Poster #79 - Parental Unpredictability and Motivation for Instrumental Learning in Adolescents

• In Event: Poster Session 18

Abstract

Background
Childhood maltreatment is associated with increased problem behaviors, and impaired social learning may be one of the cognitive pathways. What children learn from their social environments and how they use this information to inform their behaviors is a crucial factor that leads to individual differences in socio-emotional outcomes. Childhood maltreatment has been shown to alter how children use reward information in instrumental learning. But little is known about its effect on children’s use of emotional information. This pre-registered study will address this gap. Specifically, we will test the effect of parental unpredictability, a characteristic of maltreating parents that has been suggested to alter cognitive and emotional processing differently from parental harshness.

Hypotheses
First, emotional feedback can motivate children’s action learning, but it is less salient than reward information. Second, children with more unpredictable parents are less effective in using emotional information to guide their action learning. The rationale is that parental unpredictability may lead children to assume more randomness in action-outcome relationships and have more blunted sensitivity to varied emotional information.

Measurements
Dependent variable: Action learning will be assessed with a modified Go-NoGo task (de Berker et al., 2016). In this task, participants will use emotional information—positive and negative images from the OASIS dataset (Kurdi et al., 2017) to learn how to respond to different cues across 120 trials. Each cue is associated with an initially unknown correct action (Go versus NoGo) and an outcome (positive, negative, or neutral image).

Independent variable: Child-reported parental unpredictability will be measured using the Questionnaire of Unpredictability in Childhood (QUIC; Glynn et al., 2018).

Control variables: Parental harshness will be measured with Parental Environment Questionnaire Conflict Subscale (Miller & Hauser, 1989); children’s motivational sensitivity will be measured with the Behavioral Inhibition and Behavioral Activation System scale (BISBAS; Carver & White, 1994); socioeconomic status will be measured with a demographics information of household income; children’s cognitive ability will be measured with the digit span task.

Sample size
An a priori power analysis suggests that 91 participants are sufficient to achieve a power of .80 to detect an effect size of partial eta squared = 0.10 or larger.

Planned analysis
We will fit a generalized Linear Mixed Effects Model to examine whether the percentage of Go responses varies as a function of a three-way interaction between action (Go versus NoGo), valence (positive versus negative images), and parental unpredictability. For the first hypothesis, we will compare the effect size of emotional information to that of reward information in past research. For the second hypothesis, we will test whether the interaction between cue types and parental unpredictability is significant when predicting Go responses. If so, it suggests that children with varying exposure to parental unpredictability show differences in their ability to learn how to respond to different cues.

Implications
This study will help us (1) understand how children use different types of information to learn about their environments and (2) identify factors underlying differences in children’s responsivity to reward versus emotional feedback in instrumental learning.

Authors

• Kinjal Mehta, University of Wisconsin - Madison
  Presenting Author

• Yuyan Xu, University of Wisconsin - Madison

• Seth David Pollak, University of Wisconsin - Madison