Estimating Home Visit Activities: How Much Observation Is Enough?

Background: More than two million U.S. families receive home visiting services each year, yet little is known about the content and process of home visit interventions. Home visit observations are a recommended tool to address this gap in knowledge, but guidance regarding how much observation time is needed to provide a trustworthy picture of home visits is not available. Objective: This study estimated the length of observation needed to provide a reliable description of the content and process of home visits delivered through an Early Head Start (EHS) home-based program. Methods: Observations of 555 home visits involving 88 families and 17 home visitors were coded with the Home Visit Observation Form--Revised (HVOF-R). The HVOF-R describes the proportions of time spent among several subcategories within each of three major categories: Primary Interaction Partners, Content of Interactions, and Nature of the Home Visitor's Interactions. The psychometric theory of test reliability guided use of bootstrap resampling and the Spearman-Brown prediction formula to estimate the duration of observation necessary to achieve conventional psychometric standards of reliability (r = 0.90 and r = 0.80). Results: Results showed that 40-60 min of a home visit observation yields a reliable estimate of the proportion of time spent on most key features of home visits, especially those behaviors observed most frequently, including interactions among the home visitor, parent, and/or child. Conclusion: Fairly brief observations yield reliable description of home visit activities, and use of observation is recommended to guide program implementation efforts and enhance program quality.