Abstract:
The objectives of this study were to (1) compare wellness scores of youths at a juvenile training center, with different levels of self-control, and were assigned to either logotherapy group or the training center's regular activities group. (2) validate the development of wellness as a result from logotherapy group by using latent growth curve Model structure equation analysis, and (3) examine the Measurement invariance of model of wellness scores of youths who were assigned to logotherapy group and the training center's regular activities group. A quasi-experimental design was applied. A total of 180 juvenile offenders and substance abuse at the Juvenile Training Center for Children and Youth, Zone 8; Surat Thani province, and Zone 9; Songkhla province, were selected in this study. Ninety participants were in a experimental (logotherapy) group and the others was control (regular activities) group. Research instruments used were; (1) Logotherpy Group Program, (2) Wellness measures, (3) Self-Control measures. Data were analyzed by descriptive and inferential statistics i.e., standard deviations, coefficients of variation, skewness & kurtosis values, repeated measure MANCOVA, Pearson correlation analysis, Exploratory Factor Analysis using SPSS. Moreover, Confirmatory Factor Analysis, Latent Growth Curve Model Structure Equation Analysis, and Multiple Group Structural Equation Model Analysis by LISREL were used to analyze the data. The findings revealed that (1) The average of wellness status measured for four times between two groups were a significant differencely at .05. The average of wellness status measured at 2, 3, and 4 in the groups with different level in self- were not significance different at .05. (2) Latent growth curve model of wellness in both experiment and control groups was valid and well fitted to empirical data = 1.84, df = 3, / df = 0.613, p-value = 0.61 GFI = 1.00, CFI = 1.00, NFI = 1.00, RMSEA = 0.0, and SRMR = 0.022 (3) Latent growth curve model of wellness in both experiment and control groups was not invariance but there were differences in mean scores of wellness in each measurement and the variance of mean scores between 4 times of measurement. In the other words, the multiple group model was not significantly different, but the weight of each component in the measurement was different.