Pallas, A.M., Entwisle, D.R., Alexander,

Pallas, a.M., Entwisle, D.R., Alexander, K.L & Weinstein, P. (1990). Social structure and the development of self-esteem in young children. Social Psychology Quarterly 53(4).

Summarize the major objectives of the study and point out the main hypotheses that the author(s) list.

The authors investigate the structure and differentiation of self-esteem in young children, and trace the development of self-esteem from first to fourth grades. In a longitudinal study using the same population sample, the authors are able to trace the development and differentiation of self-esteem among young children. The authors predict that children will make increasingly finer distinctions in their self-concept as they grow older.

In other words, self-esteem related to body image would be further broken down into perceived weight or perceived height as the child matures. It is hypothesized, therefore, that children will differentiate among each dimension of self-esteem more as they develop. The authors also hypothesize differences among distinct population groups. In particular, the authors predict that boys and girls will score their self-esteem differently on different dimensions; that black vs. white students and privileged vs. low-income students will also reveal differences in their self-concept. The differences may or may not increase over time, but they will remain among the groups. The five broad categories of self-esteem or self-concept used in the research include character, personal responsibility, academics, athletics, and appearance.

Summarize the methods (subjects, materials, procedures, measures of behavior, etc.) used by the authors to test the hypotheses. Do you think that the methods used provide a good picture of the topic under investigation? For example, is the subject sample representative of the target population? Were the measures adequate or realistic?

The authors used a two-stage sampling scheme that was first based on the use of the Beginning School Study to glean a population of first-graders in the Baltimore area. Then, the sample was traced when the population was enrolled in second and fourth grade. The initial base group of participants was about 800 first graders, but by the end of the study in 1986, 553 children had participated in all self-esteem questionnaires. Because the authors deliberately sought a diverse sample, the population is reasonably representative with diversity in terms of gender, race, and socio-economic class. The authors refer to the research model as a "flexible confirmatory factor-analytic model," (p. 304).

Based on prior research, the authors employed a survey instrument consisting of 21 questions related to self-concept. The authors grouped the questions into five broad categories related to self-concept including character, personal responsibility, academics, athletics, and appearance. Factor loading and goodness-of-fit information is clarified in the methods section of the report.

The methods used do provide a good picture of the topic under investigation. The survey instrument is reasonably short to account for the short attention span of young children and their general inability to make abstract analyses of their self-image. However, there is no theoretical underpinning for choosing personal responsibility as one of the core self-concept groups. Children in first and even in fourth grade may not have a strong sense of personal responsibility.

What evidence (results or findings) is presented to support or refute the hypotheses of the study? How convincing is the evidence to you?

The authors found that factors were correlated on each dimension; if one aspect was rated highly then so were the others. Moreover, the authors did discover that correlations among these individual self-esteem factors did decline over time. This supported the hypothesis that "facets of the self become more distinct as children mature," (p. 307). The greatest gap among the dimensions was between first and second grade, perhaps indicating increased self-awareness and maturity in this year of social and psychological development. However, the authors do not refer to developmental psychology to illustrate their findings. The authors admit that the dimensions may not be distinguishable based solely on self-reports, and also that young children may not draw sharp distinctions among the dimensions.

Results disproved one of the research hypotheses: that between-group differences would be significant. In fact, only gender proved to be a significant variable affecting self-esteem. Blacks vs. whites and privileged vs. low-income students did not exhibit significant differences on most of the self-esteem dimensions. The most dramatic difference was between boys and girls on the athletics factor, with boys consistently reporting higher athletic self-esteem than girls throughout their development. The difference also increased with age. Girls rated themselves higher in personal character and personal responsibility than boys. The authors offer a groundless and nonsensical explanation for the findings, suggesting that "girls are given more responsibility for watching younger children than boys and are generally more conforming than boys," (p. 312). Girls also exhibited a more positive academic self-concept in the first grade than boys did, but that difference diminishes over time. Boys and girls also diverged with regards to body image, and that divergence increased over time. By the end of the fourth grade, girls had a significantly lower self-esteem related to body image than boys did. Race was less of a factor differentiating self-concept, and social class even less so. The researchers were surprised at the similarities. In general, the evidence supporting or disproving the hypothesis was convincing given the rigors of the methods used to gather and analyze the data. The authors admit the key problems with the research include that "traditional exploratory factor analyses do not produce coefficients that have standard errors, so there is no simple way to tell whether the loading for one group (girls) is statistically significant from the loading for another group (boys)," and also that "exploratory factor analysis models are underidentified" (p. 310). In other words, the research is simply exploratory in nature and no causality can be inferred.

In the discussion/conclusions section of the article, the author may explain or speculate about the meaning and importance of his/her findings. What explanations does the author provide for his/her results or conclusions? Do you agree with them (why or why not)?