RESEARCH QUESTION:
ARE SOCIAL COMPARISON, COMPETITION OR COOPERATION STYLE DESKTOP-BASED SCREEN NOTIFICATIONS PERCEIVED TO BE MOST EFFECTIVE ON WORKERS’ SEDENTARY BEHAVIOUR?

Background: An increase in sedentary computer-based tasks in the workplace, a reduction in physical jobs and peoples’ increased attachment to devices mean that people are sitting down more than ever, which is associated with numerous illnesses, such as cardiovascular disease. Notifications increase self-awareness which is key to behaviour change and therefore tackling sedentary behaviour.

Method: Using a Wizard of Oz style study with participants (N = 21) we were able to simulate participants receiving desktop screen-based notifications. After this, participants evaluated these social comparison, competition and cooperation style notifications using the Perceived Persuasiveness Scale (PPS) to determine which notification style was perceived to be more effective. WhatsApp interviews and a content analysis on data was conducted afterwards to gain insights into what participants thought of the notifications.

Results: Whilst we couldn’t find any significant differences in PPS score, indicating all three notification styles were very similar in perceived persuasiveness, interview data uncovered insights about participants’ preferences of notification and why they are motivated by them. Insights on notification preference; competition is preferred, cooperation is similarly preferred and social comparison is demotivating. Insights for reasons for notification choice are that design of the message is important – it’s recommended that notifications are trustworthy, believable, calm and private. People respond to competition with others and personalised messages with relevant quantitative information. People were not interested in social comparison and comparing themselves with others.

Why was there no significant difference in persuasiveness of notification styles? Firstly, it may be that the three notification styles are similarly persuasive. However, it also could be that our sample size was too small to detect an effect creating a type II error. It could also be that participants were in a state of change (SoC) that meant they were not open to persuasion. For example, someone in maintenance or action SoC has already decided to take action and get up and move more (see State of Change (SoC) graph below) and therefore doesn’t need to be persuaded. We also only exposed participants to each notification style once, which is unlikely to be enough to persuade participants to change their behaviour.