Dennis E. Hinkle

How Large Should the Sample Be? A Question with No Simple Answer? Or...

A recurring problem in behavioral science research is that of determining the appropriate sample size before conducting a research investigation. The more common approaches to this problem have been to apply indiscriminantly the formulae used in survey research without considering the factors necessary for using these formulae. The most important of these factors is the Effect Size. In this paper, tables for the appropriate sample sizes are presented and discussed in the context that the determination of the Effect Size must precede the determination of the sample size.

How Large Should the Sample Be? Part II—The One-Sample Case for Survey Research

In a previous article, the authors discuss the importance of the effect size and the Type II error as factors in determining the sample size (Hinkle and Oliver, 1983). Tables were developed and presented for one-factor designs with k levels (2 ≤ k ≤ 8). However, between the time the article was submitted and its publication, the authors presented these tables at several national and regional meetings. A recurring question from colleagues attending these meetings was how these tables could be used for the one-sample case (k = 1). Since they could not be, we were encouraged to develop comparable tables for the one-sample case. Thus, the purpose of this paper is to readdress the sample size question and to present these tables.

Adlerian parent education: Changes in parents' attitudes and behaviors, and children's self-esteem

Abstract The Adlerian approach to family education is receiving increasing attention. The Adlerian philosophy reflects a systematic approach with special emphasis upon the family constellation. The family is viewed as a total unit, and the individual needs of the family members are seen as family group needs with each meinber attempting to establish his/her place within the family structure. The purpose of the present study was twofold: first, to determine the reliability and validity of three measurement scales appropriate for assessing the effects of family education programs, and second, to determine the effects of a family education program using the Adlerian approach. The data indicated high reliability for the scales and the beneficial aspects of the family education program.

Children's thinking about family characteristics and parent attributes

Abstract Children (N = 61; 31 boys. 30 girls) from prcschools, first grades, and fourth grades were interviewed on their thinking about family and parental roles. Older children were more likely than younger children to recognize the multiple roles of parents and the multiple functions of a family, as well as the interrelationships between these factors. Qualitative analyses of the childrens responses clarify some of the assumed developmental differences in childrens thinking on family and parental concepts.

A Developmental Study of Ethnic Self-Concept among Preadolescent Girls

Summary The present study explored the developmental changes in the self-concepts of 634 nine-, 432 ten-, and 452 eleven-year-old girls from Euro-, Afro-, and Mexican-American backgrounds, respectively. The lower SES group had a significantly lower mean self-concept than the middle SES group. The Euros had a higher self-concept than the Afros, and both of these groups had significantly higher self-concepts than the Mexican-Americans. Also, 11-year-olds had a significantly higher mean than both the nine- and 10-year-olds. Furthermore, ANOVA showed that the two-way interaction between race and SES was not the same for the three age groups. The pattern of interaction was similar for the nine- and the 11-year-olds but different for the 10-year-olds. Self-concept scores increasing with age may be a function of development, cognitive maturity, and the mastery of social skills as well as ethnic membership, sociocultural environment, and the accessibility of opportunities.

Sexual Behavior and Guilt among Women: A Cross-Generational Comparison.

This paper reports the results of a study of the relationship between guilt and sexual behavior among 45 married women. The age breakdown of the subjects included 60 and above, 40 to 50, and 20 to 30, with 15 women included in each age category. A negative relationship was found between the age of subjects and the extent of sexual behavior while a positive relationship existed between age and sexual guilt. The younger womens sexual behavior scores were higher than both the middle and older (statistically significant) women. Scores indicated less guilt associated with sexual behavior for the younger women than for the middle-aged and older (statistically significant) women.

Ethnic Self-Concept during Middle Childhood.

Summary The present study investigated the self-concepts of 432 Euro-, Afro-, and Mexican-American 10-year-old girls. The results indicated that both ethnic and socioeconomic backgrounds exert varying degrees of influence on self-concept ratings. The middle-income girls scored significantly higher than the lower-income ones. The middle-income Euros had significantly higher self-concept ratings than their Mexican-American counterparts. No significant difference was found between the ratings of the lower-income girls from the three ethnic groups. The results were discussed in terms of cognitive development, reference group theory, ethnic boundary maintenance, and the ethnic identity movements.

Selection of models in contingency tables: A reexamination

Qualitative data often need analysis using log-linear models, as pointed out by Liu and Sanders. In this article, references to the state-of-the-art literature and software packages for analyzing such data are presented. We also illustrate the two types of log-linear models and present procedures for investigating both types.

Bayesian M-Group Regression Applied in a Community College Developmental Studies Program.

The purpose of the study was to compare classical multiple regression with Bayesian m‐group regression, complete with cross‐validation of both methods. The context of the study was to predict first‐quarter grade point average in various curricula in a comprehensive community college for those students who had completed developmental studies programs. The major findings were: (1) informal counselor prediction was the largest contributor to the final regression models, and (2) no difference was found between classical multiple regression and Bayesian m‐group regression.

Do They Differentiate Between Accredited and Nonaccredited Institutions

Voluntary institutional accreditation in public schools and higher education is unique to the United States. An institution granted accreditation has demonstrated that appropriate standards and criteria have been met, thus assuring the public as well as various governmental agencies of its quality and integrity. While it is often euphemistically referred to as voluntary, accreditation by various agencies is essential for institutional survival. A nonaccredited or disaccredited institution potentially experiences either a short life or a quick death (Stinnett, 1971). Accreditation by both the regional and professional associations has been under intense scrutiny during the past decades. Of particular note is the accreditation of teacher education programs by the National Council for Accreditation of Teacher Education (NCATE). From its stormy beginnings in 1952, NCATE has been subjected to a continuous stream of criticism regarding its effectiveness in insuring quality of accredited programs and the products of those programs. There was a brief period of relative calm following NCATE’s recognition by the National Commission on Accrediting (NCA), but it was short-lived.