Using event-related potential and behavioural evidence to understand interpretation bias in relation to worry

Abstract

Wo rry is a common experience, thought to be maintained by the tendency of interpreting ambiguous information in a consistent (e.g. negative) manner, termed "interpretation bias". This study explored whether high worriers (Penn State Worry Questionnaire score, PSWQ ≧ 56) and low worriers (PSWQ score ≦ 39) show different interpretation biases, and examined at which stages of information processing these interpretation biases occur. Participants with high and low worry levels completed interpretation assessment tasks yielding behavioural and event-related potential indices. We focused on the N400 component, reflecting whether given interpretations were in line with or violated participants' own interpretations. We found that high worriers lack the benign interpretation bias found in low worriers from the early "online" interpretative stage, reflected by the reaction time in a relatedness judgment task and the N400 in a lexical decision task, to the later "offline" stage at which participants had time for reflection. Our results suggest that a benign interpretation bias may be a protective factor in relation to worry and is likely to remain active across online and offline stages of interpretation processing.

Impact of imagery-enhanced interpretation training on offline and online interpretations in worry

“…In line with previous research on socially anxious populations using the LDT to examine interpretation bias (Hirsch & Mathews, 1997;2000), Feng et al (2019) reported that low worriers displayed a positive interpretation bias, whereas high worriers demonstrated neither a positive nor negative bias, evidenced by both reaction time and N400 ERP interpretation bias indices. In contrast to the N400 findings in Feng et al (2019), the earlier of Moser's studies (Moser, Hajcak, Huppert, Foa, & Simons, 2008) failed to find an N400 difference between high and low socially anxious groups in an interpretation assessment task, where ambiguous sentences were resolved in a negative or positive manner by the final grammatically correct words. Unlike Feng et al (2019) or the current study, Moser et al (2008) also looked at P600, a later ERP that has the potential to indicate whether expected interpretations are violated or not (i.e., expectation violation effect).…”

“…In contrast to the N400 findings in Feng et al (2019), the earlier of Moser's studies (Moser, Hajcak, Huppert, Foa, & Simons, 2008) failed to find an N400 difference between high and low socially anxious groups in an interpretation assessment task, where ambiguous sentences were resolved in a negative or positive manner by the final grammatically correct words. Unlike Feng et al (2019) or the current study, Moser et al (2008) also looked at P600, a later ERP that has the potential to indicate whether expected interpretations are violated or not (i.e., expectation violation effect). They found a difference in interpretation bias, indicated by P600, between high and low socially anxious groups.…”

“…The paucity of research into online interpretations of ambiguity in general and worry in particular has begun to be addressed. Feng et al (2019) examined offline and online markers of interpretation bias in individuals with high or low levels of trait worry. Online interpretations were captured using a lexical decision task (LDT) in which real word targets matched either positive or negative resolutions to ambiguous scenarios.…”

“…Online interpretations were captured using a lexical decision task (LDT) in which real word targets matched either positive or negative resolutions to ambiguous scenarios. Feng et al (2019) measured both reaction times and event-related potentials (ERP) as interpretation bias indices.…”

Looking on the bright side reduces worry in pregnancy: Training interpretations in pregnant women

“…In this first study of interpretation training in pregnant worriers, we successfully induced a positive interpretation bias using CBM-I. Consistent with Hirsch et al (2009) and Feng et al (2019), participants in the CBM-I condition reported fewer negative thought 3) As GAD7 scores were significantly different at baseline we re-ran the regression analysis with mean centred GAD7 scores and an interaction variable of (mean centred) GAD7 and condition. Neither GAD7 scores (p = .67) or the interaction term (p = .54) were significant predictors in the model.…”