THE PURPOSE of this article is to report the design and development of a device which has promise for partially alleviating one of the most frustrating problems with which the investigator of the teaching-learning process (as it occurs) in ? the classroom setting has to cope?the systematic and efficient collection of data in a form that lends itself to subsequent analysis. machine design reported here will scan a maximum of 40 persons at 30 second intervals over a period of 60 minutes for one of five individ ually selected responses per person. Each stu dent's responses are automatically recorded by person as a function of time. design should prove valuable to an investigator in the following areas, among others: 1) in-service training of teachers by modification of the student-teacher's behavior through feedback to the teacher of spe cific student reactions to identifiable teacher ac tions, 2) analysis through correlation of trained observer and student reactions to specific teacher behaviors, 3) a search for patterns of student re action to teacher behaviors according to varia tions in selected background factors of students, 4) analysis of effective teacher behaviors for spe cific purposes, and 5) interaction analysis. Historically, a variety of approaches have been utilized in the research studies of teacher effec tiveness. These have included 1) student listing of characteristics of most-liked and most-disliked teachers, 2) careful study of clusters of teachers judged to be effective or ineffective to identify attributes of teachers in each category, 3) evalu ation of teachers in training and correlation of factors with subsequent judgments of their suc cess in the actual teaching situation, 4) ratings by panels of experts, etc. Investigations into this area of the effective and less effective teacher were conducted for many years by Barr at the University of Wisconsin. In the 1956 Encyclopedia of Educational Research, Barr (1) reported as follows: The ultimate measure of teacher effectiveness, particularly in his teacher-pupil relations, will be found in the changes produced in the pupils under his direc tion. Hence it seems sound to attempt the evalu ation of teaching efficiency on the basis of pupil growth, but a practical procedure has not yet been developed. large number of investigations in the area of teacher competence is further evidenced by the fact that Domas and Tiedeman (3) reported 1006 annotated references to studies of teacher compe tence for the period 1929-49. In addition, Barr (2) reports there has been an increase in the fre quency of studies since 1949. Despite the volume of research efforts such as cited above, the findings reported in the literature are frequently inconclusive because of conflicting findings. complex nature of the problem is certainly a primary cause. A brief statement of the variables involved throughout the stages of the learning process suggests the many obvious sources of variation which weaken experimental controls and preclude replicating much of the published research. When the student is presented with a learning task, basic sets of conditions are identifiable: variation in the nature of the task from simple rote memorization to more meaningful material at different degrees of difficulty ; variation in teacher style procedures, and classroom organization; and variations in the perception of the learner with his set of personality characteristics includ ing intelligence, past achievement, and social at tributes. He responds during the presentation of the learning task and a measure of learning or achievement is evaluated. nature of his re sponses or response pattern is difficult to assess during the learning process. In attempting to circumvent the measurement of the complex of variables inherent in the learn ing process, researchers have frequently fallen back on 1) opinion and judgment by observers, 2) correlation studies between ratings of teacher attributes, and a) the student teacher rating, b) in-service rating, c) college grades, d) pupil growth in achievement, e) judgments of panels of experts. Two major weaknesses characteristic of much of the research include the identification