The Development And Applied Research In The Self-Esteem Scale Of High-School Students

[Objective]The goal of this study is to develop the self-esteem scale of high-school students with good reliability and validity, and to analyze the characteristics of self-esteem in the high-school, and to confirm the mediation of self-esteem between social support and trait anxiety.[Methods]First of all, we read the relevant references in order to collect the discussion of self-esteem and construct theoretical framework. After referred to the interview results, we constructed the open-ended questionnaire. Secondly, we arranged and checked consequence after tested the questionnaire. The initial self-esteem scale was formed with existing scale. Then, we tested the initial scale and the data was submitted to project analysis, exploratory factor analysis by Spss13.0.The formal self-esteem scale was generated. Finally, we tested the formal scale and the data was submitted to reliability and validity testing, confirmatory factor analysis, differences testing in self-esteem and mediating effect of self-esteem testing.[Results](1)the self-esteem scale of high-school students was formed by literature, questioning, talking and statistical analysis. The scale was constitute of five factors,such as the self-esteem of study, the self-esteem of family, the self-esteem of conversing, the self-esteem of society and the self-esteem of body. Reliability tests showed::cronbach's of self-esteem scale was 0.899,the sub-scales'a were from 0.640 to 0.871.Validity tests showed:content validity was satisfying.. Relevance between self-esteem and social support confirmed an good criterion-related validity. The sub-scales'coefficient correlation were form 0.249 to 0.524,the coefficient correlation of the dimensions and scale were from 0.565 to 0.757, these confirmed an acceptable construct validity. Confirmatory factor analysis showed:χ2/df was from 0.2 to 0.5, GFI and AGFI were nearly equal tol, RMSEA was nearly equal to 0, RMR was less than 0.1.(2)There were significant differences in grades, gender, home address on the self-esteem scale of high-school students.The grades differences analysis showed:there were significant differences in grades on scale and sub-scale. Grade one were higher than grade two on the self-esteem of study, grade three were higher than grade one on the self-esteem of conversing and body and the total self-esteem.The gender differences analysis showed:female were higher than male on the self-esteem of conversing.There were no differences between singleton and non-singleton on the scale and all sub-scale.The home address differences analysis show:urban were higher than rural on the self-esteem of family and the total self-esteem.(3) To explore the mediating effect of self-esteem, there were a regression analysis. The results showed:we conducted a regression analysis between social support and trait anxiety and social support predicted trait anxiety negatively; we conducted a regression analysis between social support and trait the self-esteem and social support predicted the self-esteem actively; we conducted a regression analysis among social support, the self-esteem, trait anxiety.There wasn't the role of prediction between social support and trait anxiety, the self-esteem predicted trait anxiety negatively. Therefore, the self-esteem has the role of complete intermediary. [Conclusion](1)The self-esteem scale of high-school students was formed by the self-esteem of study, the self-esteem of family, the self-esteem of conversing, the self-esteem of society and the self-esteem of body. Reliability and validity tests showed:the total scale and subscale's a coefficients achieved a higher standard; content validity, criterion-related validity, construct validity were good. Confirmatory factor analysis showed the structure of self-esteem was good and it met psychometric. So the scale was valid measurement questionnaires.(2)The self-esteem of high school was a dynamic level of development. Due to academic pressure, relationships, physical appearance and other factors, there was no synchronization between self-esteem and grades; due to differences between men and women, interpersonal self-esteem of girls were higher than boys'; the number of children had little effect on the level of the self-esteem; the self-esteem of urban students was lower than rural students'.(3)There was no directly causal relationship between trait anxiety and social support. Their influence was produced by the self-self.(4)The disadvantage of the scale:promotion of the results were restricted because of time and condition; the practice of self-esteem had yet started in high school.