Search

This study was designed to derive the distribution of a test statistic based on normal probability plots. The first purpose was to provide an empirical derivation of the critical values for the Line Test (LT) with an extensive computer simulation. The goal was to develop a test that is sensitive to a wide range of alternative distributions, applicable to a wide range of sample sizes, and easy to compute. The second purpose was to determine the power of LT and compare it to that of several other test statistics. Monte Carlo simulation was used to generate the critical values of LT by randomly generating 500,000 replications from a normal distribution for sample sizes 10(1)100(25)1000(250)5000. For each replication, the value of LT was calculated. The empirical critical values were determined for each of three levels of significance for each sample size. These critical values were then "smoothed" using nonlinear regression techniques. The results indicate that LT provides adequate control over Type I errors while at the same time providing statistical power comparable to the Shapiro-Wilk test. Results also indicate that LT is easy to compute, is powerful for detecting departures from normality under a wide variety of alternative distributions, and is available for sample sizes up to 5,000. An appendix contains the Statistical Package for the Social Sciences syntax for computing the LT for a given data set. (Contains 2 figures, 10 tables, and 45 references.) (Author/SLD)

The assumption of normality underlies much of the standard statistical methodology. Knowing how to determine whether a sample of measurements is from a normally distributed population is crucial both in the development of statistical theory and in practice. W. Ware and J. Ferron have developed a new test statistic, modeled after the K-squared test of R. D'Agostino and E. Pearson (1973), the g-squared test statistic. This statistic has been used to estimate critical values for sample sizes up to 100, but a more extensive derivation and validation of the critical values are required, and the power of g-squared against a wide range of alternative distributions requires study. Monte Carlo simulations were performed to investigate these areas. The main advantage of g-squared is its conceptual and computational simplicity. The power study shows that g-squared is sensitive to a wide range of alternative distributions, especially peaked distributions, having absolute power for many distributions with a large "n." G-squared could be valuable for testing univariate normality in statistical routines, but it does have some weaknesses. One of its main disadvantages is its low power with small sample sizes except for peaked distributions. While g- squared can tell you about a departure from normality, it can not tell if the departure is due to a single outlier. It is recommended that when testing for departures from normality, g-squared should be used as a supplemental quantitative measure of normality to the information obtained from histograms, box plots, stem and leaf diagrams, and normality plots. (Contains 7 figures, 12 tables, and 28 references.) (SLD)

Major elements of the Florida Parent Education Follow Through Program are: (1) training mothers (two to each classroom) in the role of combined parent educator and teacher auxiliary; (2) training the teacher to use paraprofessional personnel; (3) developing materials for family use which take into account not only the school's goals for the child, but also, and equally, the family's expectations, goals, life style, and value system; and, (4) involving the Policy Advisory Committee in all phases of the program. In 11 communities, intervention contact focused on parents but also involved teachers, students, parent educators, community program coordinators, and consultants. Community personnel participated in five workshops which encouraged each community's independence, responsibility for self-evaluation, and the development of effective use of the program advisory committee (PAC). Inservice program support was supplied by liaison officers and consultants from the University of Florida. Inservice training materials were developed, and 422 home learning tasks were performed in two Alachua County, Florida, elementary schools for use in the 11 communities. Pretest and posttest data was collected using a variety of test and report measures; results appeared to indicate that basic objectives were met for this year. Data results on parents, children, teachers, parent educators, home visits, and home learning activities, are contained in the appendixes. (SDH)

This report of the 1974 Florida Parent Education Follow Through Program contains evaluation data pertaining to parents, children, teachers, and parent educators from 11 communities. The program emphasis is on the development of: (1) nonprofessionals as parent educators, (2) materials for family use, and (3) parents as partners in the educational program of their children. Part 1 contains a rationale and an outline of the key elements of the program; Part 2 lists the program goals for parents, children, and for classroom and school. Part 3 describes program implementation procedures (including on-site training workshops and consultant visits), the Policy Advisory Committee activities, and evaluation of data concerning parents, home environment, home visits, the children, teachers, parent educators, and home learning activities. Achievement results of children in each community in the program are presented separately. Appendixes, which make up more than one-half of the document, contain detailed charts on the evaluation results and a variety of forms and materials used in conjunction with the program. (CS)

The power of randomization tests was systematically examined through simulation for typical designs that rely on the random assignment of interventions within the observation sequence. A 30-observation AB design, 32-observation AB design, and multiple baseline AB (15 observations on 4 individuals) were studied. Power estimates were generally found to be low. (SLD)

Finds a small, but statistically significant, association between exposure to U.S. entertainment programs and attitudes, perceptions, and behaviors of foreign audiences. Finds also that, when taking study characteristics into consideration, only language of the questionnaire produced a significant difference in correlation size; and that the magnitude of the relationship depends upon the type of dependent measure. (RS)

This paper extends the factor analytic research of C. W. Swartz and others (1990) on the North Carolina Teaching Performance Appraisal Instrument (NCTPAI), which contains 28 teaching practices grouped into 5 functions (management of instructional time, management of student behavior, instructional presentation, instructional monitoring, and instructional feedback). Data were collected on 305 teachers in a large city/county school system, using full-time professional teacher observers/raters. The professionally trained observers/raters rated each of the 28 practices on a 6-point scale, visited the teachers' classrooms, and made extensive notes relating to the 28 practices. One item was observed in less than half of the classrooms and was deleted, leaving a pool of 27 items. After deleting cases with missing values on the remaining 27 items, the data on 247 teachers were used. A confirmatory factor analysis, using the two factor (instructional presentation and management of student behavior) oblique solution of Swartz and others, was completed. Results were compared to those of the earlier study. Additional exploratory factor analyses were conducted to find a more appropriate model. The two factor oblique solution did not provide an adequate fit for the data. Analyses of the chi-squared test for the model and the fit indices suggest an inadequate model. The original factors of instructional monitoring and instructional feedback may be only one factor, with the other three original factors being maintained. Three tables are included. (RLC)