COMMONALITY ANALYSIS: A METHOD FOR DECOMPOSING EXPLAINED VARIANCE IN MULTIPLE REGRESSION ANALYSES

Commonality analysis is a procedure for decomposing R² in multiple regression analyses into the percent of variance in the dependent variable associated with each independent variable uniquely, and the proportion of explained variance associated with the common effects of predictors. Commonality analysis thus sheds additional light on the magnitude of an obtained multivariate relationship by identifying the relative importance of all independent variables, findings which can be of theoretical and practical significance. In this paper we offer a brief explication of commonality analysis; a step-by-step discussion of how communication researchers may perform commonality analyses using output from computer-assisted statistical analysis programs like SPSS; and we provide an extended example illustrating a commonality analysis.     

ISSUES CONCERNING THE USE OF REGRESSION ANALYSIS IN SALARY DISCRIMINATION CASES

Twenty-nine salary discrimination court cases were examined to determine ways an organization can refute a regression analysis that leads to the inference of discrimination in compensation. It was found to be equally effective for the organization to introduce its own regression, other statistics, or no statistics. The plaintiff won all of the cases when discrimination was proven in promotion. The defendant won all cases when discrimination was not proven in promotion. Defendants also won all cases when plaintiffs inappropriately treated jobs as fungible or when they failed to include important variables. Market variables may be included in regressions if defendants can show they were applied consistently to determine salaries. The implications of the courts' acceptance of some regressions for comparable worth is presented, and a list of recommendations is made for organizations that may be faced with results from regression analysis in court.     

A COMPARISON OF THE EFFECTS OF FIXED AND VARIABLE RATIO SCHEDULES OF REINFORCEMENT ON THE BEHAVIOR OF DEAF CHILDREN

The performance of all five students in an adjustment class for deaf children was compared under fixed and variable ratio schedules of reinforcement. During the fixed ratio (FR) condition, students earned checks if they were attentive and did not engage in disruptive behavior. These checks could be exchanged for the opportunity to draw a prize from a grab bag. During the variable ratio (VR) condition, they earned a draw from the grab bag according to a variable ratio schedule with a mean ratio equal to the value of the preceding FR schedule. During the VR condition, students visually attended more and engaged in disruptive behavior less often than during the FR condition. The number of math problems completed per minute was also higher during the VR condition although no specific consequences were applied to math performance.     

EFFECTS OF EQUAL AND UNEQUAL REINFORCER DURATION DURING FUNCTIONAL ANALYSIS

In the functional analysis described by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994), reinforcer duration varied across conditions (e.g., brief attention vs. 30 s of escape); this may result in unequal exposure to the establishing operations for aberrant behavior. In this study, we compared the effects of unequal and equal reinforcer duration during a functional analysis. The results showed that reinforcer duration affects the rate of aberrant behavior and may potentially alter functional analysis interpretation.     

COMPONENTS OF ARTIFACTUAL VARIANCE IN META-ANALYTIC RESEARCH

Except for a few rough estimates, the literature contains no precise data on the relative size of each correctable artifact in a meta-analysis. The purpose of the present study was to determine these proportions and see if artifacts were related to various sample and subpopulation parameters. The study confirmed the notion that sampling error accounts for the bulk of the total artifactual variance (i.e., more than 90% for small or medium samples and more than 70% for large samples, e.g., N = 500). Nevertheless, variation in the percentages of each artifact was found to be primarily a function of sample size and mean correlation. Total artifactual variance was found to be a function of subpopulation correlation differences and sample size. Practical implications of the results are presented.     

交互分析理论对改善人际交往困惑的应用

在心理咨询中,会遇到很多"老好人"来访者,这些老好人因不会拒绝别人而有苦难言。为更好地解决"老好人"式来访者的心理困惑,作者在交互分析理论的"人间游戏"基础上,发展并诠释了一个新的游戏——"老好人游戏"。通过对这一游戏模式的剖析,帮助咨询师充分了解"老好人"的行为模式及其背后的内心状态,对症咨询,真正实现心理咨询"助人自助"的目的。     

A PRELIMINARY ANALYSIS OF INTERACTIVE EFFECTS BETWEEN COMMON CLASSROOM CONTINGENCIES AND METHYLPHENIDATE

To assess the drug—behavior interaction effects with an 8-year-old boy with attention deficit hyperactivity disorder, common classroom antecedent (e.g., seating arrangement) and consequent (e.g., peer prompts) stimuli were alternated within a school day while drug conditions (methylphenidate vs. placebo) were alternated across days. The results suggested that peer attention maintained disruptive behavior when methylphenidate was absent but not when it was present.     

THE FACILITATIVE EFFECTS OF INCIDENTAL TEACHING ON PREPOSITION USE BY AUTISTIC CHILDREN

In a comparison of incidental teaching and traditional training procedures, three language-delayed autistic children were taught expressive use of prepositions to describe the location of preferred edibles and toys. Traditional highly structured training and incidental teaching procedures were used in a classroom setting, and generalization was assessed during free-play sessions. Results clearly indicate that incidental teaching promoted greater generalization and more spontaneous use of prepositions. These findings have important implications for language programming and teacher training, suggesting that incidental teaching should be included as a standard component of language development curricula for autistic and other developmentally delayed children.     

EFFECTS OF PICTURE PROMPTS ON THE ACQUISITION OF COMPLEX VOCATIONAL TASKS BY MENTALLY RETARDED ADOLESCENTS

The effects of using picture prompts on the acquisition, generalization, and maintenance of complex vocational tasks were evaluated within a multiple baseline design across subjects and tasks. Five moderately and severely mentally retarded adolescents were first trained to use picture prompts to guide their performance on one or more complex tasks. Following training, posttesting with and without the picture prompts was conducted to evaluate the effects of training and to determine maintenance effects over a 2- to 4-week interval. Generalization of performance across tasks was assessed with three of the students who were provided novel tasks (with and without picture prompts) without training. Results indicated that picture prompts can be successfully used to promote both acquisition and generalization of performance, and that subsequent training time on a novel task was reduced when the use of picture prompts had been previously trained.     

ACQUISITION OF SIGN LANGUAGE BY AUTISTIC CHILDREN II: SPONTANEITY AND GENERALIZATION EFFECTS

Autistic children typically do not use their language repertoire in order to communicate. Six autistic children who exhibited poor communication skills were trained to use their sign repertoire to make spontaneous requests of adults. Training consisted of imitative prompting, fading, and differential reinforcement, and included aspects of incidental teaching. The children displayed an increase in the rate and variety of spontaneous sign requests (Experiment 1). Generalization of spontaneity across adults (Experiments 1 and 2) and settings (Experiment 2) was also observed. We suggest that spontaneity may be facilitated when language is brought under the control of broadly defined stimuli such as adult attention rather than narrowly defined stimuli such as the presence of specific objects or verbal prompting in the form of questions. Finally, response generalization was observed as well (Experiment 1). Specifically, as spontaneity increased, self-stimulatory behavior decreased. This result may be accounted for in terms of reinforcer competition, reinforcer consistency, or discriminative stimulus effects.     

EFFECTS OF SELF-ESTEEM ON LENIENCY BIAS IN SELF-REPORTS OF PERFORMANCE: A STRUCTURAL EQUATION MODEL ANALYSIS

This study used structural equation model analysis to examine the relationship between self-esteem and leniency bias in self-reports of performance. The results showed that consistent with the prediction of consistency theory (e.g., Korman, 1970), leniency bias was positively correlated with self-esteem. The results also suggest that the relationship between self-esteem and leniency bias was more positive when self-ratings were made on ambiguous performance dimensions rather than on less ambiguous dimensions. Implications for future research and practice are discussed.     

THE EFFECTS OF HAVING BEEN SUPERVISED BY A WOMAN ON PERCEPTIONS OF FEMALE MANAGERIAL COMPETENCE1

Responses of male and female managers in state public human service organizations are analyzed to determine if perceptual differences exist concerning the competence of women as managers in these organizations depending on whether an individual has or has not been supervised by a woman. Significant differences were found in only one area: the motivation of women to manage.     

APPLICATIONS OF MODERATED REGRESSION IN JOB DESIGN RESEARCH

Applications of moderated regression analysis in five areas of job design research are described and fully illustrated with actual survey data. The five areas of research have received some attention in the literature over the past 10 years. Arguments for the selection of moderated regression over the more common subgroups analysis are presented. We conclude that moderated regression analysis is superior to a subgroups analysis for the research questions being asked by investigators interested in job design research.     

THE EFFECTS OF VARIABILITY AND RISK IN SELECTION UTILITY ANALYSIS: AN EMPIRICAL COMPARISON

To date, utility analysis research has derived point estimates of the expected utility value for human resource management programs or interventions. Utility estimates are usually quite large, but they fail to reflect the size and shape of the utility distribution. The present study investigated utility estimate variability for the selection utility of using the Programmer Aptitude Test to select computer programmers in a medium-sized computer manufacturing organization. Utility calculations incorporated financial/economic factors as well as employee flows over time. The distributions for each utility parameter were empirically estimated, and these distribution estimates were combined through a Monte Carlo analysis to yield a distribution of total utility values. Monte Carlo results were compared to three other risk assessment approaches: (1) sensitivity analysis, (2) break-even analysis, and (3) algebraic derivation of the distribution. Results suggest that the distribution information provided by the Monte Carlo analysis more completely described the variability and riskiness associated with the expected utility value. Future research suggested by these findings is discussed.     

EFFECTS OF WAIT-TIME AND INTERTRIAL INTERVAL DURATIONS ON LEARNING BY CHILDREN WITH MULTIPLE HANDICAPS

We investigated the influence of teacher wait-time and intertrial interval durations on the performance of 4 multiply handicapped students during instruction in 10 skills. Four experimental conditions were evaluated: long wait-time and long intertrial interval, long wait-time and short intertrial interval, short wait-time and long intertrial interval, and short wait-time and short intertrial interval. Instructors attempted to keep short intervals as close as possible to 1 s and long intervals as close as possible to 10 s for both variables. Results showed that student performance was superior under the long wait-time conditions irrespective of the length of the intertrial interval.     

ON THE DETERMINATION OF VARIANCE ATTRIBUTABLE TO PREVIOUSLY ISOLATED FACTORS

H olley , J. W. & E lmgren , J. On the determination of variance attributable to previously isolated factors. Scand.J. Psychol ., 1969, 10 , 74–80.—The paper describes a method for extracting variance which can be attributed to factors which have been isolated previously. An illustration is given, using hypothetical data.