Praise and criticism data were collected during a two-year correlational study of a selected sample of second- and third-grade teachers chosen for their consistency in producing student learning gains averaged over four years, and these data were analyzed to determine the effect that praise, criticism, rewards, and punishment had on learning gains. Data on motivations, incentive, and punishment differed considerably by socioeconomic status (SES). In low SES schools, praise was regularly but weakly associated with learning gains on several measures, but it was relatively unimportant in high SES classes. Criticism was negatively related in low SES classes, but positively related in high SES classes (although absolute incidence of both types of evaluative comments were low). Symbolic rewards (stars, smiling faces) were moderately effective motivators in both high and low SES classes, but verbal praise and the "reward" of monitoring duties were not. In low SES classes, neither these symbolic rewards nor punishment of any sort were strongly related to student gains. The most effective way of dealing with misbehavior in these classes was through an individual conference with the student. The main factor in the low SES classes was the teacher's ability to motivate the student to become actively engaged in the learning process to the point that he or she would answer questions in public response situations and work persistently on seatwork. Successful teachers in both kinds of schools communicated high expectations, but the successful teachers in high SES schools did so through a critical demandingness, while the successful teachers in low SES schools did so through patience and encouragement. (MJB)