Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Mary L. Kelley


This study was the first to apply a behavior analytic model of assessment to failure to thrive (FTT) by observing parent and child behavior during mealtimes at hospitalization for provisional diagnosis of FTT. Descriptive data (e.g., child growth parameters, temperament, and developmental status; maternal medical history and psychopathology; and demographic information) also were collected. The specific goal, however, was to identify feeding behaviors that differed in rate of occurrence in parent-child dyads in which the child was classified: (a) nonorganic failure to thrive (NOFTT) or Mixed FTT (e.g., physical and psychosocial etiology), (b) organic failure to thrive (OFTT), or (c) normal weight and hospitalized due to acute illness (control). Statements with regard to differential behavior patterns for control and NOFTT-Mixed FTT dyads must be viewed cautiously due to nonsignificant multivariate group comparisons. However, trends, based on effect size statistics, indicated behaviors that were the most discrepant between groups were those that accentuated the reciprocity of parent-child interaction in the development of NOFTT. Furthermore, correlations of some behaviors with child's age and feeding method, implied that "types" of FTT based on child's age (Linsheid & Rasnake, 1985) may be an appropriate conceptualization. Specifically, NOFTT-Mixed FTT parents were more likely to display a lack of active visual, verbal, or physical contact with their child during meals (i.e., Non-Interaction) and were generally less interactive than control parents (i.e., Mean Parent Behavior, Non-Negative Verbal). Relative to control children, NOFTT-Mixed FTT children displayed less non-aversive vocal or physical behavior (i.e., Social Interaction). Unexpectedly, control children tended to display food refusal more often than NOFTT-Mixed FTT children. In spite of nonsignificant multivariate group comparisons, this study provided a useful methodology for observational studies of parent-child feeding interactions in hospitalized FTT and normal weight children. Because trends in the behavioral data generally supported this study's hypotheses, the Feeding Interaction Code (FIC) may be viewed as a workable observational system to evaluate parent and child mealtime behavior.