StabilityIn some of the earliest studies of students' opinions of college teaching, Guthrie (1954) found correlations of .87 and .89 between students' rank ings of the quality of their teachers from one year to the next. He also found that such judgments were more stable than were faculty judgments of teaching quality. Lovell and Haner (1955) used a forced-choice rating scale to obtain students' opinions of their teachers. The correlation between ratings made two weeks apart was .89.More recently, Costin (1968a) found moderate to high correlations between mid-semester and end-of-semester ratings of teaching assistants in psychology, social sciences, humanities, physical sciences and biological sciences. The r's ranged from .70 to .87 for four of the dimensions mea-