Additive conjoint isotonic probabilistic models (ADISOP)

The ISOP-model or model of twodimensional or bi-isotonicity (Scheiblechner, 1995) postulates that the probabilities of ordered response categories increase isotonically in the order of subject ability and item easiness. Adding a conventional cancellation axiom for the factors of subjects and items gives the ADISOP model where the c.d.f.s of response categories are functions of an additive item and subject parameter and an ordinal category parameter. Extending cancellation to the interactions of subjects and categories as well as of items and categories (independence axiom of the category factor from the subject and item factor) gives the CADISOP model (completely additive model) in which the parallel c.d.f.s are functions of the sum of subject, item and category parameters. The CADISOP model is very close to the unidimensional version of the polytomous Rasch model with the logistic item/category characteristic(s) replaced by nonparametric axioms and statistics. The axioms, representation theorems and algorithms for model fitting of the additive models are presented.     

Proximal Impact of Two First-Grade Preventive Interventions on the Early Risk Behaviors for Later Substance Abuse, Depression, and Antisocial Behavior

We assessed the immediate effects of two universal, first-grade preventive interventions on the proximal targets of poor achievement, concentration problems, aggression, and shy behaviors, known early risk behaviors for later substance use/abuse, affective disorder, and conduct disorder. The classroom-centered (CC) intervention was designed to reduce these early risk behaviors by enhancing teachers' behavior management and instructional skills, whereas the family-school partnership (FSP) intervention was aimed at improving parent-teacher communication and parental teaching and child behavior management strategies. Over the course of first and second grades, the CC intervention yielded the greatest degree of impact on its proximal targets, whereas the FSP's impact was somewhat less. The effects were influenced by gender and by preintervention levels of risk. Analyses of implementation measures demonstrated that greater fidelity to the intervention protocols was associated with greater impact on behavior ratings and on achievement scores, thus providing some evidence of specificity in the effect of the interventions.     

Functional analysis and treatment of problem behavior exhibited by elementary school children.

A functional analysis involving antecedent events was conducted with 4 students who had been identified as having behavior problems. Off-task behavior was measured while task difficulty and level of adult attention were manipulated during analogue sessions. Results revealed two patterns: Three students displayed higher rates of off-task behavior during difficult tasks, and 1 displayed higher rates of off-task behavior during sessions with low attention. Improved behavior was observed when students were taught an alternative behavior that matched the assessment results.     

The writing of clinical papers: the analyst as illusionist

Rather than begin by defining what we mean when we call a paper 'clinical', I would like to reflect on how such a piece of work originates, how, through the interplay between the written and the spoken word, it becomes clinical and good enough for a professional journal.
In the following, I shall pay particular attention to recent publications from the editorial side, that is, those by Wharton, one of the three editors of The Journal of Analytical Psychology , and by Tuckett, editor of The International Journal of Psycho-Analysis . I shall plead that there should not only be one kind of clinical paper, namely the type that for political reasons seems to be the one that is most wanted at the present time and the aim of which is to put psychoanalysis on more secure foundations than it currently is.     

Cooperative learning and group contingencies

This paper discusses the similarities and differences between cooperative learning and group contingencies. Cooperative learning refers to any methods in which students work together to help one another learn, while group contingencies refer to rewarding students based on the performance of a group. Research on the achievement effects of cooperative learning finds that these methods are effective primarily when they incorporate group contingencies, when groups are rewarded based on the average of their members' individual learning performances. The use of group contingencies within cooperative learning is hypothesized to motivate students to do a good job of explaining concepts and skills to their groupmates, and elaborated explanation is the principal behavior found to account for achievement gains in cooperative learning.     

A BEHAVIORAL-EDUCATIONAL ALTERNATIVE TO DRUG CONTROL OF HYPERACTIVE CHILDREN

A behavioral procedure for controlling hyperactivity without inhibiting academic performance is described. Using a time-sample observational method, the hyperactivity displayed by three school children was recorded during math and reading classes. Concurrently, math and reading performances were measured. The study consisted of two baselines, one while the children were on medication and the second while they were off medication. A multiple-baseline design across the two academic subject matters was used to assess the behavioral intervention, which consisted of token reinforcement for correct academic responses in math and subsequently math and reading. Discontinuation of medication resulted in a gross increase in hyperactivity from 20% to about 80%, and a slight increase in math and reading performance. Introduction of a behavioral program for academic performance, during no medication, controlled the children's hyperactivity at a level comparable to that when they were on drugs (about 20%). At the same time, math and reading performance for the group jumped from about 12% during baseline to a level of over 85% correct. Each child performed behaviorally and academically in an optimal manner without medication. Contingency management techniques provided a feasible alternative to medication for controlling hyperactivity in the classroom while enabling the children to grow academically.     

Self-instruction: An analysis of the differential effects of instruction and reinforcement

This study investigated the impact of training 9 first- and second-grade children to use a full self-instructional regimen, and then differentially reinforced the use of self-instruction only, accuracy only, or both self-instruction and accuracy. Three comparison children received no training in self-instruction and were reinforced for accuracy only. Children improved dramatically in academic accuracy subsequent to self-instructional training, independent of the use of self-instruction and of the specific behavior consequated. Children who were reinforced for using self-instruction did use self-instruction, and those who were not, did not. Comparison group children showed little improvement until training in problem-solving strategies was given after 9 days of reinforcement for accuracy. Self-instructional training is discussed as one type of event that increases the likelihood of accurate performance. Its effectiveness may be explained in terms of a teaching strategy rather than in terms of modifying cognitive processes.     

A new estimator of the uniqueness in factor analysis

A closed form estimator of the uniqueness (unique variance) in factor analysis is proposed. It has analytically desirable properties—consistency, asymptotic normality and scale invariance. The estimation procedure is given through the application to the two sets of Emmett's data and Holzinger and Swineford's data. The new estimator is shown to lead to values rather close to the maximum likelihood estimator.     

Matching models in the analysis of cross-classifications

Inference models motivated by the combinatorial chance literature and the concept of object matching may be used in the analysis of a contingency table if the conditional assumption of fixed row and column totals is imposed. More specifically, by developing a matching reinterpretation for several problems of interest in the prediction analysis of cross-classifications—as defined by Hildebrand, Laing and Rosenthal, appropriate significance tests can be given that may differ from those justified by the more common multinomial models. In the course of the paper the distinction between a degree-1 statistic (based on the relationship between single objects) and a degree-2 statistic (based on the relationship between object pairs) is reviewed in some detail. Also, several specializations are presented to topics of current methodological importance in psychology; for instance, a number of references are made to the measurement of nominal scale response agreement between two raters.Partial support for this research was provided by the National Science Foundation through GSOC-77-28227.     

INDIRECT FACILITATION OF ON-TASK BEHAVIOR PRODUCED BY CONTINGENT FREE-TIME FOR ACADEMIC PRODUCTIVITY1

This study investigated the relationship between on-task behavior and the academic performance of three low achieving and disruptive children in an elementary school special-education class. A number of researchers have postulated that the direct elimination of off-task or disruptive behavior is a necessary prior requirement when increases in academic performance are sought in special-education classrooms composed of low achievers with below-average socioeconomic environments. The present research program investigated this issue and measured the incidence of on-task behavior and its relationship to changes in the performance of a variety of academic tasks brought about by the opportunity to engage in daily free-time activities. Three children from a class of eight “educably mentally handicapped” students were chosen as target subjects, as their repertoires seemed to be especially defective. These children were observed by the first author for 90-minute sessions three days a week, and data regarding on-task behavior were obtained by the use of an interval recording technique. In conjunction with the teacher, the first author also monitored the academic performance of the children in six structured activities—answer completion from boardwork instructions, task completion from handout assignments, comprehension of reading assignments and vocabulary recognition, accuracy of direction-following to audiotaped instructions, quality of copied handwriting, and the matching of appropriate phonetic sounds to pictured objects and situations. Following baseline sessions during which both on-task and academic performance was measured, several experimental sessions were conducted during which free-time was given noncontingently. During these sessions, no reliable changes in student performance were observed. Subsequently, the teacher and experimenters defined a specified percentage of correct completion of each of the six categories of skill performance as necessary for the achievement of the free-time reinforcer, during each experimental session. As a result, both the academic productivity and the on-task measures of behavior were observed to increase, even though reinforcement was not directly contingent upon on-task behavior. The subsequent reinstitution of noncontingent free-time was followed by lower rates of both output and on-task behavior. The final reintroduction of contingent free-time for academic productivity again produced substantial increases in both work accomplished correctly and on-task behavior. Corresponding increases in productivity were noted for most nontarget children as well. While the present investigation has shown a high correspondence between productivity and on-task behavior, we must keep in mind that on-task behavior can be defined in many different ways. It remains possible that differences in the degree of correlation between outcome measures and the topography of “work” activity may relate to the degree to which the specific on-task behaviors measured are actually required for the outcome behavior. The failure of many researchers to identify a direct relationship between “appearances” and output may be inherently a measurement problem.