Mom, you're only remembering the bad behavior! the utility of distributed lag models in single-subject research

Abstract

The present study examined if prior events/behaviors accounted for variance in parental perceptions of child behavior beyond that accounted for by those events/behaviors occurring simultaneously with the assessment of the parental perception. By utilizing a time-series analysis, this study investigated the relative effects of positive and negative child behaviors, interparental behavior, and intrapersonal behavior/events occurring simultaneously or at an earlier point in time on a mother's daily satisfaction with her clinic-referred child. Using a single-subject methodology with replication, nine mothers served as subjects. Data were collected across 30 days for participants, and subsequently, each mother's data were analyzed by multiple regression techniques. The results indicate the usefulness of distributed lag models for single-case designs in the absence of autocorrelation. Analyses indicate that for some mothers what happens on prior days with her child is as important or more important in determining her satisfaction with the child than what happens on the current day. In particular, mothers tend to carry over previous days' negative child behaviors into the present day's satisfaction rating. Previous days' positive and negative spouse behaviors and daily activities did not consistently contribute to regression equations predicting mothers' daily satisfaction ratings with their children.