The author contends that model misspecification can occur even after researchers have selected the generally most appropriate class of methods, or general linear model techniques. It is suggested specifically that canonical correlation analysis may provide more meaningful results, as compared with regression, particularly if analysis is augmented…

Part and bi-partial canonical correlations were developed by extending the definitions of part and bi-partial correlation to sets of variates. These coefficients may be used to help researchers explore relationships which exist among several sets of normally distributed variates. (Author)

Clarifies the nature of communication avoidance through a structural analysis grounded in facet theory. Presents evidence for a duplex model of avoidance in which theoretical distinctions among modalities of approach-avoidance and context proved empirically relevant. Discusses implications of these findings for the explication, treatment, and…

Describes the strengths and weaknesses of a new type of multivariate technique, covariance structure analysis (LISREL). Provides a detailed example which shows how the use of covariance structure analysis can improve research sophistication and theory development in counseling psychology. (Author/ABB)

Provides an expository presentation of multivariate analysis (MANOVA). Addresses several questions relating to why one should use MANOVA, its structure, how MANOVA test statistics and follow-up tests are obtained and interpreted, presentation of results, and alternatives to MANOVA. Uses an example data set to illustrate these points. (Author/KS)

Computer sampling from a multivariate normal spherical population was used to evaluate the type one error rates for a test of sphericity based on the distribution of the determinant of the sample correlation matrix. (Author/LMO)

The impact of test anxiety on test performance and the cognitive appraisals of test-anxious students were explored. The pattern of students' anxiety and appraisals suggests that test-anxious students experience most doubt and concern in the term. (Author/LMO)

A needs assessment was conducted to identify the health needs of school district employees in planning a comprehensive employee health service. Survey questionnaires were analyzed to determine characteristics of those individuals who may be at greatest health risk. Results indicate that degree of stress can differentiate health risk predictors.…

Describes the appropriate control of experiment-wise error rates and the researcher's role in making decisions to control for Type 1 errors. Issues related to level of significance, one- versus two-tailed tests, and the use of multivariate statistics are included. (Author/CT)

Multiple regression analysis is frequently being employed in experimental and non-experimental research. However, when data include predictor variables that are correlated, some regression results can become difficult to interpret. This paper presents a study to provide a demonstration that structure coefficients may be useful in these cases.…

The validity of the Lollipop Test: A Diagnostic Screening Test of School Readiness was examined using the Metropolitan Readiness Test (MRT), Level I, Form Q, as the criterion. Appreciable concurrent validity was found across test batteries. Implications for school readiness screening are discussed. (Author/BS)

A simplified example of a technique, conceived by Yates and later developed by Goodman and Haberman, for the determination of appropriate log-models in the measurement of qualitative data in higher education is presented. (Author/MLW)

Three significant canonical correlations were obtained from the analysis of the Wechsler Intelligence Scale for Children-Revised (WISC-R) and the Woodcock-Johnson Tests of Cognitive Ability (WJTCA) to examine test overlap with a sample of suspected learning disabled children. Results indicate a substantial portion of the variance is independent.…

To develop normative data for the revision of the Children's Manifest Anxiety Scale, the scale was administered to 4,972 children, ages 6 to 19 years, representing all geographic regions of the United States. Reliability was satisfactory for all groups except Black females below age 12 years. (Author/BW)