Exploration of computerised automatic processing tasks and their use in digital psychological health interventions in real-world settings amongst adults

Abstract

Introduction: Digital psychological health interventions, targeting both behaviour change and mental health disorders, have shown promising results at creating positive change in a variety of health areas. These interventions are based on dual-process theories, which propose that the brain uses two types of thinking to influence behaviour: automatic processing and reflective processing. Automatic processing is fast, immediate, non-conscious, and unintentional. Reflective processing centres around logical reasoning; it is slow, step-by-step, voluntary and intentional. Most health interventions tend to solely target the reflective system; even though repeated laboratory testing has found behaviour can be influenced via targeting automatic processing pathway. This happens through conditioning; which involves pairing a stimulus that naturally elicits a response with one that does not until the second stimulus elicits a response like the first (39). However, there is a lack of translation of automatic processing tasks into real-world settings and a lack of consolidated evidence on how effective these tasks are in a real-world setting. Further, automatic processing tasks are often described as ‘boring’ by participants. Having an enjoyable experience could play a significant role in amplifying the conditioning effects of automatic processing tasks. Enjoyment can be increased via gamification, which uses gaming style elements (rewards, points and leader boards) to increase engagement. This thesis aims to further investigate the use of automatic processing tasks in digital health interventions in real-world settings and address the gaps identified by the completion of two studies. Study 1 will address the lack of consolidated evidence of automatic processing tasks in real-world settings by way of a scoping review, and study 2 will explore the role of enjoyment in contributing to the effectiveness of automatic processing tasks by evaluating the role of enjoyment as a moderator in an effective gamified automatic processing app called ‘Flex.’ Methodology: Study 1 followed the methodologies proposed by Arskey and O’Malley (2005), and Colquhoun (2016) for scoping reviews. A study protocol highlighting search strategy and inclusion and exclusion criteria was developed before conducting the review. Study 2 was nested within a pilot randomised control trial (RCT) of a gamified automatic processing smartphone app called ‘Flex’. The RCT found, that at 24-hour follow-up, there was a significant between-group effect such that people in the Flex experimental group had more favourable automatic associations of physical activity than those in the Flex control group. Our nested study was interested in further exploring this association, by examining if the moderator ‘enjoyment’ was responsible for the increase in effectiveness of automatic processing of physical activity stimuli. This was done by evaluating enjoyment through a validated survey. Results: Study 1: 4,038 studies were found for possible inclusion into study 1; however, only 14 studies met inclusion/exclusion criteria. Attentional bias modification tasks were the most commonly used (n=7). The review concluded that use of smartphone technology to deliver interventions was increasing, but there were mixed findings of effectiveness automatic processing tasks. There was also preliminary evidence for positive effects of gamification. Study 2 found that automatic processing of physical activity stimuli in the Flex app was not impacted by enjoyment as a moderator and that enjoyment did not contribute to amplifying the conditioning effect observed in the RCT. Conclusions: This thesis found that there were encouraging real-world results for the use of automatic processing tasks to influence people’s attitudes. The first of these two studies found that there was some evidence of automatic processing health interventions being successful at changing people’s automatic associations towards targeted stimuli in real-world settings. This is promising given that the majority of studies into these tasks have occurred in a lab-based setting and highlights the translational potential of these tasks into the real-world. In terms of why some of these tasks did not work, one possible reason could be due to the repetitive nature and lack on engagement of these tasks. The second study found that enjoyment was not a moderator for the positive associations observed in the RCT. However, encouraging process evaluation results provided insights into participant experience of the app and what elements participants found enjoyable and frustrating. This thesis has been able to contribute to the developing area of automatic processing tasks in a real-world setting. We have been able to highlight that this field, to-date, is still very small but developing and with further research, particularly around increasing engagement, these tasks could provide an innovative solution for improving existing health interventions.