Abstract
Dual-task reaction times are a classic measure of the attentional demands of a stimulus. The paradigm has also been applied to more complex visual stimuli including film and television content, demonstrating that both low-level features (such as salience) and higher-level features (such as narrative) can drive attention. This set of 3 experiments compares dual-task reaction times to physiological measures and validates the dual-task approach against self-reported measures of engagement. In Experiment 1 (n = 24), we simultaneously obtained heart rate and skin conductance recordings, while participants completed a secondary choice tone discrimination task. We compared measures while participants watched BBC television content, to baseline periods in which no content was played. In Experiment 2 (n = 24), we assessed the direct influences of the dual task on the physiological responses with three conditions: (i) tones were present and participants had to respond, (ii) tones were present but no response was required, and (iii) a control where no tones were present. Experiment 3 (n = 165) was an online experiment where participants watched short TV clips while completing the dual-task paradigm, and rated their narrative engagement in each clip using a standard questionnaire. Dual-task reaction times were significantly correlated with heart rate and skin conductance. Skin conductance and heart rate increases were present when participants were required to make a response, but not when instructed to ignore the tones. Additionally, self-reported narrative engagement was predictive of participant’s reaction times. Our results indicate that the dual-task paradigm can be used to assess dynamic fluctuations in attention during complex dynamic visual stimuli. Dual-task reaction times also appear to be a direct and objective way to measure the wider construct of audience engagement.