Unsophisticated raters judged speech naturalness and stuttering severity in audio-only and audiovisual presentations of speech samples. For stutterer samples, raters judged the audiovisual presentation more unnatural than the audio presentation, but for the nonstutterer samples, there was no difference. Stuttering severity ratings did not differ…

Comparison of self- and others' ratings of 335 managers' job performance found moderately low agreement between the two. Managers consistently reported practicing specific skills more frequently than raters' feedback showed, across 20 management practices areas. (SK)

Relatives of patients with dementia completed questionnaires about caregiver stress and patients' functional and psychiatric problems during initial evaluation, at 4-17-month follow-up, and retrospectively. Retrospective accounts correlated well with initial reports but psychiatric symptoms were underreported in retrospective accounts. Spouses…

This article describes the Scale for the Evaluation of Gifted Identification Instruments, developed for use by school decision makers. Development of the scale is reviewed in terms of five areas of assessment: validity, reliability, propriety, respondent appropriateness, and utility. Specific guidelines and cautions in using the scale are also…

Presents general principles for cleaning data sets. Suggests techniques: check for column shifts, data completeness, marginal frequencies, and selected cross-tabulations; examines a sample of cases. Gives strategies for preventing data processing errors. (SK)

Fifteen deaf and 11 hearing children (ages 8-10) witnessed slides depicting a wallet theft and were interviewed using a free recall approach followed by increasingly directive questions. Although accuracy of the two groups did not differ in free recall, deaf children provided less accurate responses to directive questions, whereas accuracy of the…

This article presents arguments that the correlation measure "r" of halo is not conceptually more appropriate than the standard deviation (SD) measure. It also describes conditions under which halo effects occur and when the SD and r measures can be used. Neither measure is uniformly superior to the other. (SLD)

A previous paper proposed a generalized family of coefficients of relational agreement for multiple judges. It focused on the concept of empirically meaningful relationships. This paper presents an ordinal coefficient of relational agreement as a special case of the generalized family. The proposed ordinal coefficient encompasses other ordinal…

Three studies investigating the existence of sex bias in the grading of undergraduate students, by examining interrater reliability for blind and non-blind grading, are reported. Negative evidence found in the results and the confusing picture presented by previous research indicate little firm evidence of sex bias in grading. (Author/MSE)

Reports findings from a study of the confidence expressed by ministry-level decision makers in five developing countries (i.e., Somalia, Botswana, Liberia, Yemen, and Nepal) about the quality of the national-level education data available to them and reasons for the perceived 16-40 percent error rate. (DMM)

Reviews commercially available instruments to assess students' study skills/habits/behavior, finding that none completely meets the standards set by the American Psychological Association or the American Educational Research Association. Offers guidance on selecting and applying these instruments in college settings. (DMM)

Outlines and discusses the nature and assumptions of the Willingness to Communicate scale. Discusses data relating to the reliability and validity of the instrument. Concludes that the scale is of sufficient quality to be recommended for research and screening purposes. (SR)

Analyses of the internal consistency of three autism scales--the Autism Behavior Checklist (ABC), the Real Life Rating Scale (RLRS), and the Childhood Autism Rating Scale (CARS)--were conducted with 34 children with pervasive developmental disabilities. Good internal consistency was found for the CARS. Adequate full-scale consistency was found for…

Coefficient alpha was examined through computer simulation as an estimate of test reliability under violation of two assumptions. Coefficient alpha underestimated reliability under violation of the assumption of essential tau-equivalence of subtest scores and overestimated it under violation of the assumption of uncorrelated subtest error scores.…

When using assessment instruments for personnel purposes, identify the purposes of assessment, specify criteria for excellent performance, determine behavior standards, and select appropriate methodology. Correlation analysis is a useful technique for establishing the validity of an instrument. (SK)