Poster #43 - The Dual-Pathway Model and Dimensionality of Self-Control: Inhibitory Control, Delay-of-Gratification, and Delay-Aversion

• In Event: 1-061 - Poster Session 02
  In Poster Session: PS 02 Section - Developmental Disabilities

Integrative Statement

The dual-pathway model (Sonuga-Barke, 2002, 2003) proposes Attention-Deficit/Hyperactivity Disorder (ADHD) develops along two separate but interrelated developmental pathways. The first, a cognitive pathway accounting for executive function deficits as a primary deficiency in inhibitory control (IC). The second pathway is a hypersensitivity to delay (i.e., “delay-aversion”), accounting for ADHD symptoms as functional expressions of motivational system deficits.
Studies examining the dual-pathway model use choice-paradigm tasks to operationalize self-control. Like Mischel and colleagues’ delay-of-gratification (DG) tasks, in which children are given a choice between smaller/sooner (SS) or larger/later (LL) rewards. However, studies suggest DG tasks tap both components of the dual-pathway model and therefore, require some degree of IC. Delay-aversion tasks have been created to reduce the influence of IC, by preventing children from switching their response during the delay period. However, it is unclear whether IC, DG, and delay-aversion tasks are measuring truly distinct constructs. Accordingly, this study measured IC, DG, and delay-aversion to examine the dimensionality of self-control.
Preschool children (N = 134) completed multiple age-appropriate performance-based measures of IC, DG, and delay-aversion. Teachers and parents completed report-based measures of participants’ ADHD-related behaviors. Based on previous literature, we hypothesized that IC, DG, and delay-aversion would be best represented as three moderately correlated factors.
Confirmatory factor analysis (CFA) was used to evaluate the fit of one-, two-, and three-factor models of IC, DG, and delay-aversion. Table 1 shows fit indices and model comparisons for these models. The one-factor model provided poor fit to the data. Three two-factor models were examined by using all possible construct pairings. Model B, with separate Delay and IC factors, provided adequate fit, fit significantly better than the one-factor model, and had weakly correlated factors (r = -.07). Model C, with separate Delay-Aversion and IC + DG factors, provided good to excellent fit and fit significantly better than the one-factor model and slightly better than Model B (e.g., comparison of AIC). Model C had weakly correlated factors (r = .09). Model D, with separate DG and IC + Delay-Aversion factors, provided poor fit to the data and had moderately correlated factors (r = .65). The three-factor model with separate IC, DG, and Delay-Aversion factors provided good to excellent fit but did not fit significantly better fit than any of the two-factor models, and the factors had relatively small correlations (rs = .28, .07, .36).
Contrary to expectations, results indicated that measures of IC, DG, and delay-aversion were best conceptualized as two factors, rather than three distinct constructs. Specifically, Model C indicated that DG and IC tasks were best represented as a single factor, and Delay-Aversion was a distinct factor (see Figure 1). These results are consistent with previous literature indicating similarities between IC and DG tasks (e.g., Dalen et al., 2004) and suggest that modifications made to delay-aversion tasks sufficiently reduce the influence of IC during the delay period. Additional analyses will examine the common and unique relations between the identified factors and preschool children's ADHD-related behaviors.

Authors

• Brittany M Morris, Florida State University
  Presenting Author

• Christopher Lonigan, Florida State University
  Non-Presenting Author