footnote239
This study uses a quasi-experimental design and compares two groups to <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr>. First, as part of a larger longitudinal study, families who were randomly assigned to <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr> or another intervention called Developmental Education for Families (DEF) as part of another study, and who also were involved with child protective services, were compared. The study analyzed a subsample of cases from the larger study, and did not state how the subsample was selected. Therefore <abbr title="Home Visiting Evidence of Effectiveness">HomVEE</abbr> assumes that subsample selection is nonrandom and evaluates the <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr> to DEF comparison as a quasi-experimental design. Quasi-experimental designs must establish equivalence on race/ethnicity, <abbr title="socioeconomic status">SES</abbr>, and baseline measures of the outcome (and control for baseline measures) in order to earn a moderate rating, but this study does not meet all of these criteria. For the <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr> to DEF comparison, although the samples are equivalent on race/ethnicity and <abbr title="socioeconomic status">SES</abbr>, the study does not assess or control for baseline measures of the outcomes (which are all parenting practices), so that contrast received a low rating. The study also compares the <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr> group to another group of low-risk families identified as parts of other studies. The different approach to recruiting the samples for the <abbr title="Attachment and Biobehavioral Catch-Up ">ABC</abbr> and the low-risk comparison groups introduces a confounding factor that automatically rates that comparison low.