Measures

Adolescents were asked to rate the extent their parents directed verbal aggression or criticism towards them and parents were asked to rate the extent to which they directed verbal aggression or criticism towards their adolescents using three items from the Conflict subscale of the Parent-Child Relationship Inventory at Time 2 (Hetherington & Clingempeel, 1992). This subscale consists of questions such as, “How much do you yell at this child after you’ve had a bad day?” and uses a 5-point response scale ranging from 1 (extremely) to 5 (not at all). A composite was computed using an average of parent and adolescent responses on three items, reverse-scored, and averaged. Higher scores were indicative of greater parent-child negativity. Subscale reliability was acceptable with internal consistency of α = .72 at Time 2 for adolescents and α = .69 at Time 2 for parents.

The short version of the Confusion, Hubbub and Order Scale (CHAOS) was administered to adolescents and parents to measure self-reported degree of chaos in the household at Time 2 (CHAOS; Matheny, Wachs, Ludwig, & Phillips, 1995). This scale distinctly taps into the degree of ambient noise, crowding, and traffic in the household. Adolescents and parents were asked to rate six statements about their households such as, “You can’t hear yourself think in our home,” and, “We have a regular morning routine at home.” The items were rated on a 5-point Likert scale ranging from 1 (definitely untrue) to 5 (definitely true) and a composite was computed using an average of adolescent and parent responses on the six items at Time 2. Higher scores were indicative of greater household chaos. Scale reliability was consistent with prior studies (Deater-Deckard, Chen, Wang, & Bell, 2012; Deater-Deckard et al., 2009) with internal consistency of α = .64 at Time 2 for adolescents and α = .62 at Time 2 for parents.

An index of adolescents’ and parents’ and delay discounting was derived using a computerized delay discounting task (Johnson & Bickel, 2002). Adolescents completed the task at Time 1 and Time 3 whereas parents completed the task at Time 1. Participants were given a series of hypothetical monetary decisions in which they made choices between an immediate monetary reward and a larger monetary reward with a delay. The reward amount chosen was $100. Choices were presented using the following delays: One day, one week, one month, and one year. Individual indifference points were calculated using the area under the curve (AUC) approach to measurement (Myerson, Green, & Warusawitharana, 2001). AUC values can range from 0 to 1, with 0 representing extreme discounting and 1 representing no discounting. We performed ordinal transformation of AUC values to retain equal contributions of each delay to the overall AUC using the prescribed methods of Borges, Kuang, Milhorn, and Yi (2016). We then used the Johnson and Bickel (2008) algorithm for identifying and excluding cases demonstrating nonsystematic discounting from the analysis for violating the assumption of monotonic decreases in discounting function. Data were identified as nonsystematic for either or both of two reasons: (1) If following the first delay, an indifference point was greater than the previous indifference point by 20% of the larger, later reward, and (2) if the last indifference point (calculated at one year) was not any different from the first indifference point (calculated at one day). Less than 10% of parent delay discounting cases were identified as nonsystematic (n = 14), whereas less than 5% of adolescent delay discounting cases were identified as nonsystematic at Time 1 (n = 6) and less than 5% of adolescent delay discounting cases were identified as nonsystematic at Time 3 (n = 6). We computed adolescents’ delay discounting residualized change scores by regressing Time 3 delay discounting on Time 1 delay discounting, with lower scores indicating lesser declines in adolescents’ delay discounting from Time 1 to Time 3.