ACCURACY OF SELF-ASSESSMENT OF ON-TASK BEHAVIOR BY ELEMENTARY SCHOOL CHILDREN1

Being able to assess one's own performance would seem to be prerequisite to most forms of self-management. The present study investigated the extent to which children in a typical second-grade classroom could accurately assess their own academic on-task behavior. In essence, this investigation replicated an experiment by Glynn and Thomas (Journal of Applied Behavior Analysis, 1974, 7 , 299–306) and extended those findings, by studying self-assessment over a longer span of time. The 31 children in the class ranged in age from 7 yr five months to 8 yr six months. Although only eight “problem” children were observed and studied, the self-control treatment procedures were applied to all children in the class. Behavior was observed daily during mathematics lessons for over eight school weeks. The study utilized a five-phase ABABC design, where the A phases were baseline periods, B phases were self-control periods, and C phase was a postcheck. In the self-control phases, children were asked by the teacher to put a check on their own individual performance card if they were on-task when a signal sounded. If they were not on-task, they were instructed not to touch their card. Typically, 15 to 20 signals occurred per 40-min session. At the end of the lesson, children were allowed to choose games to play according to how many checks they had earned. During all experimental phases, observers measured children's on-task behavior. Other observers measured the children's accuracy of self-assessment in the self-control phases. The results showed noticeable increases in the daily mean on-task behavior scores in the self-control phases following baseline phases. Also, a high level of on-task behavior was maintained in the postcheck phase. The variances of on-task behavior scores during the self-control phases were noticeably smaller than those in the baseline phases. Generally, individual data reflected group results, with some minor differences. Levels of accuracy of self-assessment varied from very accurate (95% of the time) to relatively inaccurate (56% of the time). The overall level of accuracy for all children was 78%. No consistent individual pattern appeared across phases. Some children got better as the study proceeded and some got worse. Most subjects tended to give themselves too much reinforcement, rather than too little. Peer social surveillance appeared to be the major factor that influenced individuals' evaluation of their own behavior, even though the teacher did make some control remarks in regard to how the children marked their cards. The results also demonstrated that self-control techniques can increase on-task behavior in a classroom with no history of external reinforcement, and that these techniques can have a lasting effect for more than two months of a school year.     

Multiple encoding of word attributes in memory

Two experiments employing the release from PI method were conducted to determine whether degree of release could be predicted by the amount of denotative attribute overlap between the control and experimental groups. This theory was very appropriate for predicting the results of the first experiment, but was less successful in the second unless a connotative attribute is also introduced. It is suggested that this class of attribute should be included in semantic memory conceptualization.     

Conditioning in human opiate addicts

Eight volunteers maintained on daily methadone participated in a classical conditioning procedure to determine which if any of the elements of narcotic withdrawal could be conditioned. The unconditioned stimulus was the injection of a small dose of naloxone. The unconditioned response was a brief precipitated withdrawal syndrome. The conditioning stimulus was a tone, odor, and injection of saline. Conditioning was successful in the pilot study in 5 of 8 subjects. The conditioned response consisted of tearing, yawning, lacrimation, systolic blood pressure increase, respiratory irregularities and subjective feelings of narcotic withdrawal sickness (nausea, muscle aches, chills). A second group of 8 subjects showed, in addition to the above, evidence of conditioning of heart rate, respiratory rate and skin temperature decrease. These laboratory findings support the clinical reports of a conditioned withdrawal syndrome and suggest ways to improve treatment results by detecting and extinguishing or modifying conditioned responses.     

Comparison of the WISC and WISC-R on a psychiatric adolescent inpatient sample

The present study compared the WISC and WISC-R on a group of psychiatric adolescent inpatients. The results indicated high intercorrelations between the two test versions but with the correlations linearly related to the changes in content and administration rules of the WISC-R. Analyses of variance indicated that, on the basis of the Full Scale IQ, the WISC is an easier test than the WISC-R.The authors gratefully acknowledge the assistance of Ms. Kathleen Lennox for her aid in data analysis and computer programming.     

Herbert Blumer's critique of The Polish Peasant: A post mortem on the life history approach in sociology

Herbert Blumer's appraisal of William Thomas and Florian Znaniecki's Polish Peasant study and the panel discussion of Blumer's critique, sponsored by the Social Science Research Council in 1938, occasioned a discussion of the value of personal documents in social research which is indicative of the controversy surrounding their use in the interwar period. It is suggested that most sociologists writing in the 1930s were skeptical of the value of personal document research, because such research could not be conducted along the lines of experimental procedure in physical science. Personal document research was in decline by 1938. The critique and conference constituted a post mortem examination of this approach to social analysis.     

A comparison of BASIC language timing loops for the TRS-80 microcomputer

For users who wish to program the TRS-80 microcomputer exclusively in BASIC, the FOR-NEXT timing loop provides almost twice as fine temporal resolution as the incrementing loop. A temporal resolution of 6.62 msec is possible with the FOR-NEXT loop.