Examining the association between anxiety and attentional engagement in the context of COVID-19 public service announcements
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Waterloo
Abstract
Public service announcement (PSA) effectiveness depends on both message features and
audience characteristics. Yet, few studies have systematically examined theorized predictors of
effectiveness, and none has combined self-report with eye-tracking to assess engagement with
custom-constructed COVID-19 PSAs. The present study used self-report and eye-tracking data
from the Canadian COVID-19 Experiences Project (CCEP) to examine whether baseline anxiety
predicted engagement with COVID-19 PSAs and whether behavioral focus (masking,
vaccination, hand hygiene, distancing) moderated this relationship. Temporal bridging and
vividness factors were also examined as secondary analyses. Self-report (n = 120) and eyetracking
data (n = 101) were analyzed using confirmatory factor analysis and linear mixedeffects
models; primary models involving baseline anxiety used 116 participants with complete
GAD-7 data. Findings indicated that a two-factor model distinguishing self-report and
physiological engagement fit significantly better than a one-factor model (Δχ²(1) = 28.39, p <
.001), suggesting related but distinct aspects of engagement. Contrary to the primary hypothesis,
baseline anxiety did not show a reliable overall main effect across outcomes. However,
behavioral focus moderated several self-report outcomes. Notably, the positive association
between anxiety and subjective engagement was attenuated for vaccination PSAs (β = −0.013, p
= .041). Secondary analyses showed that vividness was strongly associated with higher selfreported
engagement in the vaccination subset, and that this relationship was stronger among
more anxious participants for the self-report factor score (Δχ²(1) = 5.73, p = .017). Temporal
bridging, however, did not meaningfully influence engagement. Overall, anxiety appears to be
associated with subjective responses to COVID-19 PSAs depending on message characteristics,
while corresponding eye-tracking differences were less apparent.