Francis J. Kelly

Suggested Inferential Statistical Models for Research in Behavior Modification

Operant Conditioning procedures have diffused to applied educational settings where results are not descriptively obvious. Inferential statistics using Multiple Linear Regression Analysis (MLR) to partial variance to answer specific behavioral modification research questions are presented. Four applications of MLR to typical behavioral modification designs are illustrated. The examples represent the case where the research questions dictate the statistical design. Assumptions regarding individual subject design and repeated measurement problems are presented and are seen as minor sources of difficulty for use of inferential statistics in behavioral modification studies.

Further normative data on tests measuring flexibility in cognitive processes.

Two University of Alberta samples and two high school samples were administered two measures of flexibility (Associations IV and Object Naming). Two measures of anxiety and convergent thinking data were obtained for three of the samples. The Alberta samples tended to score lower on Associations IV than an Illinois sample. Split-half reliability of the Associations IV task yielded Spearman-Brown correlations ranging from .60 to .75. The Alberta sample tended to score higher on the Object Naming task by comparison with the Illinois sample. The split-half reliability of Object Naming, however, indicates a questionable relationship between parts (Spearman-Brown Formula rs range from .15 to .57). Anxiety tended to facilitate performance among older Ss and hinder performance among the younger Ss. The two flexibility measures correlated more strongly with measures of convergent thinking than with each other.

REFLECTIVE JUDGMENT: A CROSS-SECTIONAL STUDY OF UNDERGRADUATE EDUCATION MAJORS '

Using a reflective judgment interview based upon Kitchener and Kings 1985 model, the present study investigated reflectivity among undergraduate education majors and student teachers. Each set of students responded during a semi-structured interview to three ill-structured problems in education. These interviews were rated using Kitchener and Kings 1985 scoring manual. No differences were found between responses of the two groups. On two of the three issues the modal response was at Stage 4, while on the remaining issues the modal response was at Stage 5. Issues regarding preparation of teachers are raised, and implications for further study are suggested.

Levels of reflective judgment among noncollege trained adults

The present study assessed the influence of life experiences upon reasoning among three known groups with no college background. Sample 1 is from rural USA, had limited life experiences and were learning independent living skills (all are within the normal range of intelligence). Sample 2 includes working women (semi-skilled, skilled and mid-management) from rural USA with a high school education or less. Sample 3 is from urban Brazil and was made up of nontraditional students working in a high school equivalence program. Samples 2 and 3 were judged to have greater conflict in life experience. Three ill-defined issues were used in a structured interview with a modified reflective judgment format. Samples 2 and 3 scored relatively high on Kitchener and Kings Reflective Judgment scale, and on some issues their scores were comparable to college juniors reported in previous studies. The results are consistent with Piagets notion that density of cognitive conflict provides the condition for cognitive growth.

Reliability of a multidimensional locus of control construct for older adults anticipating transitions in careers.

Reid and Ware in 1974 provided data on college-age adults which suggested that Rotters locus of control paradigm might be better represented as a multidimensional construct. Using a varimax rotation, Reid and Ware reported a three-factor solution was parsimonious and tended to support a multidimensional model. The three factors were labeled Self-control, Social Systems Control, and Fatalism. In light of Supers 1983 view that the first stage of planning a career includes a component of autonomy or locus of control, Reid and Wares multidimensional scale might prove useful for older adults undergoing transition in their careers. Since Reid and Wares college undergraduates represented those in early career development, the present study was conducted with older adult individuals contemplating changes in career to determine the efficacy of the three-factor solution. Only two items did not coincide with Reid and Wares factor structure. Implications for the application of a multidimensional locus of control scale to career-planning were discussed.

Does Piaget's Graphic Collection (Stage I Classification) Exist?

Summary In an otherwise fine presentation of Piagets theory, Brainerd made a number of distorted interpretations of the replication research of Piagets three successive stages in classificatory development, and concludes that “… there is no evidence that Stage I even exists.” Very young children (ages 2.5 to 4.5), according to Piaget, tend to arrange elements (geometrical figures that can be classified according to color, shape, and size) in short or long lines or to place them in some spatial or pictorial arrangement. Piaget calls these arrangements graphic collections, and they represent the first stage of classification. Of the studies cited by Brainerd, two supported the graphic collection stage, the others did not methodologically deal with the issue, and most did not include the age-appropriate Ss. In order to test Piagets classificatory stage development, a replication of the Genevan studies was conducted with 80 2- to 6-year children in Brazil, and the results were in total agreement with Piage...

Predicting speech grades from selected spoken language variables

The purpose of this study was to determine the extent to which selected spoken language characteristics predict grades for an introductory public‐speaking course. A Flesch reading‐ease score, Flesch human‐interest score, and type‐token ratio were computed for the first five hundred words of a public‐speech sample for forty‐five subjects. A multiple regression analysis was computed in which the scores on the spoken‐language characteristics (predictor variables) were regressed on the final grades in the speech course (criterion variable). Results indicated that the final grades in an introductory speech course were significantly related to the number of syllables used by students in public‐speech behavior.

PERTURBATIONS AND COMPENSATIONS IN SOCIAL COGNITIVE CONFLICT A FUNCTIONAL ANALYSIS OF COGNITIVE DEVELOPMENT

Piagets work from 1975 through 1980 added a dynamic aspect to his earlier formal structural model. This new work portrayed change in more functionalist terms and describes cognitive conflict referring to the process of perturbations and compensations to explain the regulations of the network of cognitive cycles. He suggests three types of balances or compensatory responses, alpha, beta, and gamma. These writings, however, ignore social interaction as a contributor to the process of perturbations and compensations. Doise and Mackie suggest that cognitive change and development is facilitated by peer interaction but they ignore the functional analysis of social collaboration. The present report presents data collected using selected Piagetian experiments in contradiction with Doise and Mackies model. 30 children from 8 to 10 yr. of age worked in triads composed of homogeneous and heterogeneous groups, while 20 8- to 10-yr.-olds worked alone during five independent sessions on selected Piagetian tasks on contradictions. Analysis showed in general co-construction produced significant and generalizable progress when compared with those working alone. Analysis of videotapes suggests that compensations in co-construction conditions might be detected. A brief section concerns utility of Piagets network of cognitive cycles. Those preliminary comments were taken from video recording of the social interaction.

An evaluation of a programed textbook with educable mentally retarded children.

A commercial programed arithmetic textbook and two teacher made programs, one using the rote approach, the other using the understanding approach, were evaluated with EMR students as subjects. The teacher made programs were designed to follow suggestions from the attention theory of Zeaman and House (1963). The rote approach was highly significant for learning division facts. Implications for the classroom and suggestions for revision of the commercial program are given.