2003, Number 4
Salud Mental 2003; 26 (4)
Confiabilidad y validación con niños mexicanos de dos instrumentos que miden la autoestima
Giménez FC, Cortés MM, Loaeza VP
PDF size: 377.48 Kb.
ABSTRACTThe concept of self-esteem was first envisaged through the ideas on self elaborated by William James. From his observations, a great deal of conceptual work has been built to discriminate it from other more ambiguous constructs.
Most part of these concepts can be assessed from two different perspectives. The first one relates to auto-cognition and includes terms such as self-concept, self-image and self-comprehension, all of which suggest a descriptive reference towards the self. The second, which is a preferred psycho-dynamic approach, is linked to self-affection and related concepts, i.e. self-value, selfevaluation and self-esteem. These concepts are usually employed to describe how much worth a person believes to have or how valuable believes to be in relation to his or her peers.
As a result of different contemporary suggestions, self-esteem has been defined as the negative or positive attitude each individual develops towards himself or herself, the degree of satisfaction with oneself, and the personal feeling of worthiness that each of us holds as compared to others. An important issue in the construction of self and thus, of self-esteem is the object relationship between mother and child established in early life by an infant. Thus, some theorists point out that the child uses what he/she already knows, and feels, in order to build his/her reality and to contribute emotionally to his/her own experience.
Following the critical definition of self-esteem offered by Jude Cassidy (9), and due to the lack of appropriate instruments in Spanish-speaking societies, we decided to investigate the degree of self-esteem in a population of boys and girls from Mexico City ages 8 through 12 by using two comparable indices that require open, calibrated, non-guided questionnaires in order to evaluate self-esteem.
Our study focused on the evaluation of the psychometric properties of two scales, intended to measure self-esteem reliably in a sample of Mexican school children. The first scale is the Self Esteem Index (SEI) created by Brown & Alexander (1991), on which we worked to accomplish its adaptation (back-translation), validation and reliability before applying it to Mexican children. This SEI scale is an 80 item, norm-referenced, self-report instrument designed to elicit children’s perceptions about their personal traits and characteristics. The respondent reads each item on the SEI, and uses a modified Likert-type scale to classify each item based on four possible answers: always, almost always, almost never, never. Since most of the instruments generally used to measure self-esteem discriminate between factors, this fouritem scale includes familiar acceptance, academic competence, peer popularity and personal security as relevant factors in assessing self-esteem in school-children.
The SEI was administered to an unselected group of 419 Mexican children of both sexes, aged 8 to 12 years-old, who attend private elementary schools in Mexico City. Prior to its application, the instrument was subject to English-Spanish translation, back-translation (Spanish-English) with peerreviewing. During this process, several idiosyncratic adaptations were performed in order to simplify some reagents. The SEI was administered to a pilot sample of school-children (ages 8 to 12) for final syntactic corrections.
The research protocol was presented to two private institutions in Mexico City for approval and both instruments were applied in parallel, in groups comprising twenty to twenty-five schoolchildren, ages 8 through 12, for a total of seventeen elementary school groups. Amongst these, there were four third-grade groups, four fourth-grade, five fifth-grade and four six-grade groups in all. The application of each instrument took about 40 minutes.
After giving out the questionnaire, instructions were read aloud, stressing the fact that no good or bad responses were intended, and also the need to be sincere in every answer. Both instruments were applied with a difference of three weeks between them, and then evaluated simultaneously with a psychometric analysis in order to obtain their validity and reliability for this particular population and age-group.
Statistical analysis of all data obtained was made according to the SPSS package for Windows (version 10.0). Comparative assessments were made between factors, among every reagent and between instruments, applying Cronbach’s alpha and Student’s t test. Also, the items were submitted for factorial analysis using the Varimax rotation for construct validity (16). Finally, the concurrent validity between instruments was obtained by means of Spearman’s correlation, with a significance level of .0001. Our results show that, for the SEI instrument, we obtained an alpha coefficient of 0.8586 which confirmed a good reliability. This was additionally supported in the factorial analysis with Varimax rotation at .30, suggesting that despite our previous assessment, nine items need to be reviewed for further applications in our translated version. Moreover, we submitted all items to a discriminatory validation with a Student’s t value of .000 as cutoff point and showed that all the items were satisfactorily discriminated
In a concurrent process, we elaborated a new instrument to measure self-esteem which we called Instrumento de Autoestima Infantil (IDAI). This novel index consists of 80 items, distributed in twenty-long item scales, corresponding to four distinct factors: familiar acceptance, academic competence, popularity and personal security. The respondent reads each item and uses a Likerttype scale to respond under five options, i.e. never, almost never, sometimes, almost always, always. From a bank of reagents elaborated by the authors, we selected 80 which would readily conform to the above-mentioned scales. Before its administration, the instrument was submitted to peers who provided several syntactic suggestions that improved the quality and clarity of the reagents.
The IDAI was administered to the same sample of 419 Mexican school-children. In this case, we obtained an aplpha coefficient of 0.9168, which confirms a good reliability, as well as in the factorial analysis with Varimax rotation at .30, by eliminating only two items (item numbers 20 and 60). The other items were thus clustered into four new scales that, accordingly, we named social acceptance, academic acceptance, personal assurance and academic competence.
The discriminatory validity of each item was obtained by means of a Student’s t value with a high degree of significance (.000) that reflects a satisfactory discrimination for all items tested. Correlation between the Self Esteem Index and the Instrumento de Autoestima Infantil with a Spearman standard was statistically significant (p value ‹ .0001) with a score of r= 0.71.
These results are encouraging as they suggest that both instruments can be reliably administered to Spanish-speaking children ages 8 to 12 years-old to assess their self-esteem, for purposes of improving their academic performance and facilitate timely psychological intervention.
More research studies in different Hispanic populations are needed to confirm our findings, but we hope that this work will foster further investigation in the field of developmental psychology to better understand the nature and range of selfesteem in young children.