Key points are not available for this paper at this time.
A recent review of studies of stability in teacher effectiveness (Rosenshine, 1970) could locate only four long-term investigations which included stability coefficients reflecting teacher consistency across time in producing student gains. These four studies are difficult to compare with each other but, in any case, the stability coefficients were generally low. The median coefficients, when more than one test was used, were .34 (p < .001), .36 (p < .05), .20 (ns), and .09 (ns). How stable is the effectiveness of typical teachers? Are some teachers more stable than others? The present study addressed these questions by studying only ordinary teachers working with their regular classes (i.e., no experimental intervention), and by including a time period long enough to allow us to judge stability reasonably well (three full school years).
Jere Brophy (Tue,) studied this question.