollonia, and Cohen, 1990; Braskamp, Brandenburg, and Ory, 1984; Cohen, 1981, Kulik and McKeachie, 1975; McMillan, Wergin, Forsyth, and Brown, 1986; Marsh and Dunkin, 1992)
• An instructor's ratings for a given course tend to be relatively consistent over successive years; there is not much variation in student ratings for an individual instructor regardless of whether the form is administered to current students or to alumni. (Sources: Braskamp, Bradenburg, and Ory, 1984; Centra, 1979; McMillan, Wergin, Forsyth, and Brown, 1986; Marsh and Dunkin, 1992)
• There is little or no relationship between the following characteristics of students and their ratings of instruction: age, grade point average, year in college, and academic ability. No consistent relationships have been found between student ratings and such variables as the amount of homework assigned or grading standards. (Sources: Braskamp, Brandenburg, and Ory, 1984; Centra, 1979; Kulik and McKeachie, 1975; McKeachie, 1979; McMillan, Wergin, Forsyth, and Brown, 1986; Marsh and Dunkin, 1992)

Researchers do report the following relationships:

• Students tend to rate courses in their major fields and elective courses higher than required courses outside their majors. (Sources: Kulik and McKeachie, 1975; McKeachie, 1979; Marsh and Dunkin, 1992)
• Faculty tend to receive more positive ratings than graduate student instructors. (Source: Marsh and Dunkin, 1992)
• The gender of a student has little effect on ratings. The gender of an instructor, however, may have an impact. Though some studies report no relationship between a professor's gender and student ratings, others show that adhering to a gender-appropriate teaching style may be rewarded by higher evaluations. (Sources: Basow and Silberg, 1987; Bennett, 1982; Kiersted, D'Agostin, and Dill, 1988; Marsh and Dunkin, 1992; Statham, Richardson, and Cook, 1991)
• Ratings can be influenced by class size (very small classes tend to receive higher ratings), by discipline (humanities instructors tend to receive higher ratings than instructors in the physical sciences), and by type (discussion courses tend to receive higher ratings than lecture courses). (Sources: Cashin, 1992; Feldman, 1984; Marsh and Dunkin, 1992)
• Data from classes in which fewer than 75% of the students respond or in which there are fewer than 15 students may not provide reliable information.
• It is not justifiable to infer difference in teaching effectiveness from small differences in average ratings. To obtain statistical significance, differences of .5 to 1.5 may be needed, especially at the higher end of the rating scale.

Recommendations

• When making judgments about an individual instructor, do not use university-wide norms for comparative data. Compare the ratings to similar courses (e.g., other required introductory lab courses) taught by other TAs.
• Do not use student ratings as the only measure of teaching effectiveness. They do not provide evidence in all areas relevant to teaching effectiveness (e.g., command of subject matter, appropriateness of course content and objectives).
• Do not rank TAs based on student ratings. Not only is it unjustifiable to infer differences in teaching effectiveness from small differences in student ratings, but it can have a detrimental effect on morale and inhibit TAs from actively helping each other improve teaching. Instead use student ratings to identify both potential problems and teaching excellence. TAs who receive ratings in the bottom 30% of all similar courses taught by TAs may be referred to appropriate assistance for improving their teaching practice. TAs who receive ratings in the top 30% may be commended.