Robert S. Barcikowski

Statistical Power with Group Mean as the Unit of Analysis

In 1940 Lindquist presented a discussion of an analysis of variance using the group mean as the unit of analysis. Unfortunately, most modern writers on statistical methodology have ignored this important topic and some of the best text-book book writers make the error of using the individual as the unit of analysis when the group mean would be more appropriate. The reluctance to use group mean as the unit of analysis is due in part to the belief that the resultant drop in the number of observations per treatment greatly reduces the probability of detecting a treatment effect. This concern is put in perspective in this paper and equations are presented to facilitate power estimates when the group mean is the unit of analysis.

Decisions in Single Group Repeated Measures Analysis: Statistical Tests and Three Computer Packages

Abstract Repeated measures analysis is a widely used tool of applied statisticians. However, its users are confronted with what seems to be a myriad of decisions, even in its simple application. Should a univariate or a multivariate mode be used? Should an adjusted univariate test be used, and if so, which one? Should a pooled or partitioned error term be used? How important is the sphericity test? Given answers to these questions, which of three major statistical packages, and which programs within a package, could be used? These questions are discussed and resolved in this presentation.

Barriers to nurse/nursing aide communication: the search for collegiality in a southeast Ohio nursing home

AIM The aims of this study were: (1) to identify barriers to nurse/nursing aide communication and to discuss and provide solutions to some of these problems through a focus group; and (2) to determine the effectiveness of this focus group on job satisfaction-related outcomes. BACKGROUND Numerous studies have attributed problems with nurse staffing and turnover to faulty communication between nursing supervisors and nursing aides. METHODS Registered nurses (RNs) (n = 2), licensed practical nurses (LPNs) (n = 10) and nursing aides (n = 19) were interviewed; narrative themes from the interviews were then used to guide focus group discussions. A job satisfaction survey in a pre-test/post-test fashion was used to test the efficacy of the focus group. RESULTS Qualitative data from the interviews highlighted: (1) anger and condescension in communication; and (2) lack of mentoring, empathy and respect. The job satisfaction survey data showed that the attitudes decreased significantly among participants in the focus group but not in the control group (P < 0.05). CONCLUSION Nurse supervisors, both RN and LPN, are key to educational reform and must be taught collegial methods when delegating authority. IMPLICATIONS FOR NURSING MANAGEMENT Findings from this study can be used to develop a multidisciplinary educational tool to foster communication and collegiality, thereby reducing job turnover in nursing homes.

LD Consultant Information in Mainstreaming: Help Or Hindrance?

The purpose of this study was to assess the impact of information commonly disseminated by consultants on teacher expectations for mainstreamed LD students. After receiving varying combinations of positive and negative communications in the areas of reading achievement, classroom behavior and learning style, subjects completed the Regular Educator Expectancy Scale-Revised — an instrument developed to measure four areas of mainstreaming (Instruction, Effects on Others, Effects on the Child, and Behavior Management Strategies). Results indicated that teachers were differentially influenced according to the type of information received. Implications for the LD consultant are discussed.

Predictors of teachers perceived self‐competence in classroom management

The purpose of this exploratory study was to identify predictors of teacher manageability beliefs (the degree to which teachers believe they can personally manage a behavior). One hundred and eighty-two teachers completed the 39-item, nine-subtest Teacher Manageability Scale (TMS), which asked participants to rate each behavior (on a 5-point scale) by responding to the question “How easily can you manage the behavior when any student…?” Results of a multivariate regression analysis evaluating the relationship between the nine subtests (dependent variables) and eleven teacher-reported independent variables indicated that between 13% and 18% of the variance was predicted for the six TMS subtests achieving significance. The variables related to the greatest number of dependent measures were student academic achievement and socioeconomic status. Surprisingly, principals disciplinary support and class size (enrollment) failed to predict even a single subtest. Implications for professional educators working with students identified for special educational services and future research are discussed.

Calculating the Statistical Power of the Univariate and the Multivariate Repeated Measures Analyses of Variance for the Single Group Case Under Various Conditions

Repeated measures designs find frequent application in behavioral research. Researchers using such designs must choose between the univariate approach and the multivariate approach for a test statistic. The statistical power of these two approaches has been shown to be varyingly different depending upon the structure of the variance-covariance matrix relative to the pattern of mean differences. However, the guidelines for choosing the most powerful test for a given situation have not been well defined. A FORTRAN program is described which calculates the statistical power of the omnibus tests for the univariate and the multivariate approaches to repeated measures in the single group case. Program input, output, applications, and availability are discussed.

The Factorial Validity of the Piers-Harris Children's Self Concept Scale for a Sample of Intermediate-Level EMR Students Enrolled in Elementary School.

Several factor analytic studies have obtained encouraging but inconclusive evidence for the construct validity of the Piers-Harris Childrens Self-Concept Scale. In this study scores on the Piers-Harris scale from a sample of elementary-school-age, educable mentally retarded subjects were factor analyzed. Partial support for the construct validity of the Piers-Harris scale was obtained by replicating Piers-Harris factors bearing on physical appearance and behavior. Some support was also found for Piers-Harris factors reflecting popularity, academic ability, and anxiety. The factor analytic results are compared with those from other studies. Suggestions for future research include more clearly defining emotionally toned factors, broadening the response structure of the Piers-Harris scale, considering maturational effects in attempting to replicate factors across age periods, and searching for a general self-concept factor.

Optimum Sample Size and Number of Levels in a One-Way Random Effects Analysis of Variance.

In most behavioral science research very little attention is ever given to the probability of committing a Type II error, i.e., the probability of failing to reject a false null hypothesis. Recent publications by Cohen (2, 3) have led to a great deal of insight on this topic for the fixed-effects analysis of variance and covariance. These publications should prove invaluable in aiding social scientists to carry out and better understand their research. It is the purpose of this paper to provide social scientists with some insight in dealing with Type II error, and therefore, optimum sample size and number of levels, in the random-effects analysis of variance.

The Effects of Item Discrimination on the Standard Errors of Estimate Associated with Item-Examinee Sampling Procedures

A Monte Carlo study was conducted using item-examinee sampling procedures to examine the standard error of estimate for a given tests mean and variance. The main variables considered were test length, item difficulty, and item discrimination. The results indicate that optimal estimates, i.e., smallest standard error, of both mean and variance from a single item-examinee sampling plan may not be possible.

A Computer Program for Selecting Optimum Sample Size and Number of Levels in a one Way Random Effects Analysis of Variance

effects, but in the variance of the population from which these effects were randomly selected. For example, suppose that a population of school teachers is available to teach reading to first grade pupils using a certain method. It is decided that the method will be adopted for use provided its success is not heavily dependent on the personalities of individual teachers. A random sample of teachers is drawn and randomly assigned to a random sample of classes of first grade pupils. The dependent variable (a standardized reading test) is selected and a one way analysis of variance, random effects model, is to be used to test the hypothesis that there are no significant differences among the teachers. Researchers faced with the preceding problem are often ignorant of how many teachers to select and the number of students to as-