This article is available in: PDF HTML Gambling Data and Modalities of Interaction for Responsible Online Gambling: A Qualitative Study

Journal Information
Journal ID (publisher-id): jgi
ISSN: 1910-7595
Publisher: Centre for Addiction and Mental Health
Article Information
Article Categories: Original Research
Publication date: Spring 2020
Publisher Id: jgi.2020.44.8
DOI: 10.4309/jgi.2020.44.8

Gambling Data and Modalities of Interaction for Responsible Online Gambling: A Qualitative Study

George Drosatos Department of Computing and Informatics, Bournemouth University, Poole, Dorset, UK Institute for Language and Speech Processing, Athena Research Centre, Xanthi, Greece
Emily Arden-Close Department of Psychology, Bournemouth University, Poole, Dorset, UK
Elvira Bolat Department of Marketing, Strategy and Innovation, Bournemouth University, Poole, Dorset, UK
Raian Ali College of Science and Engineering, Hamad Bin Khalifa University, Education City, Doha, Qatar

Abstract

Online gambling, as opposed to land-based gambling and other mediums of problematic and addictive behaviour such as alcohol and tobacco consumption, offers unprecedented opportunities for monitoring and understanding users’ behaviour in real-time. It also provides the ability to adapt persuasive messages and interactions that would fit the gamblers usage and personal context. These features open a new avenue for research on the monitoring and interactive utilization of gambling behavioural data. In this paper, we explore the range of data and modalities of interaction which can facilitate richer interactive persuasive interventions, and offer additional support to limit setting, with the ultimate aim of aiding gamblers, who gamble at low to moderate levels, to stay in control of their gambling experience. The exploration is based on our previous research on online addiction and interviews with experts (ne = 13) from different relevant multidisciplinary backgrounds and different points of view. We also interviewed gamblers (ng = 6) about their perception of the utilization of their data for aiding more conscious gambling. Directed at multiple stakeholders, including the gambling software providers, compliance and responsible gambling personnel, as well as policymakers, this paper aims to provide a basis and a reference point for empowering future responsible gambling socio-technical tools through the capture and utilization of relevant online gambling behavioural data.

Keywords: online gambling, problem gambling, behavioural data, limit setting, persuasive interventions

Résumé

Le jeu en ligne, contrairement aux formes de jeu hors ligne et à d’autres types de comportements problématiques et de dépendance comme la consommation d’alcool et de tabac, offre des possibilités sans précédent de surveillance et de compréhension du comportement des utilisateurs en temps réel, ainsi que la capacité d’adapter des messages persuasifs et des interactions adaptées à l’utilisation des joueurs et au contexte personnel. Cela ouvre une nouvelle voie pour la recherche sur la surveillance et l’utilisation interactive des données comportementales relatives au jeu. Dans cet article, nous explorons à cette fin la gamme de données et les modalités d’interaction qui peuvent faciliter des interventions persuasives interactives plus riches et permettre un soutien accrû pour l’établissement de limites, dans le but ultime d’aider les joueurs de niveaux faibles à modérés à demeurer en contrôle de leur expérience de jeu. L’exploration est basée sur nos recherches antérieures sur la dépendance en ligne et sur des entretiens avec des experts (ne = 13) issus de différents contextes multidisciplinaires pertinents et ayant différents points de vue. Nous avons également interrogé des joueurs (ng = 6) à propos de leur perception de l’utilisation de leurs données pour contribuer à un jeu plus conscient. Ce document vise à fournir une base et un point de référence pour l’autonomisation de futurs outils socio-techniques du jeu responsable grâce à la saisie et l’utilisation de données pertinentes sur les comportements de jeu en ligne, et il est destiné à de multiples parties prenantes, notamment des fournisseurs de logiciels de jeu, du personnel de conformité et de jeu responsable ainsi que des décideurs.

Introduction

Online gambling is on a continuous upward growth trajectory (Gambling Commission, 2016), and the DSM-5 now recognizes gambling disorder as an addictive disorder (American Psychiatric Association, 2013). Online gambling is easy to gain access to, and is enhanced by creative technology that makes the medium increasingly appealing and fascinating to users. The ubiquitous accessibility, through desktop and mobile devices, makes the scale and complexity of the problem even higher compared to traditional gambling machines, such as fixed-odds betting terminals (FOBTs). This situation is exacerbated by the social computing features that can add further problematic capabilities. Examples include accompanying forums that allow gamblers to communicate to share tips and betting stories. Such techniques, along with peer pressure, may extend exposure, stimulate relapse, and prevent efforts to maintain gambling at an acceptable level. This integration of social computing into gambling reflects the increasing socialization of gambling. Television advertising promoting the gambling industry, for example, often highlights the social aspects of activities rather than the potential monetary gain. (An instance might be playing online bingo with friends.) Furthermore, the usage of persuasive techniques in online gambling (e.g., badges and leader boards) may in turn create an even more engaging medium, and therefore increase the risk of gambling being used as a method for avoiding real-life difficulties. It is therefore important to ensure that gambling remains controlled and responsible from the start. Given the limited usage and experience with data-driven technology-assisted tools for responsible gambling, perspectives of both experts and gamblers are needed to determine first the acceptability of such technology. Such a contribution would help the managed induction and introduction of the solutions relying on it. For example, certain gamblers can see it as an enabler for more self-awareness and responsible gambling, while others see it as an enabler for optimizing gambling.

The features that make online and mobile gambling more simultaneously impressive and attractive also provide significant potential to combat the problem of gambling (Zhao, Marchica, Derevensky, & Ivoska, 2018). The accessibility and persuasive techniques utilized in online gambling could equally be used as behavioural change mechanisms to prevent potential problematic behaviour. For example, gambling behavioural data can be used to reveal players’ patterns of both loss of control and loss chasing as experienced by themselves or other players, or by peers in the case of using peer support groups and group therapy. Our research indicates that in online gambling money tends to be perceived as less real. Hence, gamblers can be shown visualizations of the actual value of the money used for gambling. An example would be a progress bar displaying how the spent money compares to 10% of the monthly income and how much of certain goods and services this money could buy. Consequently, the online medium provides a unique opportunity to empower classic behaviour change as it offers real-time responses, interactivity, traceability of usage data, intelligence, personalization and the ability to be context-aware. Building on the established research on influence (Davidson et al., 1999), help seeking and behaviour change (Moos & Moos, 2004), and online addiction labels (Ali, Jiang, Phalp, Muir, & McAlaney, 2015), we advocate persuasive approaches for assisting responsible online gambling behaviour instead of relying solely on compulsive ones (e.g., self-limitation and self-exclusion; Chagas & Gomes, 2017).

Self-regulation theory (Baumeister, Schmeichel, & Vohs, 2007) introduces the concept of self-regulation systems, which are information systems for conscious personal management that involve the process of guiding one’s own thoughts, behaviours, and emotions to achieve a self-designated goal. These systems can be advocated to prevent problematic online behaviour given the nature of the medium which allows various workarounds when classic and coercive approaches are enacted, e.g., using a different device or account. Self-regulation systems are focused on those users who have an active role in changing their own behaviour, as supported by such psychological theories as goal-setting (Fenner, Straker, Davis, & Hagger, 2013), self-monitoring (Miller & Thayer, 1988), and implementation intentions (Hagger & Luszczynska, 2014). A basic assumption and premise would be that people understand the benefits of maintaining control of their behaviour. Furthermore, such self-regulation systems for regulating online behaviour can derive benefit from the online medium itself, to monitor the behaviour (e.g., user’s interaction with an online platform) and introduce addiction mitigation technologies, e.g., interactive warning labels and persuasive techniques such as timers and avatars (Ali et al., 2015). For example, Webb and colleagues (Webb, Sniehotta, & Michie, 2010) state that goal-setting theory provides clear implications for promoting change in addictive behaviours. Goal setting, along with feedback and advice, is a core component of interventions to reduce problem drinking and facilitate smoking cessation (e.g., Scott-Sheldon, Carey, Elliott, Garey, & Carey, 2014; Whitlock et al., 2004). Also, recent studies suggest that limit-setting approaches using pop-up message reduces in the majority of cases (apart from those cases with a financially focused self-concept) gambling expenditures (Tabri, Hollingshead, & Wohl, 2019). Therefore, in our study, we will explore what types of limits could be utilized in a self-regulation system to control the online gambling experience.

The data can be used within an individual or personalised setting (e.g., individual limit setting and plans); within a social setting (e.g., group therapy and online peer support groups), and in a blended modality. Self-regulation systems can be facilitated by persuasive technology techniques exploiting principles of social influence such as authority, social proof, likeability, and commitment (Cialdini, 2006). In addition to helping to regulate individual performance, people feel a sense of belonging when they receive assistance from others, either directly through dialogue support (Oinas-Kukkonen & Harjumaa, 2009) or indirectly through showing others success stories, e.g., data-driven graphs displaying a decline in problematic patterns of other gamblers over time which can act as a social proof and give hope to the afflicted persons.

Furthermore, interventions could be designed based on the Theory of Planned Behaviour (Ajzen, 1991), which holds that attitude, subjective norm (perception of how others feel about the behaviour), and perceived control over a behaviour all influence the intention to perform that behaviour. This process eventually affects whether the individual does in fact execute that behaviour. Feedback on regularity of gambling and amount of bets in relation to others could help individuals regulate their behaviour, in line with the theory of social norms and social comparisons (Festinger, 1954). Comparably, enhanced awareness of how behaviour differs across contexts (e.g., increasing people’s awareness of how their gambling differs based on time and place) could increase perceptual control of gambling. These types of context-aware and social influence interventions are one of the main research questions that will be explored in the present study.

As a primary step towards the generation of self-regulation platforms which collect and use online gambling behavioural data, it is necessary to carry out an in-depth exploration of the range of data and modalities of interaction that can facilitate richer interactive persuasive interventions and variations of limit setting, with the ultimate goal of making gambling a more conscious and informed experience for those who are able to maintain control over their gambling. In line with the principles of usability testing, an assessment of the target group’s views regarding the content and format of an intervention, which is best achieved by conducting qualitative research, is needed during intervention development (Yardley, Morrison, Andreou, Joseph, & Little, 2010). We will also need to explore gamblers’ perceptions of such automated and semi-automated data collection and utilization. This paper builds on our previous work on online gambling and addresses the above point by exploring views from experts in responsible gambling and gamblers through interviews and qualitative analysis. We aim to provide a basis and reference points for future platforms to empower responsible gambling through the capture and utilization of gambling behavioural data.

Gambling Data Flow for Enabling Responsible Gambling

In addition to marketing, personalization and trend analysis, gambling data can be used for responsible and informed gambling. This type of data includes visited pages, navigation paths, played games, tournaments of interest, live betting event status, login status, login frequency, location, computing device used, limits set so far, and tendency to comply with them. Furthermore, these data can be obtained for both past and real-time events. More complex data can additionally be obtained using the gambler’s personal devices. For example, data indicating emotional status can be obtained through affective computing and multimodal interaction techniques (Kostoulas, Chanel, Muszynski, Lombardo, & Pun, 2017).

Enjoying access to these data in a way that is practical and timely for processing would necessitate real-time streaming and formatting of such data, so it could be used as input to algorithms meant for responsible, informed and conscious gambling. The algorithms could then visualize the data in various ways (e.g., charts and infographics), and send recommendations to the gamblers or relevant others to take action. This process would take place under specific contractual constraints and settings and with gamblers’ informed consent.

In practical terms, this means the data would be subject to retrieval by automated and programming means (such as Application Programming Interfaces or APIs) and would also need to be put in place for use of the data by third-party applications or other beneficiaries, such as family members, counsellors and, when authorized by the gambler, therapists. This data sharing stream for the well-being of gamblers and their families is shown in Figure 1. In a typical data flow scenario, the gamblers enjoy the ability to retrieve their personal data located in the gambling operators and also any other third-party data provider, such as a bank and a healthcare provider. Thus, they could use it in their personal device for enabling responsible online gambling through a self-regulation mobile application. Additionally, this application could combine this data with additional multimodal data from the gambler’s environment and the device usage. Potentially, all this data could also, with the gambler’s consent, be shared with other recipients, such as the gambler’s family and friends, researchers, or any other responsible gambling services, for the gambler’s benefit.

image

Figure 1. The data flow of gamblers to third-parties.

Method

Design

Qualitative semi-structured interviews were used to explore the following research questions:

RQ 1. What are the main types of limits gamblers could set to stay in control of their online gambling experience?

RQ 2. What are the main types of online interactive interventions, corrective measures, visualization techniques and infographics that could be applied help gamblers to play responsibly?

RQ 3. What are the main types of data that could measure those limits and inform the gamblers about their activity and level of problem gambling?

These research questions were developed with experts from different relevant multidisciplinary backgrounds, such as computer and data science, and psychology, and who possessed various areas of expertise pertinent to the research, such as persuasive technology or addiction (see Table 2). Experts included the four following persons or groups. First, academics from the UK with established track records established through peer-reviewed publications and projects within the domain of online gambling and its core areas, such as decision making and cyber-psychology. Second, the CEO of an addiction rehabilitation charity. Third, the head and a manager of a gambling rehabilitation centre. And, fourth, three directors of responsible gambling units from within three gambling companies in Europe. The head and the directors had a longstanding experience with responsible gambling, one which enabled them to take a leadership role in their companies.

Gamblers with problematic gambling experience were then interviewed about their perception of the utilization of their data for aiding more conscious gambling and their views about the experts’ responses to the research questions (RQ1–3).

Experts were interviewed to identify potential behaviour change techniques that could be applied to online gambling using the online gambling behaviour data, and gamblers were interviewed to determine their perception of the collection of their data and the application of such techniques.

Participants

In our study, the interviews were conducted in two different groups of participants. Participants in the first group were experts in multiple subject areas in relation to addiction, persuasive technology and the gambling industry. This was used to explore RQ1 to RQ3. Demographic information for these participants is presented in Table 1 and details of each participant are shown in Table 2. Participants in the second group were gamblers (ng = 6, one female) that were recruited via (1) an open call on social media which was shared by organizations and charities working on gambling awareness and responsible gambling, and (2) snowball sampling through the gamblers. The interviewed gamblers (4 online gamblers, 1 in-person only gambler, 1 gamer/gambler) were identified as individuals with gambling disorder in recovery and were interviewed to express their perception about the findings of the first group. Individuals with gambling disorder in recovery were specifically chosen as they unique could reflect back on their own gambling experiences and provide more insightful comments than those who are currently experiencing problems with gambling or those who are just casual or social gamblers. In total, the two groups comprised a sample of 19 participants (12 males, 7 females).

Table 1 Demographic information of experts

Table 2 Details of the participating experts

Data Collection

Approval for this project was granted by Bournemouth University Research Ethics Committee in the UK, and all participants provided written informed consent prior to their respective interviews.

The semi-structured interviews with experts lasted 40–70 minutes and were remotely conducted using teleconferencing services during August to September 2017 by one of the authors (GD). This author holds 12 years of experience in both computing and health care through working in this area in various European projects.

The semi-structured interviews with the gamblers lasted 1–2 hours and were conducted either face-to-face (2 of 6) or by teleconferencing (4 of 6), by another author, a health psychologist with experience in qualitative interviewing and web-based interventions (EAC).

All interviews were audio recorded and transcribed verbatim. The interviews with experts began with questions about their profile (i.e., education, expertise and experience in gambling). During the interviews, the research questions mentioned above were elaborated and exemplified with a wide range of data and techniques that are typically used in literature concerning software-assisted behaviour awareness and change. This approach was based on the project team and interviewer’s experience in the general areas of Digital Addition, Persuasive Technology and Behaviour Change, and was meant to aid interviewees with a basis for a more specialized discussion tailored to the domain of online gambling. Examples from previous research (Ali et al., 2015; Alrobai, Dogan, Phalp, & Ali, 2018; Alrobai, McAlaney, Phalp, & Ali, 2016b), as well as an explanatory video showing the architecture, were given to familiarize the interviewees with the overall architecture and processes of the solution proposed, e.g., how data can be collected and manipulated by users and their surrogate software for limit setting and regulated usage. In the interview induction phase all participants were shown that the authentication is executed through the usual log in to the operator site and that they are given full control over the transfer of the data to any additional tools or personnel and have the right to revoke any permission they have given at any time. The illustration was done though displaying a rich picture explaining the architecture and the data flow, as well as through a video we specifically developed to explain the whole process and its guiding policies and constraints. The conversation was also focused on individual interviewees’ expertise to maximize the quality of input. The interviews with gamblers (the second group) explored (1) their experience of gambling and (2) their views regarding the acceptability of the limits, intervention techniques and data suggested by the experts, which are presented here.

Data Analysis

Data analysis was carried out using content analysis and thematic analysis. Content analysis was especially appropriate for the interviews with experts as it remains the best technique in research for the identification and categorization of pertinent data points (i.e., RQ3: gambling operators’ data, multimodal sensors’ data) (Woodrum, 1984).

Data from the expert interviews in the first group were analysed using content analysis to identify particular techniques around each research question (RQ1–3) (Joffe & Yardley, 2004). As the topic is relatively new, the method allowed us to analyse the data with relative few assumptions and to be open to new types of gambling related data, rationale, and modalities of collection. It also allowed us to unify the relatively diverse terms used by the participants. The content analysis was mainly performed for data categorization, and with the ultimate aim of eliciting the types of online gambling behaviour data and other non-gambling related data. Moreover, it was also important for a holistic understanding of the gamblers’ status and behaviour. The findings were then sent to the experts to comment on and to debate over a four-week period through a shared online document.

Data from the interviews with gamblers in the second group were analysed using thematic analysis (Joffe & Yardley, 2004). Thematic analysis was a preferred option for the analysis of these interviews as it allowed the combination of deductive and inductive approaches to identification of themes (Fereday & Muir-Cochrane, 2006). In particular, the deductive a priori template of codes was developed based on the interviews with the experts (i.e., RQ1: ‘limit setting’ code-theme). However, we have not ignored the data-driven themes that provided us with the richness of descriptions and explanations (i.e., RQ2: feelings regarding the comparative visualization). For the purposes of this study, only content specifically relating to the research questions RQ1–3 was included. In the following section, the anonymized results of the interviews are presented based only on those questions. It is important to highlight that the data collection was driven by theoretical saturation. With the first group of participants (experts), this goal was achieved quickly, because the main points were mentioned by the majority of the participants, and also because it is typical for experts to provide factual and evidence-based statements. For the second group (gamblers), we interviewed a range of participants, leading us to conclude that persuasive techniques were broadly acceptable to gamblers, but we acknowledge that interviewing a wider sample could further refine our general findings and contextualize it more. Hence, the saturation here is up to the point where the collection of data and their utilization for responsible gambling was generally accepted in principle.

Results

Findings about Limit Setting (RQ1)

Interviews with Experts

Experts in psychology highlighted that limits that should be set to aid gamblers in regulating their online gambling activity would need to follow the SMART approach to goal setting (Doran, 1981). The limits were as follows:

  • Specific (simple, sensible, significant): target a specific area for improvement.
  • Measurable (meaningful, motivating): quantify or at least suggest an indicator of progress.
  • Assignable (agreed, attainable): specify who will do it.
  • Realistic (reasonable, results-based): state what results can realistically be achieved, given available resources.
  • Time-related (time-based, time-limited, time- or cost-limited, timely, time-sensitive): specify when the results shall be achieved.

Additionally, our experts mentioned that the individuals should set their own limits to increase their autonomy (as per Locke and Latham, 2004), but that each limit should consider the individuals’ expectations of reaching it and its value to them (as per Atkinson, 1964). For instance, the individuals should not set limits that are unachievable (e.g., quitting without any support in place) or too easy (e.g., never gambling more than £1 million in one bet), as they would be set to fail (if unachievable or too hard), or would not feel satisfaction at achieving this limit (if too easy). That is, limits offered by the platform should be realistic but challenging and could be based on betting history. Finally, our experts mentioned that individuals may find it difficult to set SMART goals (as per Yardley et al., 2013), so they could also be set in collaboration with therapists or family members.

Table summarises our findings from RQ1. These results were obtained through twelve of the thirteen interviewees. (One of the thirteen preferred instead to not answer because of lack of expertise.) We organized our findings in six groups. The first three groups present the subject of the limits that should be specified (money, time and access limits). The fourth group (“Who should set the limits?”) discusses which party should be the one to set the limits, while the fifth group (“limit duration”) explores the time by which a limit should be achieved, and the sixth group (“special considerations”) presents best practices on how limits should in fact be set.

Table 3 Findings about limit setting

Interviews with Gamblers

In relation to the findings in Table 3, the interviews with gamblers indicated that they were positive about setting money, time and access limits. They believed time limits would enable gamblers to stay in control, as certain of them reported gambling for over eight hours, and not wanting to stop until they had won:

That experience [playing with a particular operator] ... was very helpful because it forced me to take a break for a significant period. I think 24 hours ... The longer I am gambling, the less likely I am to make rational choices around my gambling and the more likely I am to gamble problematically and place stupid bet stakes, lose control basically. (Pg3, online gambler)

Similarly, money limits were understood as helpful in limiting damage. Participants believed that setting spending limits would reduce damage, as they understood that when they were losing they lost an important sense of logic.

It’s so dangerous to be allowed to gamble to the extent that I was allowed to. I had a £20,000 spin one night. (Pg2, online gambler)

Regarding access limits, participants also argued it was important to restrict the time between bets, in order for them to locate the time to take a meaningful pause:

You must equally take into consideration the time [between bets in online roulette] and stop it being 20 seconds and make it at least a minute if not 90 seconds. (Pg2, online gambler)

The interviewed gamblers believed time limits should be set by the platform, as they thought they would be unable to do this themselves. The subjects also believed that the platform should provide a guide to setting money limits, e.g., calculating disposable income based on occupation and income bracket. However, nominating a family member was not considered to be helpful for everyone, as many gamblers hide or deny information from their families, so this feature would in fact need to be voluntary.

Findings about Interactive Persuasive Interventions (RQ2)

Interviews with Experts

In this section, we summarize the results about the online interventions, corrective measures, visualization techniques, and infographics, all of which can be applied, based on the gambling behavioural data, to help the gambler to play responsibly. The results answering research question RQ2 are presented in Table 4 and detailed below.

Information for empowerment

This category includes information we need to display to the gamblers to empower them through graphs or any other forms of visualizations. Aesthetics is one of the key factors in enhancing engagement with web-based interventions, and visualizations are more aesthetically pleasing to users than pure writing (O’Brien & Toms, 2008).

Comparative information

This category included comparisons of gamblers’ activity with other multimodal data (e.g., emotions and locations) that could help them to understand their behaviour and change it. The category also incorporated comparisons of their gambling activity with gambling activities of others (Auer & Griffiths, 2015). Social norm theory suggests that gamblers are likely to underestimate how much they gamble relative to others, based on research around alcohol use in students (Perkins, 2002). Normative feedback—that is, feedback about regularity of gambling and amounts gambled relative to others—could help individuals to regulate their behaviour. This approach has been effective in reducing alcohol consumption among university students, across a variety of studies (e.g., Neighbors et al., 2016). A personalized normative feedback intervention led to reductions in gambling problems in university students (Neighbors et al., 2015).

Infographics about user’s level of gambling

This category included infographics that could make gamblers better understand their gambling behaviours. Such graphics, we argued, might clarify the nature of the information being provided, particularly to individuals of lower educational levels.

Notifications and messages

This category concerned different types of notifications and messages (in certain cases framing the situations as part of a game) according to gamblers’ limits, to inform gamblers about their achievements, encourage them to not play more, and instruct their families. Persuasive system design enhances adherence to web-based interventions (Kelders, Kok, Ossebaard, & Van Gemert-Pijnen, 2012).

Communication mediums

This category summarized how such notifications and messages could be communicated to the gamblers through different mediums. Using a range of mediums, preferably tailored to the user’s interest, this category, we reasoned, was likely to enhance adherence to the intervention.

Educational materials

This category includes educational materials that could be provided to gamblers as knowledge at appropriate times during betting. Such examples of knowledge are related to gambling consequences, how to reduce stress, games’ probabilities and the risks of addiction. Education is an essential part of interventions to reduce addictive behaviour, as knowledge is an essential first step in bringing about behaviour change. In the stages of change model (Prochaska, DiClemente, & Norcross, 1993), knowledge is required to move from precontemplation (no intention to change behaviour) to contemplation (intention to change behaviour within the next 6 months).

Special considerations

This category concerned thoughts about the appropriate selection of infographics and visualizations, as well as the timing of the provided notifications and messages.

Table 4 Findings about interactive persuasive interventions

Interviews with Gamblers

Gamblers reported that visualizations about their gambling activity would be helpful, as they would enhance awareness of gambling behaviour, possibly leading to further reflection:

Having a visual look of what I spent, it makes it real then, wow I didn’t realise I spent £500 a day for the past 2 weeks on [gambling operator’s] website. (Pg3, online gambler)

On the other hand, there were mixed feelings regarding comparative visualizations. While certain gamblers argued they would help raise awareness, others believed that a focus on the gambler as an individual would be more helpful, as others might in fact be experiencing different circumstances:

I wouldn’t really care what other people were gambling actually ... maybe they haven’t got enough time, maybe they’ve got plenty of money. (Pg6, online gambler)

Similarly, notifications and messages were received with mixed views. On the one hand, they were perceived as a way to enable change of focus, and therefore viewed positively:

It [a message] would really have been helpful at the time because anything that gives you a reason to switch your whole attention from what you’re doing. I could literally have been playing roulette and there could be a fire and I would have said, “Don’t worry, I’m not using the fire, I’m watching this screen here.” If you get a message whether it be oral or visual, it just distracts you. (Pg2, online gambler)

On the other hand, certain participants believed they would find pop ups annoying, and would instead be likely to click on them and then ignore them:

Similar to the pop-up messages that appear on fixed betting terminals ... they’re a *** nuisance. What I would do ... would be just switch them off ... I’m speaking from someone who ... when he’s gambling just wants to gamble, doesn’t want to be interfered with. (Pg3, online gambler)

However, the interviewed gamblers maintained it would be helpful to receive notifications from the platform if they appeared to be betting in an unusual manner. They particularly liked the idea of telephone calls, as they thought they would provide a personal touch. Providing emoticons as a method of giving feedback on betting activity (e.g., a smiley face if they had achieved their limits, a sad face if they had had a net loss) were not understood as helpful. Certain participants believed they would worsen low mood and contended they would trivialize the problem.

Findings about Interactive Persuasive Interventions (RQ3)

Interviews with Experts

In this section, we report the data needed to support limit setting and the different types of interventions with the aim of enabling more responsible online gambling. The results of the research question RQ3 are presented in Table 5. The resulting types of data mentioned were organized into the following groups.

Gambling operators’ data

This category contained data that were generated or recorded by the gambling operators’ platforms. Data concerning betting history, including real time data, could, we argued, enhance gamblers self-awareness regarding the pervasiveness of their behaviour. Problem gamblers tend to place higher confidence in their bets and believed they command greater control over their bets than non-problem gamblers (Goodie, 2005). Data showing gamblers how much they have been betting and winning or losing over a particular period may break their illusion of control, thus acting as a catalyst for behaviour change. Such information would be useful for individuals from when they start gambling.

Multimodal sensors’ data

This category consisted of data that were produced in the user-side and in gamblers’ personal digital devices (i.e., smartphones and sensors). These data could enable the platform to inform the gambler about their gambling behaviour in relation to location and time, to increase self-awareness of automatic behaviours (Banos et al., 2016). Information regarding how their behaviour varies across contexts (e.g., creating awareness of their differential gambling activity based on time and place) could also increase perceived control over gambling.

Web presence data

This category represents data describing the general online activities and behaviour of the gambler.

Third party data

This category includes data outside the boundary of the gambling operators, that can be collected with gamblers’ consent from third party systems, such as financial and health-related institutional systems. Data could also be used to facilitate limit setting. Providing third party data such as bank statements could facilitate the platform in setting SMART goal type limits for the individual (Locke, Shaw, Saari, & Latham, 1981). Self-reported information, such as the gamblers’ financial situation, could enable the platform to set the gambler money limits.

Self-reported data

This category contained data that can be reported by the gamblers themselves using appropriate forms spontaneously, after an event or in a specific time frequency. Data about the gambler’s emotional state, daily activities, or both could enable the platform, over time, to determine when the gambler is likely to carry out problem gambling. In these situations, the platform could inform the gambler via instant messaging, and possibly suggest alternative activities, to enhance self-awareness and break the habit (Banos et al., 2016).

Self-administered measures

This category consisted of self-administered measurement and tasks that, when completed, can provide indicators and quantification of gambling addiction and psychological status of a gambler. Such questionnaires could further increase gamblers’ awareness of their behaviour.

Table 5 Findings about the relevant gambling behavioural data

Interviews with Gamblers

Gamblers believed it would be important to collect data across gambling operators, as many used a range of websites. Data from multimodal sensors (e.g., regarding location, emotion, stress) was perceived as helpful by certain gamblers, as they thought it might facilitate the platform in detecting potential issues:

I can only see it [app sensing gambler’s location] as a positive especially if somebody’s got a problem. (Pg1, online gambler)

On the other hand, certain gamblers contended that the platform having access to this level of information about them would be too intrusive. The platform having access to third party data (such as bank statements) was, however, perceived as having the potential to be helpful, if provided with consent:

I know people, through GA, who keep track of what they spend and what they’ve done and can prove they haven’t gambled and have spreadsheets and all sorts ... It works for them. (Pg1, online gambler)

Self-report data about emotions were also understood to be of use. Participants mentioned being more likely to gamble when they were depressed or had had what they perceived as a bad day. They concluded this information would aid the platform in gaining knowledge of their behaviour. They also argued that this information would be helpful to look back on in the future to maintain control:

It’s nice to have a record of how bad it [binge] was because, sometimes, I’m reading through my journal and it can motivate me to stop or it can motivate me to stay stopped because I can just forget how bad these binges were. (Pg4, gamer)

Finally, questionnaires to assess gambling activity were also perceived as helpful, particularly for those new to gambling. However, participants were concerned filling in questionnaires would take quite a lot of time, which gamblers might feel would be better spent gambling. Incentives such as prize draws were suggested as a possible solution to this problem.

Discussion

This paper aimed to identify the limits, potential interventions, and types of data that could help online gamblers maintain control of their gambling via a self-regulation platform. For this reason, in our study we interviewed two groups: experts and problem gamblers.

Our interviews with experts identified three types of limits (money, time and access) and identified areas to consider such as who should set these limits and their duration. We also identified seven areas to consider when designing interventions (such as visualization of the gambling data) and six areas regarding the type of data collected (such as betting history and location).

Our interviews with gamblers revealed mixed and sometimes contradictory views. For instance, whereas certain of the gamblers concluded that obtaining multimodal data (e.g., information on geolocation, heart rate, and emotions) could be useful to predict when someone is likely to gamble, other gamblers instead found that this process would prove too intrusive. More research is needed to investigate which approaches would work best for which specific individuals. We recommend in-depth interviews with service users before the development of interventions. However, gamblers did mention the importance of any intervention having a range of tools to manage responsible gambling, as they had used a range of strategies in recovering from their addiction.

Usability is not the only requirement for self-regulation applications that process the data and help gamblers stay in control. A wide range of other human factors seem to be prominent. For example, reactance (Miron & Brehm, 2006) could be one of the issues when gamblers believe that their freedom to take decisions has been compromised. When the data is used in comparison setting (that is, among groups of players) trivialization and normalization can happen. An example of that process is when one amount of expenditure compares well to others despite the different in affordability. The risk also exists of technology being understood as a remedy rather than an assistance and this may lead to players blaming the software for not curing them. Most of these risks have been studied in previous research in the general context of digital addiction (Alrobai, McAlaney, Phalp, & Ali, 2016a).

Additionally, there are several arguments about the power and risks associated with self-regulation mediated by technology. We still do not possess strong scientific evidence of their effectiveness and, in particular, the sustainability of change that they can bring (Leigh & Flatt, 2015). Delivering interventions within peer group settings could possibly be harmful because of group dynamics and structure factors. This problem may in turn lead to reinforcing negative behaviour (Dishion, McCord, & Poulin, 1999), such as social loafing and compensation (Karau & Williams, 1993), along with conformity effects (Allen, 1965). Persuasive technologies may cause people to feel frustrated, anxious, pressured by peers, and guilty when they do not comply with the system or have to deceive (Hamari, Koivisto, & Pakkanen, 2014). Despite the potential opportunities of using online gambling behavioural data to help gamblers regulate their gambling, caution is required, as the possibility exists that the change may go in unforeseen directions.

According to Hing, Russell, and Hronis (2016), a responsible gambling concept involves responsible provision of gambling and responsible consumption. This in turn places responsibility for duty of care in hands of all players within the gambling industry, operators and customers. It is clear that transparency on data sharing is heavily imposed on gambling operators; however, accountability for data sharing is not fully in place (Bachmann, Gillespie, & Priem, 2015). According to Bachmann et al. (2015) accountability is paramount to building trust between customers and organizations. The principles of data sharing and modalities for persuasive interactions proposed in this paper, such as requiring gambling operators to interact and share data within their various divisions and with the gamblers, compose the first steps to building transparent and accountable data sharing. This sort of transparency can maximize compliance with the European Union General Data Protection Regulation (GDPR). It can also enable socially responsible practices across the gambling industry that will effectively lead to more trust.

To conclude, we hope this paper will stimulate discussions not only in the gambling industry but also in the software and well-being industries, as well as policy makers, to develop strategies towards more responsible gambling. It is also our aspiration that the results of this qualitative study will constitute both a meaningful basis and set of reference points for future self-regulation information systems, systems that should be developed following the principles of security and privacy by design (Vicini et al., 2016), and with the ultimate goal the empowerment of responsible gambling through the capture and utilization of gambling behavioural data.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. https://doi:10.1016/0749-5978(91)90020-T

Ali, R., Jiang, N., Phalp, K., Muir, S., & McAlaney, J. (2015). The emerging requirement for digital addiction labels. In Fricker S. A & Schneider K (Eds.), Requirements engineering: Foundation for software quality (LNCS Vol. 9013, pp. 198–213). Cham, Switzerland: Springer International Publishing. https://doi:10.1007/978-3-319-16101-3_13

Allen, V. L. (1965). Situational factors in conformity. Advances in Experimental Social Psychology, 2, 133–175. https://doi:10.1016/S0065-2601(08)60105-7

Alrobai, A., Dogan, H., Phalp, K., & Ali, R. (2018). Building online platforms for peer support groups as a persuasive behavior change technique. In Ham J, Karapanos E, Morita P. P, & Burns C. M (Eds.), Persuasive technology (LNCS Vol. 10809, pp. 70–83). Cham, Switzerland: Springer International Publishing. https://doi:10.1007/978-3-319-78978-1_6

Alrobai, A., McAlaney, J., Phalp, K., & Ali, R. (2016a). Exploring the risk factors of interactive e-health interventions for digital addiction. International Journal of Sociotechnology and Knowledge Development, 8, 1–15. https://doi:10.4018/IJSKD.2016040101

Alrobai, A., McAlaney, J., Phalp, K., & Ali, R. (2016b). Online peer groups as a persuasive tool to combat digital addiction. In Meschtscherjakov A, De Ruyter B, Fuchsberger V, Murer M, & Tscheligi M (Eds.), Persuasive technology (LNCS Vol. 9638, pp. 288–300). Cham, Switzerland: Springer International Publishing. https://doi:10.1007/978-3-319-31510-2_25

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (DSM5) (5th ed.). Washington, DC: Author. https://doi:10.1176/appi.books.9780890425596

Armstrong, T., Donaldson, P., Langham, E., Rockloff, M., & Browne, M. (2018). Exploring the effectiveness of an intelligent messages framework for developing warning messages to reduce gambling intensity. Journal of Gambling Issues, 38, 67–84. https://doi:10.4309/jgi.2018.38.4

Armstrong, T., Rockloff, M., Browne, M., & Blaszczynski, A. (2019). Development and validation of the Protective Gambling Beliefs Scale (PGBS). International Gambling Studies, 19, 36–53. https://doi:10.1080/14459795.2018.1500624

Atkinson, J. W. (1964). An introduction to motivation. Oxford, UK: Van Nostrand.

Auer, M. M., & Griffiths, M. D. (2015, March 23). Testing normative and self-appraisal feedback in an online slot-machine pop-up in a real-world setting. Frontiers in Psychology, 6, 339. https://doi:10.3389/fpsyg.2015.00339

Bachmann, R., Gillespie, N., & Priem, R. (2015). Repairing trust in organizations and institutions: Toward a conceptual framework. Organization Studies, 36, 1123–1142. https://doi:10.1177/0170840615599334

Bagby, R. M., Taylor, G. J., & Ryan, D. (1986). Toronto alexithymia scale: Relationship with personality and psychopathology measures. Psychotherapy and Psychosomatics, 45, 207–215. doi:10.1159/000287950

Banos, O., Villalonga, C., Bang, J., Hur, T., Kang, D., Park, S., … Lee, S. (2016). Human behavior analysis by means of multimodal context mining. Sensors, 16. https://doi:10.3390/s16081264

Baumeister, R. F., Schmeichel, B. J., & Vohs, K. D. (2007). Self-regulation and the executive function: The self as controlling agent. In Kruglanski A. W & Higgins E. T (Eds.), Social psychology: Handbook of basic principles (2nd ed., pp. 516–539). New York, NY: Guilford Press.

Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity to future consequences following damage to human prefrontal cortex. Cognition, 50, 7–15. https://doi:10.1016/0010-0277(94)90018-3

Chagas, B., & Gomes, J. (2017). Internet gambling: A critical review of behavioural tracking research. Journal of Gambling Issues, 36. https://doi:10.4309/jgi.2017.36.1

Cialdini, R. B. (2006). Influence: The psychology of persuasion (Rev. ed.). New York, NY: Harper Business.

Davidson, L., Chinman, M., Kloos, B., Weingarten, R., Stayner, D., & Tebes, J. K. (1999). Peer support among individuals with severe mental illness: A review of the evidence. Clinical Psychology: Science and Practice, 6, 165–187. https://doi:10.1093/clipsy.6.2.165

Dishion, T. J., McCord, J., & Poulin, F. (1999). When interventions harm: Peer groups and problem behavior. American Psychologist, 54, 755–764. https://doi:10.1037/0003-066x.54.9.755

Doran, G. T. (1981). Theres a SMART way to write managements goals and objectives. Management Review, 70, 35–36.

Fenner, A. A., Straker, L. M., Davis, M. C., & Hagger, M. S. (2013). Theoretical underpinnings of a need-supportive intervention to address sustained healthy lifestyle changes in overweight and obese adolescents. Psychology of Sport and Exercise, 14, 819–829. doi:10.1016/j.psychsport.2013.06.005

Fereday, J., & Muir-Cochrane, E. (2006). Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International Journal of Qualitative Methods, 5, 80–92. https://doi:10.1177/160940690600500107

Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. https://doi:10.1177/001872675400700202

Gambling Commission. (2016, November 26). New figures show online gambling is largest gambling sector in Britain. Retrieved from: http://www.gamblingcommission.gov.uk/news-action-and-statistics/news/2016/New-figures-show-online-gambling-is-largest-gambling-sector-in-Britain.aspx

Goodie, A. S. (2005). The role of perceived control and overconfidence in pathological gambling. Journal of Gambling Studies, 21, 481–502. https://doi:10.1007/s10899-005-5559-1

Gratz, K. L., & Roemer, L. (2004, March 1). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26, 41–54. https://doi:10.1023/B:JOBA.0000007455.08539.94

Hagger, M. S., & Luszczynska, A. (2014). Implementation intention and action planning interventions in health contexts: State of the research and proposals for the way forward. Applied Psychology: Health and Well-Being, 6, 1–47. https://doi:10.1111/aphw.12017

Hamari, J., Koivisto, J., & Pakkanen, T. (2014). Do persuasive technologies persuade? A review of empirical studies. In Spagnolli A, Chittaro L, & Gamberini L (Eds.), Persuasive technology (LNCS Vol. 8462, pp. 118–136). Cham, Switzerland: Springer International Publishing. doi:10.1007/978-3-319-07127-5_11

Hing, N., Russell, A., & Hronis, A. (2016). Behavioural indicators of responsible gambling consumption. Melbourne, Australia: Victorian Responsible Gambling Foundation.

Joffe, H., & Yardley, L. (2004). Content and thematic analysis. In Marks D. F & Yardley L (Eds.), Research methods for clinical and health psychology (pp. 56–68). London, UK: SAGE.

Karau, S. J., & Williams, K. D. (1993, 10). Social loafing: A meta-analytic review and theoretical integration. Journal of Personality and Social Psychology, 65, 681–706. https://doi:10.1037/0022-3514.65.4.681

Kelders, S. M., Kok, R. N., Ossebaard, H. C., & Van Gemert-Pijnen, J. E. (2012). Persuasive system design does matter: A systematic review of adherence to web-based interventions. Journal of Medical Internet Research, 14, e152. https://doi:10.2196/jmir.2104

Kostoulas, T., Chanel, G., Muszynski, M., Lombardo, P., & Pun, T. (2017). Films, affective computing and aesthetic experience: Identifying emotional and aesthetic highlights from multimodal signals in a social setting. Frontiers in ICT, 4, 11. https://doi:10.3389/fict.2017.00011

Leigh, S., & Flatt, S. (2015). App-based psychological interventions: Friend or foe? Evidence-Based Mental Health, 18. https://doi:10.1136/eb-2015-102203

Locke, E. A., & Latham, G. P. (2004). What should we do about motivation theory Six recommendations for the twenty-first century. Academy of Management Review, 29, 388–403. https://doi:10.5465/amr.2004.13670974

Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and task performance: 1969–1980. Psychological Bulletin, 90, 125–152. https://doi:10.1037/0033-2909.90.1.125

Miller, M. L., & Thayer, J. F. (1988). On the nature of self-monitoring: Relationships with adjustment and identity. Personality and Social Psychology Bulletin, 14, 544–553. https://doi:10.1177/0146167288143013

Miron, A. M., & Brehm, J. W. (2006). Reactance theory: 40 years later. Zeitschrift für Sozialpsychologie, 37, 9–18. https://doi:10.1024/0044-3514.37.1.9

Moos, R. H., & Moos, B. S. (2004). Help-seeking careers: Connections between participation in professional treatment and alcoholics anonymous. Journal of Substance Abuse Treatment, 26, 167–173. https://doi:10.1016/S0740-5472(03)00190-9

Neighbors, C., Lewis, M. A., LaBrie, J., DiBello, A. M., Young, C. M., Rinker, D. V., . . . others (2016). A multisite randomized trial of normative feedback for heavy drinking: Social comparison versus social comparison plus correction of normative misperceptions. Journal of Consulting and Clinical Psychology, 84, 238–247. https://doi:10.1037/ccp0000067

Neighbors, C., Rodriguez, L. M., Rinker, D. V., Gonzales, R. G., Agana, M., Tackett, J. L., & Foster, D. W. (2015). Efficacy of personalized normative feedback as a brief intervention for college student gambling: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 83, 500–511. https://doi:10.1037/a0039125

O’Brien, H. L., & Toms, E. G. (2008). What is user engagement A conceptual framework for defining user engagement with technology. Journal of the American Society for Information Science and Technology, 59, 938–955. https://doi:10.1002/asi.20801

Oinas-Kukkonen, H., & Harjumaa, M. (2009). Persuasive systems design: Key issues, process model, and system features. Communications of the Association for Information Systems, 24, 485–500. https://doi:10.17705/1CAIS.02428

Perkins, H. W. (2002). Social norms and the prevention of alcohol misuse in collegiate contexts. Journal of Studies on Alcohol, 2002 (s14), 164–172. https://doi:10.15288/jsas.2002.s14.164

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1993). In search of how people change: Applications to addictive behaviors. Addictions Nursing Network, 5, 2–16. https://doi:10.3109/10884609309149692

Rodda, S. N., Dowling, N. A., Knaebe, B., & Lubman, D. I. (2018). Does SMS improve gambling outcomes over and above access to other e-mental health supports A feasibility study. International Gambling Studies, 18, 343–357. https://doi:10.1080/14459795.2017.1388831

Scott-Sheldon, L. A. J., Carey, K. B., Elliott, J. C., Garey, L., & Carey, M. P. (2014). Efficacy of alcohol interventions for first-year college students: A meta-analytic review of randomized controlled trials. Journal of Consulting and Clinical Psychology, 82, 177–188. https://doi:10.1037/a0035192

Stinchfield, R., Govoni, R., & Frisch, G. R. (2007). A review of screening and assessment instruments for problem and pathological gambling. In Smith G, Hodgins D. C, & Williams R. J (Eds.), Research and measurement issues in gambling studies (pp. 179–213). Burlington, MA: Elsevier. https://doi:10.1002/9781118316078.ch8

Tabri, N., Hollingshead, S. J., & Wohl, M. J. A. (2019). A limit approaching pop-up message reduces gambling expenditures, except among players with a financially focused self-concept. International Gambling Studies, 19, 1–12. https://doi:10.1080/14459795.2019.1567806

Vicini, S., Alberti, F., Notario, N., Crespo, A., Pastoriza, J. R. T., & Sanna, A. (2016). Cocreating security-and-privacy-by-design systems. In 11th International Conference on Availability, Reliability and Security (ARES) (pp. 768–775). New York, NY: IEEE, . https://doi:10.1109/ARES.2016.74

Webb, T. L., Sniehotta, F. F., & Michie, S. (2010). Using theories of behaviour change to inform interventions for addictive behaviours. Addiction, 105, 1879–1892. https://doi:10.1111/j.1360-0443.2010.03028.x

Whitlock, E. P., Polen, M. R., Green, C. A., Orleans, T., & Klein, J. (2004). Behavioral counseling interventions in primary care to reduce risky/harmful alcohol use by adults: A summary of the evidence for the U.S. Preventive Services Task Force. Annals of Internal Medicine, 140, 557–568. https://doi:10.7326/0003-4819-140-7-200404060-00017

Woodrum, E. (1984). “Mainstreaming” content analysis in social science: Methodological advantages, obstacles, and solutions. Social Science Research, 13, 1–19. https://doi:10.1016/0049-089X(84)90001-2

Yardley, L., Morrison, L. G., Andreou, P., Joseph, J., & Little, P. (2010). Understanding reactions to an internet-delivered health-care intervention: A accommodating user preferences for information provision. BMC Medical Informatics and Decision Making, 10, 52. https://doi:10.1186/1472-6947-10-52

Yardley, L., Williams, S., Bradbury, K., Garip, G., Renouf, S., Ware, L., Little, P. (2013). Integrating user perspectives into the development of a web-based weight management intervention. Clinical Obesity, 2, 132–141. https://doi:10.1111/cob.12001

Zhao, Y., Marchica, L., Derevensky, J. L., & Ivoska, W. (2018). Mobile gambling among youth: A warning sign for problem gambling? Journal of Gambling Issues, 38, 268–282. https://doi:10.4309/jgi.2018.38.14

*******




Submitted December 25, 2019; accepted January 28, 2020. This article was peer reviewed. All URLs were available at the time of submission.

For correspondence: George Drosatos, PhD, Institute for Language and Speech Processing, Athena Research Centre, P.O. Box 159, University Campus Kimmeria, Xanthi 67100, Greece. E-mail: gdrosato@athenarc.gr

Competing interests: None declared (all authors).

Ethics approval: The Bournemouth University Research Ethics Committee approved on June 5, 2017 the EROGamb 1.0 project with title “Empowering Responsible Online Gambling with Predictive, Real-Time, Persuasive and Interactive Intervention” (approval no. 16022) and on June 3, 2019 the GamInnovate project with title “Evaluating Interactive Persuasive Technology for Gambling Awareness” (approval no. 26568).

Acknowledgements: This work has been partially supported by GambleAware (UK) through the EROGamb 1.0 project and the International Center for Responsible Gaming (ICRG, US) through the GamInnovate project. The funding bodies have had no involvement with any part of the research. The authors alone are responsible for the views expressed in this article.

Article Categories:
  • Original Research

Related Article(s):


Copyright © 2020 | Centre for Addiction and Mental Health
Editor-in-chief: Nigel E. Turner, Ph.D.
Managing Editor: Vivien Rekkas, Ph.D. (contact)