Fathers' Involvement, Early Childhood Development and Risk: Employing the Language Environment Analysis (LENA) Technology in Understanding Fathers' Verbal Interactions with Their Young Children

Background and Purpose: Early verbal engagement in the home predicts young children’s cognitive development and has been associated with home environments that predict healthy psychosocial development. Knowledge about early verbal engagement in the home has been largely focused on mother-infant interactions, although fathers’ early interactions with their very young children have very recently come into focus. Of the extant studies, only a handful have employed observational assessments of fathers’ interactions, and of these, almost none have done so in a low-income diverse sample of young families. Given this, we examined fathers’ verbal interactions with their very young children and self-reported measures of their involvement in a study of a “father inclusion” intervention for low-income urban families. We employ a novel technology called the LENA (‘Language Environment Analysis”) system that audio-records parents’ and babies’ interactions, and use a validated computer algorithm, translating these into quantitative metrics of mothers’ and fathers’ word counts and infants’ vocalizations.

Methods: We present data from both a small quasi-experimental study and a subsequent larger multisite cluster randomized controlled trial examining a “father inclusion” intervention (dubbed Dads Matter-HV) for perinatal home visitation services. The first quasi-experimental study enrolled 24 families (biological mothers, biological fathers and their infants) comparing standard home visitation versus standard home visitation plus Dads Matter-HV. In the multisite RCT, we enrolled 204 families served by 19 Chicago-based programs and randomly assigned home visitation teams across study conditions. In both studies, we collected self-reported and LENA baseline data from biological mothers and biological fathers at the point of service initiation, and again after the delivery of the Dads Matter-HV enhancement at 4- months. Samples across both studies consisted of approximately two-thirds Latine families, and one-quarter African American families. All families were enrolled during early childhood (from prenatal to child under 24-months old). Prior validation studies have shown that the LENA system equally reliably codes word counts in Spanish and in English.

Results: Congruent with findings from self-reported measures, the initial quasi-experimental study found pre-to-post Dads Matter-HV intervention mean increases of 12.1 father words with their infants per five minute segment (sd = 22.3), while the standard services comparison group showed a mean decrease of 20.3 father words per five minute segment (sd = 41.7), resulting in an overall effect size d = 0.69, favoring the Dads Matter-HV intervention group. While outcome analyses continue from the larger multisite RCT, fathers’ verbalizations (word counts) positively correlate with the fathers’ self-reported involvement with their child (r = 0.25), positively correlate with the quality of the mother-father relationship (r = 0.31), and negatively correlate with abuse in the mother-father relationship (r = -0.28), providing evidence of convergent validity across LENA indices and self-reported hypothesized predictors and correlates.

Conclusions and Implications: This study shows the utility of the LENA algorithmic technology as an important measure of early parent verbalization for both descriptive and intervention studies of fathering in early childhood.