and DANIBL B. QtU8B, Umverstty of ArkansasSeveral recent studies have investigated the effectiveness of social remforcement in modifying the performance of children m repetitive motor tasks The introduction of social reinforcers has been found to increase the frequency of the reinforced response and to be more effective following social depnvation than following satiation (Gewirtz & Baer, 1958a, 1958b) The level of response has been found to be increased to a greater degree by social reinforcement for institutionalized children than for normal children (Zigler, Hodgden, & Stevenson, 1958) and to be affected by the degree of premstitutional social deprivation experienced by feebleminded 5"s (Zigler, 1958)The above studies have been concemed only with the shortterm effects of social reinforcement and have not considered tbe effect of E's absence from the situation or of E's making critical, rather than supportive comments about 5"s performance The present studies seek to investigate the effects of such variables on the performance of normal and feebleminded children EXPERIMENT I The purpose of the first study was to investigate the effectiveness of social reinforcement in modifying performance over a five-day period Three experimental conditions were employed (a) a reward condition m which E made supportive comments conceming 5's performance, (fc) a condition in which E was attenbve to 5's performance but made no comments, and (c) a condition in which E instructed S and left the room during 5's performance The three conditions made it possible to determme the degree to which the supportive comments were effective in modifying behavior, and also

The distribution of scale values of 1647 items scaled for social desirability was presented. The biomodality of the distribution of scale values for the items was noted and interpreted as showing that social desirability judgments of personal concepts are infrequently judged as neutral and tend to be either undesirable or desirable. Presentation of these items as a personality test showed the typical high correlation between social desirability scale values and frequency of endorsement.

Research evaluating student ratingsofprofessorsrevealscontinuedcontroversy. Interpretations of student ratings of professors in terms of face validity are marred by halo affects, the apparent inability of even skilled raters to judge complex behaviors adequately, the salience of personality features in judging tasks, and a host of other variables. Research shows student ratings to be reliable, but design flaws for simple, first order, predictions usually omit the teacher as a cause. Interpretations of research are confusing because of justifications that indiscriminately involve nomological and applied models. Rating scale peculiarities, questionable validity, and scholastic homogeneity lead to diverse professional attitudes towards student opinions of professors, with a learner or consumer emphasis occupying the extremes. Several evaluation schemes are noted along with behaviors that tend to produce favorable student opinions.Interest in students evaluating professors in order to improve teaching has a long history and began early in American education (Boyce, 1915). The idea is based on the belief that students, because of their exposure to professors, should know best whether teaching is adequate and whether they are learning.The implementation of the idea is simplicity itself. A set of questions is constructed, questions that refer to the professor's clarity of speech, organization, involvement in student learning, presentation of different viewpoints, fairness in grading, work assignments, and other characteristics. The students are given these questions near the end of the semester, but before the final examination, and they anonymously rate the professor. The ratings are then analysed, related to some performance index, and the professor is, ipso facto, evaluated.The program is so simple, straightforward, and appealing that rating techniques have been developed and used at a number of colleges and universities (Centra, 1979). The ratings are used in selecting professors for salary increases, promotions, and tenure, and are sometimes published for the benefit of the students in course selection. Although student ratings are only one part of evaluation programs and separate fromchairperson evaluations, faculty peer ratings, informal student opinions, and other traditional methods of evaluations, they deserve special attention because of their widespread use and possible misinterpretatio m After three-quarters of a century, controversy still exists over interpretations and use of student ratings (Cohen, 1983;Dowell and Neal, 1983;Dunkin and Barnes, 1986;Murray, 1980). Diversity of opinion ranges from "Teacher evalu-

In Experiment I some discriminative functions of food pellets were studied by developing a multiple schedule of reinforcement (mult FR 30 Fl 3) in which the delivery of a standard laboratory food pellet as a reinforcer set the occasion for reinforcement on every 30th response (FR 30), and the delivery of a sucrose food pellet as a reinforcer set the occasion for reinforcement after a 3-min interval (Fl 3). Discriminative stimulus control by the type of pellet was also demonstrated by reversing the operant discrimination and having the standard pellet control the FI 3 and the sucrose pellet control the FR 30. In Experiment II a mult FR 30 FR 30 with two bars was developed; a standard food pellet was followed by an FR 30 on Bar 1 and extinction (ext) on Bar 2, while a sucrose pellet was followed by an FR 30 on Bar 2 and ext on Bar 1. A control rat was placed, for comparison, on a mixed (mix) FR 30 FR 30 schedule with two bars, but neither bar correlated with the type of food pellet. In Experiments I and II the similarity between pellet controlled multiple schedules and multiple primed schedules was discussed, as was the comparability of transitions and effectiveness of control between pellet controlled multiple schedules and multiple schedules providing continuous exteroceptive stimuli.The purpose of this study was to investigate some discriminative functions of reinforcing stimuli by using two types of food pellets as reinforcers and following each with a different schedule of reinforcement. These procedures involve the differential reinforcement of responses as a consequence of the prior type of pellet. If the discrimination is effective the process results in changes in the rate of responding occasioned by stimuli correlated with presentation of the reinforcer and with the appropriate consummatory behavior.EXPERIMENT I A multiple schedule of reinforcement is characterized by programming reinforcements on two or more schedules, each schedule being correlated with an appropriate controlling stimulus (Ferster and Skinner, 1957). The first two-ply multiple schedule developed in this study involved presenting a standard formula (St) pellet and following it with a reinforce-

No abstract