Characteristics of useful evaluation data: An example from community mental health centers

The results of an empirical study of the use of evaluation data in community mental health centers are reported. A mailed survey on evaluation use was conducted among the directors of 164 community mental health centers in 19 states; 140 completed questionnaires were returned. Results indicate that certain types of data have important impacts in a majority of centers. Systems resources management data were most highly used, followed by need assessment data, and client utilization data. Least used were data on outcomes of intervention and community impact. Data use appears closely tied to the utility of the data in carrying out priority management tasks in a center. Findings have important implications for community psychologists who plan, administer, or evaluate mental health services. The broader role of evaluation in community psychology is also discussed.     

Integrating Systems of Care in California for Youth with Severe Emotional Disturbance IV: Educational Attendance and Achievement

Empirical results are presented from school-based interventions in three California counties that are pioneering participants in a carefully evaluated effort to create an innovative system of care for youth with severe emotional disturbance. Data are presented about the educational attendance and achievement of youth enrolled in clinical and academic programs designed to provide collaborative mental health and education services. Attendance levels were uniformly high. The youth were below expected grade level as measured by standardized tests at program admission. Across programs, two counties demonstrated grade level increases of one year or more for one year in school on all subscales of established measures of educational achievement. The third county demonstrated increases of slightly less than one year on two of three subscales of the educational achievement measures. These results are the first from a multisite demonstration of integrated education and mental health programs embedded within a broader system of care for youth. The findings, combined with those from our prior studies, illustrate that is possible to reduce and control placements in restrictive levels of care while improving the academic performance of a vast majority of the youth enrolled in specialized programs within the care systems.     

Race,gender, and the response to stress: Autoworkers' vulnerability to long-term unemployment

A three-wave panel study of auto plant closings focused on the mental health effects of unemployment on blue-collar workers. This paper explores how the impact of long-term unemployment varies across race and gender. We also examine whether other demographic factors can themselves modify the impactsof race and gender. Dependent variables include two measures of distress and two drinking measures. Results showed that the effect of long-term unemployment on distress and drinking was more severe among less educated workers, and responses of blacks were especially sensitive to level of education. In addition, men showed a greater association of long-term unemployment with depression (and to some extent anxiety) than did women. Marriage affected the responses of men but not of women, and of whites but not of blacks. Explanatory variables—the worker's experiences of financial hardship, other negative life events, and lack of a confidant—largely accounted for male-female differences. We conclude by discussing theoretical implications of these effects and address the limitations of the traditional term “vulnerability” in describing them. A prior version of this paper was presented at the North Central Sociological Association meeting in Dearborn, Michigan, April 25, 1991. Funds for the research were provided by the Michigan Health Care Research and Education Foundation and by the International Union-UAW. We are especially grateful to the men and women of the UAW, whose experiences and views form this investigation. We thank Kiseon Chung, Scott McKearney, David Rauma, and the anonymous reviewers for their comments on earlier drafts.     

Are Machines Gender Neutral? Gender-Stereotypic Responses to Computers With Voices

This study tested whether computers embedded with the most minimal gender cues will evoke gender-based stereotypic responses. Using an experimental paradigm (N = 40) that involved computers with voice output, the study tested 3 gender-based stereotypes under conditions in which all suggestions of gender were removed, with the sole exception of vocal cues. In all 3 cases, gender-stereotypic responses were obtained. Because the experimental manipulation involved no deception regarding the source of the voices. this study presents evidence that the tendency to gender stereotype is extremely powerful, extending even to stereotyping of machines.     

Eyewitness Testimony: The Effects of Discussion on Recall Accuracy and Agreement

While some previous research has suggested that group discussion may facilitate eyewitness accuracy, other research has drawn attention to the potential dangers of such discussion. The present studies examine this controversy and the interaction between accuracy and agreement following group discussion. It was found that individuals tend to agree in their testimony following group discussion, and are liable to change their original replies to bring themselves into agreement with group leaders' recall. A second experiment focused upon the actual items which witnesses agreed on following discussion. Changing the mode of questioning from interrogative to narrative eliminated differences in magnitude of agreement between discussion and no-discussion groups. As narrative recall leads to reporting of "easy" items, it was suggested that only certain "difficult" items are susceptible to discussion effects. The need for caution regarding the reliance on testimony following witness discussion was stressed.     

Criterion-referenced and norm-referenced assessment of minority group children

A variety of problems have been experienced with psychological assessment of minority children. Traditional norm-referenced measurement has repeatedly received criticism concerning cultural unfairness or bias. Responses to such accusations primarily have been in the form of new instrumentation aimed at attaining a culture fair assessment. Little response has been evident from a conceptual standpoint addressing the issues of purpose and use of test results Although many have turned to criterion-referenced measurement as an answer to the problems of norm-referenced evaluation, cultural bias is not necessarily avoided in this framework either. Issues of who determines criteria and what those criteria include must be addressed if criterion-referenced measurement is to meet adequately the challenge of multicultural evaluation.     

Between-chains and within-chain speed gradients as functions of length

Two experiments examined between-chains and within-chain speed gradients as functions of length. The first, involving straight runways, revealed hyperbolic gradients within chains for each of three lengths, .61, 1.22, and 2.44 m. Between-chains comparisons showed a goal gradient effect confined to the start box segment and a long-chain depression effect in the last two segments. The second experiment, involving discrete-trial fixed-ratio bar-pressing schedules of FR(1 + 4), FR(1 + 8), and FR(1 + 16), showed goal gradient effects for starting speeds and for speeds across the first segment beyond start. Also, confounding effects apparent in the runway, of acceleration, deceleration, and long-chain depression, were diminished or absent in the bar-pressing situation.