For the past couple of days I have been attending the Bitish Psychological Society’s Cyberpsychology Section Virtual Conference. My colleagues and I are well represented, with four talks from projects that I worked on (two by me, and one each by my colleagues Elvira Perez Vallejos and Virginia Portillo) and two others from colleagues I have the great fortune to be working on other projects with (Mat Rawsthorne and Camilla Babbage).
On the first day, I presented our work from the ReEnTrust project, on attitudes towards online wellbeing and trust in younger and older adults. The video can be seen here. One of the major aims of the ReEnTrust project was to identify the most important issues that effect trust in users’ online service interactions, and how these interactions affect wellbeing. Our work package related especially to how attitudes and experiences of these issues differed across younger and older adults. We carried out a series of 3 hour workshops with 2 age groups: 4 workshops with 35 young people aged 16-25 year olds, and 5 workshops with 40 older adults aged 65 and over. As part of these workshops participants completed pre- and post-session questionnaires focusing on trust and wellbeing, as well as digital literacy. We measured both the fulfilment of the basic psychological needs for autonomy, competence, and relatedness, to examine eudamonia, or the experience of purpose in life, and the other looking at subjective wellbeing, measuring experiences of positive and negative emotions online. High levels of need fulfilment and high levels of positive affect lead to a fulfilled life, or what may be termed ‘flourishing’. We also asked participants to rate statements related to their trust in the internet, and how important trust is when online, and measured what we called ‘digital confidence’ using a 6 item scale, which aimed to get users to rate their own online digital literacy.
Whilst both groups did experience considerable benefits of being online, and recognised the potential for both positive and negative effects of the online world, young people are more concerned about the wellbeing effects of being online. Older adults seemed to focus more on the positives, including increased ability to communicate with friends and family, and opportunities to take part in things that they could not do offline. However, some older adults did mention a concern for others who they perceive as using the internet more, especially younger adults and teenagers. Negative factors for young people often surrounded the type of content they saw, the potential for negative social comparison, and a lack of control over their information. Lower levels of autonomy among young people were also related to higher negative affect, but not for older adults.
Older adults more explicitly related negative experiences to a feeling of being overwhelmed and a lack of competence. Indeed, competence online and digital confidence were the major differences between the two groups, with older adults being more adversely affected. Young people had both higher digital confidence and competence fulfilment than older adults; higher levels of competence fulfiment in young people and old peoples’ digital confidence were related to lower levels of negative affect. Older adults were more bothered by their own (perceived) lack of understanding and this related both to their trust in websites and their sense of wellbeing; they also placed more importance on trust than young people did.
Whilst there were interesting differences between the two age groups, there were also striking similarities in how young people and older adults consider and experience their wellbeing when online. Overall the results suggest that both young people and older adults experience moderate levels of wellbeing and need satisfaction. Both groups have similar levels of autonomy and relatedness satisfaction, and encounter similar levels of positive and negative experiences online, with the positive slightly outweighing the negative. Being online has the potential to satisfy basic psychological needs and contribute to human flourishing, however both groups also talked about stress, anxiety and pressure, as well as the time consuming nature of being online. Increased autonomy was related to higher levels of positive affect, lower negative affect, and increased relatedness, but the study revealed relatively low scores for autonomy in both groups. Many users’ speak of a sense of lack of control when they are online and this needs to be addressed.
In terms of trust, both groups only had moderate levels of trust in the online world. Although older adults place more importance on trust online, both groups felt quite strongly that websites have a responsibility to act in a trustworthy manner towards their users, and that websites do not do enough to ensure this. Both groups related trust back to familiarity, reputation, safety and security, and data issues, often referring to the design and content of the websites they use. Whilst these experiences of trust bared little relationship to their levels of basic needs or subjective wellbeing, their responses did resonate with concerns about autonomy, competence and relatedness. This suggests that basic psychological needs are a useful lens through with to understand the experiences of internet users, and to frame discussions of wellbeing. More work needs to be done to relate this directly to measuring online wellbeing and trust, to ensure that the future design of platforms enhances the human experience and allows people to flourish.
Also on the first day, my colleage Elvira discussed results from our UnBias project looking at impact of algorithmic decision-making processes on young people’s well-being. Algorithms rule online environments and are essential for performing data processing, filtering, personalisation and other tasks. The algorithms that govern online platforms are often obfuscated by a lack of transparency in their online T&C and user agreements. This lack of transparency speaks to the need for protecting the most vulnerable users from potential online harms. Little attention has been given to children and young people’s experiences of algorithmically-mediated online platforms, or the impact of them on their mental health and well-being, despite one third of internet users being children below the age of 18. ‘Youth juries’ are youth-led interactive sessions that encourage participants to share and discuss their personal experiences and opinions of the online world in a safe space. We carried out a series of youth juries with a total of 260 children and young people (13-17 years old) to bring their opinions to the forefront, and elicit discussion of their experiences of using online platforms and perceived psychosocial impact on users. Perceived benefits include convenience, entertainment and personalised search results. Negative aspects include participants’ concerns for their privacy, safety and trust when online. We recommend that online platforms acknowledge and enact on their responsibility to protect the privacy of their young users, recognising significant developmental milestones, and the impact that technology may have on young users. We argue that governments need to incorporate policies that require technologists and others to embed the safeguarding of users’ well-being within the core of the design of Internet products and services to improve the user experiences and psychological well-being of all, but especially those of children and young people.
On Day Two I presented our ReEnTrust work on online trust amongst older adults. The video is here. Despite the increases in the number of older adults over 65 years old using the internet, this group are often neglected from these discussions. We therefore set out to explore the factors that affected the online trust of older adults. We draw on data from a total of 40 participants across five workshops with adults aged 65 years and over. Co-created scenarios based on everyday online tasks – online shopping and seeking information – were used to facilitate discussion about trust on the internet. For each scenario they were asked to identify points that they felt were related to trust, whether they were positive or negative. Specifically, they were asked: “What are the most important factors related to trust here? How do you feel about it? How do you respond when this happens? And What do you think websites should do about it?”
Reputation was often highlighted, linked to intertwining factors including recognition of the brand, being a well-established company, and being a platform that they had used previously. For some, brand reputation was also related to having real-world connection such as a familiar bricks-and-mortar store. The platform’s reputation appears to be a protective factor for the user which offers security because of the platform’s need to nurture a good reputation.
Participants often reported difficulties in understanding how familiar behaviours in the real world might be translated online, for example in how search results are produced versus looking something up in the Yellow Pages. Participants also felt that platforms were being dishonest by obfuscating information relating to their business, such as hiding data collection behind cookie notifications, or simply not making clear where a company was based (having products appear weeks later from China was frustrating!). Activities such as profiling, tracking and surveillance also strongly impacted user’s trust online. Whilst some found profiling useful in terms of recommendations, often they had frustrations about inaccurate assumptions, repetitive advertising, and concerns about being placed in a ‘filter bubble’. Profiling of users led to a sense of losing privacy, with participants relating their experiences to being watched by ‘Big Brother’. They also had concerns about the extent to which they were being tracked online, and how others might find information about them, especially on social media, although some felt that they had little to hide and if they weren’t doing anything ‘dodgy’ it was not a problem.
Another common concern was that the internet is simply not safe to use. Many outlined various protective strategies, including looking for visual indicators of security, such as the padlock icon in web addresses. There were also a lot of concerns about a lack of control of what they were shown online, and in their choices of what websites and services to use. They often felt forced to do things like accept cookies, create user accounts, or accept permissions when downloading apps. They also felt a power imbalance due to the dominance of just a few companies, and participants often highlighted that they often felt compelled to abandon their personal values for the convenience offered by using them. For example, Amazon was often referred to, including in terms of benefits such as being able to rely on established policies and procedures, but also in terms of concern about the platform’s dominance and especially what they saw as dubious ethical practices. This raises the question of whether this represents an abandonment of trust in favour of convenience.
We recommend that the concept of trust, or more importantly, trustworthiness, is incorporated into the design of products, technologies and services to build user confidence and increase the wellbeing of users. To do this, different user groups must be consulted and involved from the very start of the design stages. This is especially important for groups who are not traditionally seen as online users, and may become neglected. The over 65s are often such a group in online research and our research has shed light on their online experiences.
My colleague Virginia also presented work from our work in UnBias: “Transparency in the age of Big Data: What do children want to know?” In the UK 99% of 12-15 year olds are online. Despite this, children’s voices have been often overlooked when making recommendations about the lack of transparency in data management (collection and usage) by online platforms. We explored children’s experiences of interacting with online services that shape their lives, in particular recommender systems (Google, YouTube, Netflix, etc.), and their ideas for a more fair and transparent online environment, through the youth juries I already described above. Recommendations predominantly revolved around how platforms use the information they collect from users, in particular a desire to be informed with what is collected from them, who is using it, and why. Participants also highlighted the benefits and barriers the Internet has on their lives and the importance of education to allow users to understand how the online world works. Children also wanted more choice and control of how their data is used. Meaningful transparency and education is required to allow people to reflect, question and develop their own ideas on key issues related to Internet technologies, and regulation to ensure transparency is both meaningful and maintained.
Mat presented “Prototyping an unobtrusive measure of online psychological flexibility in a moderated mental health peer support forum”. His abstract:
Using data from the REBOOT study (RCT of the Effectiveness of Big White Wall Compared to Other Online Support), explore the potential of analysing language used by contributors in internet support groups to gauge their ability to respond to new circumstances and possibly predict outcome
Design/Background: Current methods for analysing online conversations are labour intensive and automated linguistic inquiry methods utilising key word counts and collocations do not scale to provide the full context (and therefore accurate meaning) of concordances. However, computer science advances in these areas often not informed by psychology. Self-report measures for digital mental health are prone to bias so unobtrusive techniques may enable triangulation.
Methods: Applying Natural Language Processing of items from relevant clinical questionnaires to bootstrap the training of an algorithm to classify statements by how people relate to themselves and others (informed by Relational Frame Theory account of empathy and perspective-taking, and mechanisms of social comparison). Create a collaborative machine learning model to incorporate human expertise to refine its ability to label different types of post and test relationship with outcomes. Combining clinical knowledge and service user lived experience of anxiety and depression to assess and improve the face validity and transparency of the categorisation decisions.
Analysis: Classifier accuracy (area under the curve, confusion matrix) and comparison with non-posting participants in both active and control arms where outcome data exists
Conclusions Expected Implications: Initial assessment of whether non-professional conversational processes can be linked to wellbeing, and therefore whether there are types of interaction moderators should monitor and encourage.
Camilla presented “Developing an app to improve wellbeing for young people with Tourette Syndrome: Interviews with young people and professionals”. Her abstract:
Many young people with neurodevelopmental disorders who show reduced quality of life, will also experience co-occurring emotional and behavioural difficulties. Young people with Tourette Syndrome (TS) which causes involuntary tics, report emotional dysfunction to be more impairing than their tics. Digital self-help interventions targeting mood management are effective for young people, recommended in guidelines, and could combat resource deficiencies. Currently no such intervention exists, therefore the aim of this research is to explore what young people and professionals would desire and consider useful in a wellbeing app for young people with TS.
Design/Background: The study design included semi-structured interviews analysed using thematic analysis. Methods: 15 young people aged 9-17 with TS and Tic Disorders were interviewed via video-call or face-to-face. 16 professionals with an average 9 years TS work-experience were interviewed face-to-face, by video-call or phone call.
Results: Both samples derived themes that centred on desired features of the app. For young people this included psychoeducational and reminder functions, and calming elements like music and games. Professional themes highlighted a need for features of the app that would facilitate the use of tic and mood-management strategies, help young people to plan ahead and for family psychoeducation of TS.
Conclusions: Professional and young people showed overlaps and differences in themes relevant to features desired in the app. In order to develop wellbeing apps that are both engaging and effective for young people with neurodevelopmental disorders, including both perspectives is important.
Great work everyone!