A SUPERVISORY STRATEGY TO IMPROVE WORK PERFORMANCE FOR LOWER FUNCTIONING RETARDED CLIENTS IN A SHELTERED WORKSHOP

A multiple component strategy was investigated for aiding staff responsible for supervising production of lower functioning retarded clients on contract tasks in an institution-based sheltered workshop. The strategy was assessed in a combined multi-element, multiple baseline across groups design with a reversal component. Production performance increased during the production supervisory strategy with all 16 clients, with the range of increase varying from a few percentage points to 150% of baseline production. Both the clients and staff (responsible for conducting the research) preferred working under experimental conditions rather than under baseline conditions which approximated those found in “typical” sheltered workshops.     

Children's knowledge of the separation of variables concept

Children 8, 10, and 13 years of age were either given prior experience constructing seriations or matrices with two-dimensional stimuli, or no prior experience. Knowledge of the separation of variables concept was then assessed by examining children's use of controlled tests of a problem variable's effect (probe questions), and recognition of the inadequacy of confounded variable tests (negative counter-suggestion questions). Although many children either passed both questions or failed both questions, a significant number of children correctly used controlled variable tests, and yet failed to recognize the inadequacy of a confounded test. These results suggest that although a preliminary form of the separation of variables concept is characterized by use of controlled variable tests, subsequent refinement of the concept includes recognition of the inadequacy of confounded tests. Relative to seriation, prior experience constructing matrices facilitated recognition of a confounded test's inadequacy. Constructing matrices may have focused attention on the simultaneous variation of problem elements on multiple dimensions and thus facilitated recognition of the necessary role of both variation of stimuli on test dimensions and equality of stimuli on nontest dimensions, e.g., a controlled test.     

A new measure for determining children's subjective organization of speech

A new technique for determining children's subjective organization of speech is described. The technique consists of teaching a child a sentence and then repeatedly (1) reading the sentence up to a selected point, (2) sounding a tone, and (3) having the child supply the next word. The time from the tone to the onset of the child's voice is the dependent variable. In the first study 6- and 8-year-old children memorized the following sentence: “John likes to feed the little puppies in the barn.” Their response times suggested that the dominant segmentation pattern consisted of four phrases: a subject phrase, a verb phrase, an object phrase and a modifier phrase. Within each phrase differences between reaction times were on the order of 30 msec. Across phrases they were on the order of 120 msec. The same trend emerged for subsamples of the children with various short-term memory sizes, and for adults. It also appeared for subsamples of the total scores (e.g., first six trials and last six trials). In order to provide further evidence of the technique's validity, different sentences of the same general type were presented to children with short-term memory spans of either 3, 4, or 5. As predicted, it was found that when children's spans corresponded to the number of hypothesized chunks (i.e., 4), they could repeat the sentences verbatim on the first exposure. They could not do so, however, when their spans were lower than that value or when the number of phrases was increased. It was concluded that the new technique is a promising one for assessing young children's patterns of speech segmentation.     

Developmental trends in distractibility: Is absolute or proportional decrement the appropriate measure of interference?

Since baseline performance varies with age, diminished interference effects found with increasing age do not provide evidence for a developmental trend in distractibility unless their relationship to baseline performance is known. In the present study, baseline difficulty was varied in two speeded classification experiments to investigate whether interference effects due to irrelevant information could best be characterized by absolute or proportional decrement models. In Experiment 1, first graders, third graders, and adults sorted cards on the basis of square size, both in the presence and absence of extraneous stimulus information. In Experiment 2, first graders and adults sorted cards on the basis of the density of a regular array of dots and two kinds of irrelevant information were used. Robust interference effects were found which declined with age. Manipulating discriminability of the relevant stimulus dimension resulted in large changes in sorting time, but interference effects did not vary with baseline difficulty. These results were interpreted as strongly supporting both an absolute decrement model and a developmental trend in distractibility.     

Methinks they do protest too much

A serious methodological error was committed by Stephens et al. (1972), the protestations of the present critics notwithstanding. When studying the relationship between two sets of variables, it is clearly wrong to hold chonological age constant in one set and allow it to vary in the other. This error spuriously reduces the correlation between the two sets. A demonstration that the two sets define different first-order factors provides very little information about the degree of relationship present, which must be derived from the intercorrelations of those factors. In the original research, these were also biased by the methodological error. The conclusion of Humphreys and Parsons that Piagetian tasks and intelligence tests could be used interchangeably is reinforced. The hypothesis that the relatively small amount of unique variance in the Piagetian composite is relatively unimportant psychologically can only be discarded on the basis of good empirical research, not debater's arguments.     

Effects of neutral and positive stimuli on stuttering: “Calling attention to stuttering” revisited

This study investigated the effects of presumably positive and neutral response-contingent stimuli on the occurence of stuttering. Eight adult stutterers spoke spontaneously during a 60-min experimental session, which was divided into three 20-min periods. During Condition I no stimuli were delivered. During Condition II a digital counter and a pair of yellow lights were activated each time the subject stuttered. During Condition III these same response-contingent stimuli were delivered, but with the additional factor that the subject would earn 5¢ each time the counter moved forward and the lights were illuminated. Conditions II and III were counterbalanced across subjects. The experimenter was unaware of which of the two contingencies were operative during Conditions II and III. Results showed no significant changes in occurence of stuttering due to either response-contingent stimulus. A discussion of theoretical and experimental implications is given.     

Psychopathy and structure of primary mental abilities

The Comprehensive Ability Battery (CAB) was used to assess 18 primary mental abilities in 53 white, male prison inmates for whom highly reliable ratings of psychopathy were available. None of the correlations between ratings of psychopathy and scores on the CAB tests was significant, and the overall pattern or structure of abilities was the same for inmates with low and high ratings of psychopathy. The possibility is raised that the personality characteristics of psychopaths make them appear to be brighter and more creative than they actually are. Compared with norms for male high school students, the inmates as a group generally performed well, falling near the 50th percentile on 11 of the tests. The inmates performed significantly better than the normative sample on tests of verbal ability, esthetic judgment, aiming, and representational drawing, and significantly worse on tests of ideational fluency, spontaneous flexibility, and word fluency.This research was supported by Grant MT-4511 from the Medical Research Council of Canada.     

Statistical significance tests for binormal ROC curves

Statistical significance tests are derived and evaluated for measuring apparent differences between an obtained and an expected binormal ROC curve, between two independent binormal ROC curves, and among groups of independent binormal ROC curves. A binormal ROC curve is described by two parameters which represent the spread of the means and the ratio of the standard deviations of the two underlying Gaussian decision variable distributions. To test the significance of apparent differences between or among ROC curves, approximate χ² statistics for each of the three tests were constructed from maximum likelihood estimates of the two parameters defining the binormal ROC curve. The performance of each test statistic was evaluated by simulating five-category rating scale data with equal numbers of noise and signal-plus-noise trials (set at 50, 250, and 500) for each of three typical ROC curves. For the significance test involving only one ROC curve, rating scale data were generated from the chance diagonal of the ROC space also. Although test performance was found to be somewhat dependent on the number of trials and on the location of the ROC curve in the ROC space, comparisons of the obtained and expected fractions of (falsely) significant results at various α levels showed the proposed statistical significance tests to be reliable under practical experimental conditions.     

A set-theoretic random utility model of choice behavior

A random utility model of choice was developed by combining the basic ideas of the well-known theories of Thurstone and Restle. The new model has exactly the same number of free parameters as Tversky's Elimination-by-Aspects model. Furthermore, both models were found to fit, with equal accuracy, the data reported by Rumelhart and Greeno, and Tversky. It was concluded that although the two theories are not identical, they may be difficult to discriminate empirically.