The use of extreme groups in assessing relationships

The relationship between variables in applied and experimental research is often investigated by the use of extreme (i.e., upper and lower) groups. Earlier analytical work has demonstrated that the extreme groups procedure is more powerful than the standard correlational approach for some values of the correlation and extreme group size. The present article provides methods for using the covariance information that is usually discarded in the classical extreme groups approach. Essentially, then, the new procedure combines the extreme groups approach and the correlational approach. Consequently, it includes the advantages of each and is shown to be more powerful than either approach used alone.     

Note on use of extreme criterion groups in item discrimination analysis

The purpose of this note is to reconsider the Kelley-Cureton definition of optimal extreme groups for estimating item-criterion correlations. Optimal tail per cents are derived, using the criterion of minimum sampling variance of the tetrachoric correlation coefficient, and the findings are related to earlier work of Mosteller. It is shown that upper and lower 27 per cent groups yield the most precise estimate of the tetrachoric coefficient only when the population correlation is close to zero. When the population value is .4, extreme 20 per cent groups provide estimates with the smallest sampling error variance. It is further shown, however, that 27 per cent extremes yield highly efficient estimates. Thus no change is recommended in traditional item analysis procedures.     

Drawing out resistance: The use of the house-tree-person test to facilitate communication in verbal therapy groups

Resistance as an impediment in both individual and group therapy is discussed, and the special resistances common to groups are described through a review of the literature. An adaptation of John Buck's House-Tree-Person projective test is suggested as a method for overcoming resistance that can be used by group therapists, including those unfamiliar with art therapy techniques. Case examples are presented of a group session in which resistance is a prominent deterrent to progress and another in which the use of the projective drawing technique facilitates the dissolution of resistance.     

The relationship between the validity of a single test and its contribution to the predictive efficiency of a test battery

Typical selection or classification testing programs should provide for improvement of the predictive efficiency of the test battery. Such provision calls for the administration of experimental tests along with the operational battery administration and follow-up analysis to determine the value of the experimental material. It is possible to determine without waiting for criterion data what the validity of the experimental test must be in order to improve the battery validity. The method together with the proof is presented.     

A test of the Olson Circumplex Model: examining its curvilinear assumption and the presence of extreme types

The debate over the usefulness of different family models continues. Recent attention has been paid to comparisons between the Olson Circumplex Model and the Beavers Systems Model. The present study seeks to contribute evidence that bears directly upon one of the most fundamental points of controversy surrounding the Olson model--the linear versus curvilinear nature of the cohesion and adaptability dimensions. A further contribution is an examination of the actual occurrence of the Circumplex Model's extreme types in a clinical population.     

The relationship of adolescent peer groups to the incidence of psychosocial problems

This study combined a treatment sample (n = 127) and a control sample (n = 114) of adolescents (ages 13-17) to investigate the relationship between adolescent peer groups and incidence of psychosocial problems. A content analysis resulted in four separate types of peer groups. The group with the lowest level of involvement in school activities was labeled by other adolescents in negative terms. An analysis of variance indicated that the least involved and most negatively labeled group generally had the most positive attitudes toward alcohol and drug use, the lowest levels of perceived harm due to alcohol and drug use, and the highest levels of alcohol abuse, drug use, delinquency, and depression. This group also had the lowest level of self-esteem, most external locus of control, least perceived access to occupational opportunities, and highest level of societal estrangement. The results are interpreted as providing support for both control and labeling theories. Implications for theory-based intervention are discussed.     

Use of the extreme groups approach: a critical reexamination and new recommendations

Analysis of continuous variables sometimes proceeds by selecting individuals on the basis of extreme scores of a sample distribution and submitting only those extreme scores to further analysis. This sampling method is known as the extreme groups approach (EGA). EGA is often used to achieve greater statistical power in subsequent hypothesis tests. However, there are several largely unrecognized costs associated with EGA that must be considered. The authors illustrate the effects EGA can have on power, standardized effect size, reliability, model specification, and the interpretability of results. Finally, the authors discuss alternative procedures, as well as possible legitimate uses of EGA. The authors urge researchers, editors, reviewers, and consumers to carefully assess the extent to which EGA is an appropriate tool in their own research and in that of others.     

In support of extreme groups design for studying masculinity-femininity and intelligence

Humphreys (this issue) reanalyzed the data of Welsh and Baucom (1977) and stated that their conclusions were unsupportable. Humphreys's criticisms are inappropriate since his findings are based upon a measure of masculinity-femininity (M-F) which is different from the previous approach and based upon a sample which is somewhat different. Thus he addresses issues which are inconsistent with Welsh and Baucom's focus. Finally, Humphreys's insistence on correlational analysis to the exclusion of extreme groups design is viewed as inappropriately limiting.     

The effects of alcohol and extreme provocation on the use of a highly noxious electric shock

Intoxicated and nonintoxicated subjects were given the opportunity to administer a potentially injurious level of shock to an opponent who behaved in either a provocative or an extremely provocative manner. The provocative opponent attempted to administer high-intensity shocks to the subject. The extremely provocative opponent attempted to administer high-intensity shocks and the potentially injurious level of shock. The intoxicated subjects in the extreme provocation condition evidenced the greatest increase in the use of the potentially injurious shock. It was concluded that aggression is a function of the interaction of alcohol consumption and level of provocation.     

Relative costs and statistical power in the extreme groups approach

The relationship between variables in applied and experimental research is often investigated by the use of extreme (i.e., upper and lower) groups. Recent analytical work has provided an extreme groups procedure that is more powerful than the standard correlational approach for all values of the correlation and extreme group size. The present article provides procedures to optimize power by determining the relative number of subjects to use in each of two stages of data collection given a fixed testing budget.The opinions or assertions contained herein are those of the writers and are not to be construed as official or reflecting the views of the Navy Department.Edward F. Alf, Jr. is also affiliated with the Department of Psychology, San Diego State University.     

Differentiation of political parties by two extreme political groups

Abstract.— The Rokeach hypothesis, that one tends to differentiate better around ones own ideological position, was tested by letting socialist and conservative subjects pass similarity judgments on eight nonvegian political parties. As expected, the conservative subjects differentiated better among the conservative parties than the socialist subjects did, and vice versa for the socialist parties. Factor analysis of the similarity matrix within each subject group yielded, as predicted from a secondary hypothesis, two factors, socialistic and nonsocialistic.     

The use of configural analysis for the evaluation of test scoring methods

A method based on configural analysis has been given whereby test scoring techniques can be evaluated to see if they have optimal validity. Configural analysis has also been used to show how three well known item scoring techniques, multiple regression, total score, and multiple cut-off, imply (for optimal validity) certain conditions on the answer pattern means. The method is illustrated by a worked example.We are indebted to Professor James G. Taylor for his helpful suggestions.