Dynamic changes in the use of online advice in response to task success or failure

Abstract

A range of responsible gambling systems have been developed, but they are not necessarily used by consumers. This requires an understanding of factors (e.g. risk and loss) influencing response to advice. Therefore, an experiment examined the use of decisional support in response to induced success or failure. A total of 21 university students played computer Blackjack at two levels of risk, and were informed as to whether the odds were in their favour ('card count'), and were given online advice as to card selection (Basic). Winning streaks led to overconfidence, as indicated by larger mean bets, more Basic errors, and a failure to adjust wagering upon cessation of the winning streak. Participants were more cautious in response to losses, betting less, and were more likely to comply with advice. Participants attended and complied with advice earlier at low risk, but at high risk noticed advice later and did not comply. Induced emotional states (e.g. winning or losing) are important as they influence the likelihood that people will listen to online advice. Designers of Behaviour Change Support Systems should consider methods of contingently implementing advice. ARTICLE HISTORY

Process-based Guidance for Designing Behavior Change Support Systems: Marrying the Persuasive Systems Design Model to the Transtheoretical Model of Behavior Change

“…Based on the theoretical analysis and insights from the literature review in step two, we supplemented our example from step one as follows: the design principle suggestion is not applicable / not recommended for designing BCSS that target the first transition. In fact, from our literature review, we found that users tend to be averse to advice when they do not experience any problems yet (Phillips & Landon, 2016). As for the second and third transitions, Nguyen et al (2018) and Song et al (2017) implemented suggestions to reveal the benefits associated with a changed behavior and an approach to take the first steps toward it.…”

“…In the first transition, the design principles reduction, self-monitoring, rehearsal, praise, rewards, reminders, suggestion, social comparison, social facilitation, cooperation, competition, and recognition cannot address users' needs. The model does not recommend the design principle suggestion for the first transition because it can lead users to reject the precontemplation stage when they do not have problem awareness yet (Phillips & Landon, 2016). Design principles such as rehearsal, praise, and reminders require a certain awareness of the problem that people in this stage lack.…”

“…Suggestion offers specific applications of the target behavior, such as an exercise plan based on preferences and goals (Lehto & Oinas-Kukkonen, 2015). Advice proves especially effective when users experience some form of loss or the situation involves low risk (Phillips & Landon, 2016). Rehearsal explicitly helps people prepare for real situations (Oinas-Kukkonen & Harjumaa, 2009) as a training technique (Langrial et al, 2014).…”

Thinking without knowing: Psychological and behavioral consequences of unjustified confidence regarding blackjack strategy

“…One further limitation of this work is that we did not investigate the factors that produce unjustified confidence (i.e., the antecedents to unjustified confidence in Figure 1 ), instead taking confidence as a given (Study 1) or manipulating it in a manner distinct from what would happen in real-life (Study 2). Nonetheless, it is informative to compare our results to that of a paper by Phillips and Landon (2016) . In their work, they manipulated whether participants had a winning streak or losing streak and found that participants who had a winning streak were more likely to bet more on subsequent trials and were less apt to take advice designed to improve play than were participants on a losing streak.…”

“…More generally, this research stream points to the need to consider the ramifications of situations when confidence increases independent of knowledge. Phillips and Landon’s (2016) work showing that winning in blackjack can cause an increase in betting and less willingness to take blackjack advice provides a striking demonstration of this concern; even a positive outcome can have a long-term negative impact under certain circumstances. Note this concern holds even for situations where knowledge increases, but at a rate less than the increase in confidence.…”