Members of adult Sunday School classes from seven churches in Bowling Green, Kentucky, were administered Rotter's Internal-External Locus of Control Scale. The seven churches were also ranked by ministers on a continuum of doctrinal closedness-openness. The hypothesis stated that there would be a difference among churches according to mean internal-external control scores. It was also hypothesized that the members of the more doctrinally closed churches would tend to score as more externally controlled. Analysis of covariance indicated that the churches did differ significantly on the internal-external control scale but the doctrinally closed churches tended to be more internal than the doctrinally open churches.

Volunteer college students participated either as members of two experimental experiential-Gestalt growth groups or as members of a control group. The experimental groups met for one two hour period per week for eight weeks. It was hypothesized that the experimental group members would change significantly in the direction of increased self-actualization as measured by the Personal Orientation Inventory (POI) while the nontreatment control group members would not. A comparison of pre- and posttest scores for the groups indicated no significant effects of the group process. Increases occurred over the nine week period for both experimental and control subjects. The importance of proper control groups is stressed. Implications for further research indicate a need to identify factors which cause control group members to change in the direction of increased self-actualization without participation in an intensive group process.

Change in self-actualizing, assessed by the Personal Orientation Dimensions (POD), was examined for two sections of a sophomore level psychology class. The subjects were administered the POD at the beginning and the end of the semester. No significant differences were found in the amount of change in self-actualizing between the two groups. It was suggested that college may be a growth experience and that students at the sophomore level will not engage in growth exercises if they feel it is not related to their grades. It was further suggested that more research is needed on the POD and of ways to combine affective with intellectual components in educational settings.

Twelve food-deprived male rats were trained to barpress for food pellets in one of three operant chambers. The chambers were of standard size, double in length, or triple in length. After training, cod cup full of pellets was placed in the corner opposite the operant bar, and the eating behavior of the rats in this choice situation was observed. The dependent measure was the percentage of the total amount of food consumed that had been obtained by barpressing. Individual comparisons between the mean percentage of food earned over the four test sessions revealed that those rats in the longest chamber barpressed for a significantly greater amount of the food consumed than did those in the shorter boxes. Results are discussed in terms of the discriminability of the two food cups and the large individual differences. An approach for future research to follow is suggested.

The effects of intervention programs on the self concept of rural preschool children were examined using 86 children, ages 5 to 6 1/2 years, who had and had not had preschool experience. The relationships between the Total Self Concept and the subscales of Body Image, Competence, and Social Interactions to the variables of socio-economic level, sex, and preschool experience were determined. Significant differences (p <.001) were found between the different population variables and the Total Self Concept score, and the subscale scores of Body Image, Competence, and Social Interactions. The locations of these differences were determined by a Newman Kuel analysis. Significant effects (p<.05) on sex differences were found on Total Self Concept scale and on the Competence subscale. There were no significant (p <.05) interaction effects. Implications of these findings were discussed.

The relationship between the drinking patterns of college students and personality factors, which have previously been shown to different:late between alcoholic and non-alcoholic populations, was examined in this study. Subjects were categorized into six groups based on self-reports of the frequency and amount of alcohol consumption. The MacAndrew Alcoholism Scale and the 16PF personality test were administered to all subjects. Pearson Product-moment correlations were calculated to assess the relationship between levels of drinking and personality factors demonstrated by the subject population. It was found that the MacAndrew Alcoholism Scale score was significantly correlated with drinking level among young adults. This suggests that the MacAndrew Scale is potentially useful as a predictor of alcohol abuse and the findings are consistent with previous research in the area. Seven Personality factors, measured by the 16PF, were found to be significantly related to the subjects' drinking levels. However, These factors were inconsistent with factors which have previously been shown to differentiate alcoholics from normal populations. The personality characteristics, which the young-adult heavy drinker demonstrated on the 16PF, appear more closely related to the antisocial personality type than to the neurotic personality type, as previous research had indicated. Further research is needed, especially longtitudinal studies, to assess fully the effects of specific personality characteristics on the etiology of alcoholism.

Previous research concerning the relationship between locus of control and assertiveness has suggested that internals are more assertive than externals, hut the findings of earlier studies are inconsistent and inconclusive. The present study focused on the social desirability needs of subjects as related to locus of control and assertiveness. It was predicted that individuals who scored in the internal direction on the I-E scale and had a low social desirability need would receive higher scores on a test of assertive behavior than individuals who scored in the internal direction and had a high social desirability need. The Marlowe- Crown Social Desirability scale, Rotter's Internal-External scale, and the College Self Expression scale were administered to 69 male and 111 female college undergraduates. The results of the multiple regression analyses indicated that social desirability did not contribute significantly toward the ability of locus of control to predict assertiveness. Other findings indicated that males as a group were more assertive and more internal than females. However, for females only internals were more assertive than externals. It was suggested that perhaps what was being measured was expected male and female sex roles.

This thesis investigates the sociological determinants of the degree of satisfaction with undergraduate education expressed by alumni of Western Kentucky University. Subjects were selected by means of a random sample (h = 1759) of graduates from the years 1922 through 1962. The effects of two major types of independent variables upon alumni satisfaction are explored using Multiple Classification Analysis. The first group of explanatory variables are background factors which include father's educational level, parental income while the subject was attending college, sex, average college grade, and whether or not the subject's present occupation is in the same field as the college major. The second group of predictors are life-experience variables which include the subject's current occupation, income, marital status, number of children, community of residence, and age. The findings indicate that the subject's age is the most powerful predictor of the degree of satisfaction expressed toward Western Kentucky University. The subject's present occupation and income were the next most influential explanatory variables. Results support the major conclusion that events in the lives of alumni subsequent to graduation from college are more important in explaining their satisfaction with undergraduate education than are their own social backgrounds or events occurring while the subject is attending college.

Data were obtained from 186 applicants at a large Midwestern bank to test the hypothesized advantage of employing a continuous as opposed to a dichotomous criterion when predicting consumer installment credit risk. A continuous criterion of cost was derived and regressed on 21 background items obtained from the application blank (e.g., number of dependents, total monthly income, present indebtedness, etc.). The maximum step of the regression analysis yielded a multiple squared correlation coefficient of .24. This data was then subjected to multiple discriminant analysis employing a median split criterion which yielded an eta squared correlation coefficient of .13. As can be inferred from these results, a greater degree of predictive accuracy may be attained when predicting the continuous variable of risk. Further indicative of the hypothesized advantage of employing a continuous opposed to a dichotomized criterion were the results of the bivariate correlational analyses. Nine background items correlated significantly with the continuous criterion while only five correlated significantly with the dichotomized criterion.

The purpose of this study was to investigate the validity of a 25-item abbreviated version of the Mathematics Anxiety Rating Scale (Richardson & Suinn, 1972). Convergent and discriminant properties of the Mathematics Anxiety Rating Scale -Abbreviated (Alexander, Cobb, & Martray, 1986), as well as its sensitivity to individual differences were examined. Convergent validity was examined by studying the correlation between the two math anxiety scales, the MARS -A and the Fennema-Sherman Math Anxiety scale (FSMA). Discriminant validity was examined by studying the correlations between: (a) the MARS-A and a test anxiety scale (TAI), and (b) between the MARS -A and a general trait anxiety scale (STAI). Stepwise multiple regression analysis and Pearson Product -Moment correlation coefficients were used to investigate sensitivity to individual differences. The MARS-A functioned as the criterion variable. The predictor variables were American College Test (ACT) math scores, math coursework grade (Grade), confidence towards learning mathematics as measured by the Fennema-Sherman Confidence Scale (FSC), race (Race), sex (Sex), and age (Age). Convergent validity was demonstrated by a relatively high correlation between scores yielded by the two measures of math anxiety, viz., the MARS-A and FSMA (r = -.61). Evidence of discriminant validity was demonstrated by positive but compared to the convergent validity coefficient, lower correlations between the MARS -A and TAI (r = .44), and yet still a lower correlation between the MARS -A and the STAI (r = .31). The positive correlations among these instruments indicated that instruments of general trait anxiety, test anxiety, and mathematics anxiety measure various aspects of anxiety; however, the order in which the correlations rank suggest that these instruments do not all measure the same trait. The degree of correlation among scores supports the idea that as the instrument becomes more item specific (from measuring general trait anxiety to test anxiety to math anxiety), the correlation between scores obtained on the instruments become stronger in magnitude. FSC was the only predictor variable to enter the stepwise multiple regression prediction equation. The relationship indicates that confidence towards learning mathematics is the single, best predictor of scores obtained on the MARS -A. The negative direction of the correlation between FSC and MARS-A scores suggests that the more positive one's confidence is towards learning math, the lower one's math anxiety level. Although the MARS -A correlated significantly with other predictor variables, viz., ACT, Grade, and Age the shared variance between FSC, ACT, and Grade, and between ACT and Age imply that once the contribution of FSC to MARS-A scores was accounted for, remaining variables made no unique contribution.