Development and reliability testing for measures of psychosocial constructs associated with adolescent girls' calcium intake

Abstract

Early adolescence is an important period for assuring adequate calcium intake to achieve optimal bone health. This study was conducted to develop a measure of psychosocial constructs likely to be associated with young adolescent girls' consumption of high-calcium foods and to evaluate the test-retest reliability and factor structure of the measure. Methods involved: (a) generating a pool of items; (b) a cross-sectional pretest of the survey with a small group; (c) longitudinal administration of the revised survey with a larger group; and (d) statistical analysis to determine psychometric properties of the measures. Seventeen girls completed the pretest and debriefing, and 217 girls participated in the pilot test and test-retest (completion rate 95%). The study took place in Hawaii in late 2002. Coefficient alpha was used to test internal consistency and test-retest reliabilities were calculated. Cluster analyses of the items were performed to evaluate the factor structure of the measure. The resulting measure was a 55-item questionnaire to assess key psychosocial variables in three categories: Social and Environmental Factors (Availability, Social Influence); Attitudes and Preferences (Health Benefits; Preferences; Taste; Weight; Tolerance; Convenience); and Knowledge. The survey had high internal consistency reliability (alpha>.75) and good test-retest reliability (0.73 to 0.78 except for Knowledge, where Spearman's rho=0.41). The study developed a measure with high internal consistency and test-retest reliability to assess constructs that can be used in descriptive, prospective, and intervention research and as possible explanatory or mediating variables.