Avoidance of tobacco health warnings? An eye-tracking approach.

Aims: Among three eye-tracking studies, we examined how cigarette pack features affected visual attention and self-reported avoidance of and reactance to warnings. Design: Study 1: smoking status × warning immediacy (short-term versus long-term health consequences) × warning location (top versus bottom of pack). Study 2: smoking status × warning framing (gain-framed versus loss-framed) × warning format (text-only versus pictorial). Study 3: smoking status × warning severity (highly severe versus moderately severe consequences of smoking). Setting: University of Bristol, UK, eye-tracking laboratory. Participants: Study 1: non-smokers (n = 25), weekly smokers (n = 25) and daily smokers (n = 25). Study 2: non-smokers (n = 37), smokers contemplating quitting (n = 37) and smokers not contemplating quitting (n = 43). Study 3: non-smokers (n = 27), weekly smokers (n = 26) and daily smokers (n = 26). Measurements: For all studies: visual attention, measured as the ratio of the number of fixations to the warning versus the branding, self-reported predicted avoidance of and reactance to warnings and for study 3, effect of warning on quitting motivation. Findings: Study 1: greater self-reported avoidance [mean difference (MD) = 1.14; 95% confidence interval (CI) = 0.94, 1.35, P < 0.001, η _p ² = 0.64] and visual attention (MD = 0.89, 95% CI = 0.09, 1.68, P = 0.03, η _p ² = 0.06) to long-term warnings, but not for reactance (MD = 0.14, 95% CI = –0.04, 0.32, P = 0.12, η _p ² = 0.03). Increased visual attention to warnings on the upper versus lower half of the pack (MD = 1.8; 95% CI = 0.33, 3.26, P = 0.02, η _p ² = 0.08). Study 2: higher self-reported avoidance of (MD = 0.70; 95% CI = 0.59,0.80, P < 0.001, η _p ² = 0.61) and reactance to (MD = 0.37; 95% CI = 0.27, 0.47, P < 0.001, η _p ² = 0.34) loss-framed warnings but little evidence of a difference for visual attention (MD = 0.52; 95% CI = –0.54, 1.58, P = 0.30, η _p ² = 0.01). Greater visual attention, avoidance and reactance to pictorial versus text-only warnings (all Ps < 0.001, η _p ² > 0.25). Study 3: greater self-reported avoidance of (MD = 0.37; 95% CI = 0.25, 0.48, P < 0.001, η _p ² = 0.33) and reactance to (MD = 0.14; 95% CI = 0.05, 0.23, P = 0.003, η _p ² = 0.11) highly severe warnings but findings were inconclusive as to whether there was a difference in visual attention (MD = –0.55; 95% CI = –1.5, 0.41, P = 0.24, η _p ² = 0.02). Conclusions: Subjective and objective (eye-tracking) measures of avoidance of health warnings on cigarette packs produce different results, suggesting these measure different constructs. Visual avoidance of warnings indicates low-level disengagement with warnings, while self-reported predicted avoidance reflects higher-level engagement with warnings.