A comparison of the efficiency and accuracy of BILOG and LOGIST

Comparisons are made between BILOG version 2.2 and LOGIST 5.0 Version 2.5 in estimating the item parameters, traits, item characteristic functions (ICFs), and test characteristic functions (TCFs) for the three-parameter logistic model. Data analyzed are simulated item responses for 1000 simulees and one 10-item test, four 20-item tests, and four 40-item tests. LOGIST usually was faster than BILOG in producing maximum likelihood estimates. BILOG almost always produced more accurate estimates of individual item parameters. In estimating ICFs and TCFs BILOG was more accurate for the 10-item test, and the two programs were about equally accurate for the 20- and 40-item tests.I am grateful to Robert J. Mislevy, Martha L. Stocking, and Marilyn S. Wingersky for many helpful comments on an earlier version of this paper. I would also like to thank Hamid Kamrani and Bongmyoung Park for getting LOGIST and BILOG running and keeping them running under changing computer systems at CTB/McGraw-Hill.     

SUPPORTED EMPLOYMENT AS AN INTERVENTION PACKAGE: A PRELIMINARY ANALYSIS OF FUNCTIONAL VARIABLES

In this two-experiment investigation, the long-term (at least 6 months) employment of 51 moderately mentally retarded clients who were placed into 64 supported employment positions was first evaluated relative to 10 training and posttraining components that comprised a supported employment training package. In Experiment 1, chi-square analyses were used to identify three components that differentiated successful (employed for at least 6 months) from unsuccessful clients. In Experiment 2, 4 successful clients were further evaluated in a nonconcurrent multiple baseline design to determine whether the three variables identified during the group analysis (client advocate, collateral behavior, and follow-up plan) were included in the training packages. These results are discussed in terms of the need to establish better the functional variables of supported employment training programs.     

BENEFITS AND COSTS OF SUPPORTED EMPLOYMENT: AN ANALYSIS OF THE ILLINOIS SUPPORTED EMPLOYMENT PROJECT

This investigation presents a cost-benefit analysis completed for one of 27 states implementing supported employment as a result of federal funding. Based upon the benefits and costs detailed, society realized a $0.75 return for every $1.00 invested in supported employment in Illinois. Supported employees realized a 37% increase in their earnings over a comparable period. The results of this study are discussed and recommendations are made for similar analyses across states.     

Development of a structured psychiatric Interview for children: Agreement on diagnosis comparing child and parent interviews

Similar structured diagnostic interviews about the child were given by different interviewers to a cohort of 307 mother-child pairs. A diagnosis was made by computer on each interview, using specified criteria. Diagnoses on mother-child interviews were compared using the kappa statistic. Kappas of .30 or higher were found for the diagnosis of antisocial personality, conduct disorder, enuresis, mixed behavior-neurotic disorder, and possible depression. Comparisons were made for sex and age. Possible depression and enuresis were diagnosed reliably at all age levels and for both sexes. The limitations of the interview and diagnostic system used are discussed. The findings support the need for further efforts to develop diagnostic research interviews for use with children and adolescents.     

Learned helplessness: Not dependent upon cognitions, attributions, or other such phenomenal experiences

Subjects shot a light gun at a target with a photorecepter cell in the bull's-eye, with the only information regarding their accuracy being provided by reinforcing tone signals. Half the subjects received reinforcers contingent upon their hits. The others were yoked to the contingent subjects, receiving non-contingent reinforcers in the same patterns. Experiment 1 compared contingent or noncontingent positive or aversive reinforcers in their effect on subsequent anagrams performance. Phenomenal experiences, such as cognitive awareness, attributions, and moods, were assessed. Subjects exhibited a strong helplessness effect independent of their phenomenal experiences. In Experiment 2 the independent variables of contingent/noncontingent reinforcement and awareness of noncontingency were manipulated orthogonally by informing half the subjects that their reinforcement had been noncontingent in the target-shooting. Actual noncontingency produced a strong helplessness effect whereas “awareness of noncontingency” did not.     

Learned helplessness and defensive strategies: A rejoinder

The results of experiments reported by Oakes and Curtis (1982), Tennen, Drum, Gillen and Stanton (1982), and Tennen, Gillen, and Drum (1982) are seen as inconsistent with the cognitive learned helplessness theory of Seligman and his associates (Abramson, Seligman, & Teasdale, 1978; Alloy & Seligman, 1979). Comments on the Oakes and Curtis studies by Alloy (1982) and by Silver, Wortman, and Klos (1982) are seen as employing three defensive strategies: (1) Declaring the research findings to be irrelevant to the theory; (2) declaring the experiments to be flawed; and (3) modifying the theory to accommodate the research findings. This rejoinder argues that the research findings are relevant and that the criticisms are of questionable validity. It is suggested that the questions of the validity of the theory raised by these findings not be declared to be resolved, but that additional data bearing on the questions be sought.