The psychometric characteristics and construct validity of an inventory of expectancy statements about breast- and formula-feeding were assessed to refine a measure of the predictors of breastfeeding initiation and maintenance. Respondents were a tri-ethnic sample of mostly lower income mothers within two days after delivery. Ninety-six expectancy items reflecting eight hypothesized domains in expectancies of breastfeeding (BF) and formula-feeding(FF) were submitted to maximum likelihood factor analysis. Four factors were determined: “Benefits of BF,” “Relative Advantages of FF,” “Inconveniences of FF,” and “Relative Inconveniences of BF.” All factors were significantly related to actual BF of the new infant in the expected directions using unweighted factor scores in a multiple logistic regression analysis. Predicted BF closely matched actual BF for all ethnicities combined with a total error rate for misclassification of only 14%. This match was high for Anglo- and Black-American mothers. Each of the factors was determined to be reliably estimated by the ten most highly loaded items without an appreciable loss of internal consistency. Further research must assess expectancies prior to delivery in order to assess their predictiveness of behavior, and must further clarify the meaning of the expectancy factors to facilitate intervention.
ASJC Scopus subject areas
- Applied Psychology
- Public Health, Environmental and Occupational Health