Listening to young people online

Mental health and social media

By Tom Harrison @_Tom_Harrison


Introduction

Social media and digital technologies are now an integral part of the lives of young people in the UK. A recent Royal Society for Public Health report confirmed this, finding that 99 percent of 12–15-year-olds are active online for over 21 hours a week with 91 percent of them using social networking sites.

The fast pace of digital change reflects Eddie Obeng’s scenario in World After Midnight, which suggests that the pace of change in contemporary society has outstripped our ability to understand the effect of that change while it is occurring.

Indeed, we know that 75 percent of mental illnesses are diagnosed before the age of 25 and there has been wide public discussion surrounding the effect social media is having amongst young people with reports linking overexposure to social media with emotional distress and mental illness.

Alongside a greater emphasis on supporting young people to account for this, there has been an intense focus on the correlation of the two, with Simon Stevens, Chief Executive of NHS England, recently urging technology companies to account for the “epidemic of mental illness” that they share accountability for.

Research continues to build the evidence base with a systematic review conducted by the Education Policy Institute finding a “clear association” between prolonged exposure times spent on social media and mental health problems. ONS data suggests children who spend three hours or more on social media on a normal school day are twice (27%) as likely to report mental ill health than those spending no time online (12%). A PISA wellbeing study found that ‘extreme internet users’ were more likely to report bullying than ‘moderate internet users’.

Responding to this, the RSA Lab embarked on a social media ‘listening exercise’, analysing the online conversations of young people between July 2017 and May 2018 to better understand how young people talk about mental health issues online.

The exercise, which analysed the conversations of young people online using key words associated with mental health on platforms such as Facebook, Twitter, Instagram and through various online forums, aimed to build upon the existing evidence base and highlight gaps in our understanding of the digital worlds young people are spending more and more of their lives operating in. Based on this, we offer new ideas and approaches that public services, in collaboration with technology platforms and third sector mental health service providers, can pilot.

During the exercise there has been a growing recognition of the issues we found online. Commenting on a recent cover story in The Times Dr Jon Goldin, vice-chairman of the child and adolescent faculty at the Royal College of Psychiatrists spoke of the “completely unhelpful” groups online that encourage young people to self-harm which was reflected in the imagery of some posts the research found.

The listening exercise identified various themes and topics of conversation amongst young people; imagery associated with mental illness online; demographic splits within people under 25; times within a typical week when conversations are most prevalent; and areas where services can improve engagement with young people. We listened in more detail to some conversations young people were having in Greater Manchester to give us a more localised picture.

Greater Manchester was selected for several reasons, the most prominent of which was Manchester Resilience Hub, set up after the Arena bombing in 2017, which has been highly valued as one of the best services of its kind in the country, serving anyone affected by the bombing. The Hub has demonstrated, as recognised by the Kerslake Review,the benefits of good communication and engagement techniques to identify those who may need their service, particularly young people.


Summary

We found examples of peer-support and knowledge sharing amongst young people but that access to quality advice and expertise was very varied. This mirrors a worrying trend that the RSA has previously explored. Our report Getting the message on mental health highlighted that only 34 percent of people using mental health services in their local authority were able to access information and advice provided by organisations such as Healthwatch or Citizens Advice, exacerbating long waiting times and later diagnosis through lack of credible advice.

This briefing reflects on the progress that anti-stigma and mental health awareness-raising campaigns have made to encourage open and honest conversations about mental health unimaginable to previous generations. No longer a synonym for weakness, we should embrace the increasing confidence amongst young people addressing these important issues.

However, the need to tackle a lack of timely, credible and impartial advice on offer to young people, especially who those that might be presenting and talking about their mental health for the very first time online, is urgent and necessary.

Whilst services need to more effectively engage in dialogue with young people talking about and experiencing mental health difficulties, the aim of this engagement should be to increase the provision of reliable information and improve the quality of offline interventions and support.

New responses for statutory and community services in providing relevant and timely support for young people online are more important than ever and an online service with a focus on responding to young people’s concerns and conversations about mental health online, working to a similar aim as the NHS funded 111 telephone service, is worth serious exploration.

We believe that, not only will this help to improve access to appropriate referrals to services, but also improve the quality of information about mental health, especially the information shared between young people online. We hope too that by advocating an approach that champions listening over the traditional ‘broadcasting’ prevalent in many public health campaigns, we can scale up effective prevention and early detection, resulting in reduced burden for stretched services. We also hope this approach will more effectively help young people access trustworthy and appropriate information about mental health.

Aims

In its 2014 green paper, Closing the gap: priorities for essential change in mental health, the Department of Health said: “We need a truer, more up-to-date and more detailed picture of mental health and wellbeing nationally and in each area”. We hope that this exercise can respond by achieving this more detailed picture.

Throughout the exercise we have sought to assess:

  1. How young people use social media to access advice and support about mental health issues
  2. How services for young people living with mental health problems interact with young people talking about mental health online
  3. How social listening might allow public services to update policy and practice to support young people with mental health problems pre-diagnosis.

Main findings

  1. Mental health is a significant topic of online conversation amongst 19–24-year-olds.

2. The most common topic of conversation amongst both men and women was depression, but problems with sleep and issues surrounding tiredness and fatigue were more prevalent amongst young men.

3. Young people tend to talk about mental health in the evenings during the working week.

4. Young people tend to share resources among each other (peer-to-peer) rather than engage directly with local service providers.

5. Local service providers showed limited engagement with young people presenting with mental health problems online, taking a ‘broadcast’ approach as opposed to a dialogue approach.

6. Forums are still an area where young people discuss mental health issues anonymously. We found impartial advice about treatment and signs and symptoms of mental illnesses hard to come by on these forums and on other platforms.

7. Young men dominated online conversations surrounding suicide, where there was some evidence of imagery which ideated self-harm and suicide.


Methodology

The listening exercise identified approximately 134,000 unique social media users aged between 14–30 years old who were participating in conversations using specific terms relevant to mental health, between 16 July 2017 and 9 May 2018, across the UK. To better target conversations and connections between young people talking about mental health issues online, we also analysed the conversations of young people living in Greater Manchester, which made up 6 percent of the UK conversation during the timeframe of the research.

The RSA Lab commissioned data analysis firm Foco Global Limited to conduct the listening exercise. Using Pulsar TRAC, the search parameters and filters were set to identify all conversations relating to mental health among young people across the Greater Manchester area. The keywords we used to identify the volume of conversation were the specific words identified as being frequently used by online participants when talking about issues relating to mental health such as ‘depression’, ‘anxiety’ and ‘suicide’. Other search terms such as ‘scared’ were identified when used in a post alongside the term ‘mental health’ to identify the volume and veracity of conversations related to mental health. The exercise gathered and analysed data from social media sites such as Twitter, Instagram, news sites and forums.

Amongst forums we found that The Student Room generated significant conversation amongst young people seeking advice and support about mental health. Yahoo’s answers’ forum was also identified as a source of information among peer groups and BenzoBuddies which is a forum providing advice to people who have been addicted to the drug benzodiazepine was highlighted as a good source of peer-support for young people experiencing mental health consequences of addiction.

Analysis of location

For Twitter its Profile Geo enrichment was used, which looks at the profile country code or profile region. Twitter’s Profile Geo enrichment attempts to determine the best choice for the geographic place described in the author’s profile location. The result may not be accurate in all cases due to factors such as multiple places with similar or ambiguous names. If location is not provided in a user’s profile location field, Twitter will not attempt to make a classification and, as such, some tweets will not be captured.

The exercise also looked for geo-tagged tweets, where at the time of posting the user has tagged the tweet with a place or location that matches the specific country.

For news, blogs, and forums location using the country code of the domain or the author location were filtered.

For reviews, the country filter was based on the source (eg all reviews from Amazon.com will automatically be with country USA).

Since there’s currently no location information available for these sources, we automatically disabled data from Tumblr and YouTube when specifying location at search setup with these sources enabled.

For Instagram we collected all the data matching specified Instagram tags regardless of the location, and then performed post processing, where we used the location information available in the post such as geo-coordinates to determine the location of the author. We found that Instagram location is not always available. The reason for this is that unlike Twitter or Facebook, where location information can be found in an author’s profile, users on Instagram only provide location information when they geo-tag their posts.

For Facebook Creative Insights the location is determined by Facebook and takes into account the location information available in the author’s profile.

Time period

The social media listening covered the period between 16 July 2017 and 9 May 2018 only.

Online and social media ‘influencers’

Influencers are authors who not only produce content with the highest volume of mentions, but have a greater than average visibility or exposure, and exert influence around a subject by engaging with other users.

Influence is based on credibility, trust, network quality and timing. Therefore, influence is measured contextually, by looking at how impactful people are about a specific topic and within specific social networks and categories. This enables us to identify and target any online engagement precisely at those users who will amplify your message the most. The exercise identified influencers around the topic of mental health whose posts were geo-located in Greater Manchester and were mapped within the Influencers network graph (see Figure 6).


Key findings

This exercise highlighted many issues which recent research into social media and mental health has uncovered but evidenced a lack of effective dialogue with young people presenting with mental illnesses online. We witnessed inappropriate referrals to support suggesting this is a significant problem amongst young people who talk about their mental health online. We believe that this problem, stemming from a high volume of conversations about mental health amongst young people, could benefit from a national service that works to improve impartial and authoritative information about mental health online.

The listening exercise reflected the existing literature findings that mental health is a significant area of interest for young people and echo the Royal Society of Public Health’s Status of Mind report which identified eight health and wellbeing issues young people themselves highlighted as important:

  1. Awareness and understanding of other people’s health experiences
  2. Access to expert health information you know you can trust
  3. Emotional support (empathy and compassion from family and friends)
  4. Anxiety (feelings of worry, nervousness or unease)
  5. Depression (feeling extremely low and unhappy)
  6. Loneliness (feelings of being all on your own)
  7. Sleep (quality and amount of sleep)
  8. Self-expression (the expression of your feelings, thoughts or ideas).

These reflect the tone and subject of conversation the listening exercise uncovered, most significantly access to trustworthy information is highlighted as the most significant barrier to young people looking for mental health support online.

We found that:

  1. Mental health is a significant topic of online conversation amongst 19–24-year-olds.
Figure 1. National demographic profile of young people talking about mental health

Of the 134,000 young people we found to be talking about mental health issues, most young people talking about mental health were within the age range of 19–24 (70.5 percent). This was followed by 14–18-year-olds (16 percent) and 25–30-year-olds (13.5 percent).

We measured the conversation using key words identified by the RSA Lab which provided a focus for the listening exercise in order to find a greater density of conversation between different groups.

Of the keywords gathered during 16 July 2017 and 9 May 2018, ‘depression’ represented a 9 percent share of the conversation making up 56,000 imprecisions, with ‘anxiety’ representing 6 percent; ‘suicide’ representing 4 percent; ‘sleep’ representing 4 percent, ‘tired’ representing 3 percent and ‘self-harm’ representing 3 percent of the conversation. In total this makes us 88.5 percent of the conversation.

Figure 2. Total key words associated with mental health

2. The most common topic of conversation amongst both men and women was depression, but problems with sleep and issues surrounding tiredness and fatigue were more prevalent amongst young men.

Conversations we analysed during the listening exercise, which focused on young people with social media accounts living Greater Manchester highlighted that the words ‘depression’, ‘sleep and tiredness’, ‘suicide’ and ‘anxiety’ stimulated the most conversations. However, they were used differently when it came to gender. For example, the analysis found that young men tend to use words such as ‘help’, ‘stress’ and ‘tired’ in their conversations as opposed to women who frequently use words such as ‘sleep’ and ‘depression’.

Figure 3. Key words most commonly used by young men in Greater Manchester when talking about mental health
Figure 4. Key words most commonly used by young women in Greater Manchester when talking about mental health

3. Young people tend to talk about mental health in the evenings during the working week.

We found conversations about mental health were most frequent on Tuesdays between 4.00pm and midnight, Thursdays between 5.00pm and midnight, and Friday evenings between 4.00pm-11.00pm. Overall, most conversations were taking place in the afternoon and evenings.

Figure 5. Conversations about mental health by volume — national

4. Young people tend to share resources among each other (peer-to-peer) rather than engage directly with local service providers.

Using analysis generated from the online conversations between young people living in Greater Manchester, which took up 6 percent of the national conversation, we found individual bloggers achieved higher levels of engagement and reach than service providers [See end note 1]. Individuals such as 26-year-old blogger Alex Davies who writes at Pushing back the shadows was of particular interest given his long-standing commitment to talking about mental health issues and his age. Within his blog he signposts to forums and feeds such as My Depression Scars Facebook page, which have over 324,000 followers.

Other influential accounts that were of interest were ‘Lucy H’, who blogs specifically about miscarriage, mental health and wellbeing and Sassy Wyatt who is a 27-year-old award-winning blogger writing about disability and mental health.

Particularly amongst young people using Twitter accounts to talk about mental health, we found a more mixed picture with organisations such as Anxiety UK, BBC 5 Live and Manchester Evening News having high engagement scores alongside bloggers and young people tweeting individually.

Figure 6. Social media accounts widely shared amongst young people talking about mental health in Greater Manchester

5. Local service providers showed limited engagement with young people presenting with mental health problems online, taking a ‘broadcast’ approach as opposed to a dialogue approach.

In the target area of Greater Manchester, there was a noticeable difference in the use of social media between young people and support services. Mental health services such as the Improving Access to Psychological Therapies (IAPT), Child and Adolescent Mental Health Services (CAMHS), Time to Change, Young Minds and Childline showed no discernible engagement, defined as responsive digital engagement, troubleshooting problems and answering questions between the 3,720 young people identified as talking about mental health regularly in Greater Manchester.

We found that while there were instances of information being shared by service providers, there was very little evidence of direct engagement and participation by service providers with young people on topics about mental health. For example, one of the main forums where young people shared their thoughts and sought advice was The Student Room, yet there was no evidence during the research period of service providers participating in these conversations within this forum.

Furthermore, we concluded that service providers were (a) not using social listening and audience insight tools to effectively identify the relevant target audiences, conversations and channels to engage with them; and (b) using social media to their advantage to sign-post people to appropriate sources of information which could be done via targeted advertising on platforms such as Instagram and Twitter.

6. Forums are still an area where young people discuss mental health issues anonymously. We found impartial advice about treatment and signs and symptoms of mental illnesses hard to come by on these forums and on other platforms.

Figure 7. Examples of presentations of mental distress found in The Student Room on forums as well as comments from users

We found that amongst meaningful resources for first-hand accounts of mental illness, which help readers understand what services and support is available, there are many widely-shared blogs, articles and websites that don’t include signposts to support services. Many of these too serving commercial interests, including encouraging young people to taking up therapies or treatment for a fee. Perhaps unsurprisingly, young people appear to share more widely content that has been suggested by peers rather than formal services, and we found websites such as www.internetbusinessideas-viralmarketing.com/anxiety.html shedding more ‘heat’ than light on the issue which is indicative of many of links that were popular amongst young people talking about mental health online. These links, or others such as manchesterpsychotherapist.com/addiction/, can be helpful, but often to serve a commercial interest alongside a plausibly genuine interest in the area.

Figure 8. A selection of links shared for information from young people amongst other young people

7. Young men dominated online conversations surrounding suicide, where there was some evidence of imagery which ideated self-harm and suicide.

CN: Due to the graphic nature of the selected posts, we have omitted them from final publication

In the research period there were 716 posts shared on the subject of suicide from young people with accounts linked to Greater Manchester, generating 609,000 impressions. The largest share of voice discussing suicide was found among 19–24-year-olds at 64 percent. However, with 14–18-year-olds sharing 25 percent of the conversation, compared to 16 percent of the total we captured demonstrates that this is a significant issue for this group. The remaining 11 percent of the conversation was taken up by 25–30-year-olds.

The exercise also came across posts associated with self-harm and suicide which used romanticised language alongside imagery that ideated conditions such as post-traumatic stress disorders that achieved large reach amongst young people. We did not identify any coherent or consistent approach from providers of services that would proactively engage with young people when talking about depression, anxiety, self-harm, suicidal thoughts and other mental health issues.


Conclusion and recommendations

The evidence is clear that the topic of mental health generates substantial interest online amongst some young people. The risks attributed to being insufficiently responsive to increased awareness of mental ill-health amongst young people are large. Not only must we improve the way we engage with young people online, this must be reinforced with increased provision, earlier and far more proactively.

Our research uncovered a huge amount of energy and appetite for peer-to-peer support and self-expression online and there is much to celebrate in that. However, a culture of openness amongst young people can lead us to being hostages to fortune as more awareness leads to more conversations that risk confusing and misdirecting young people to sources of information and advice that lack credibility.

As recent calls for technology companies to “share child mental health costs” become louder, we have an opportunity for central and local government and platform providers to work effectively together to increase the coherence of mental health support online.

Apart from clinically led platforms like Big White Wall, only accessible to those who live in an area where the Clinical Commissioning Group or local authority has access to a licence and is signposted to it, significant forums for conversation and information sharing about mental health was found on unmediated sites that serve to provide more ‘heat’ than light for young people about the issue.

Modelling the clinical excellence demonstrated by online services such as Big White Wall, we can evidence the benefits of an online-first approach to mental health. Indeed, a recent independent evaluation of its LiveTherapy programme found that 67 percent of people referred by a GP achieved reliable improvement and 57 percent were moving to recovery. The national average recovery rate is 45 percent.

While these platforms remain expensive, they can be more effectively used if there is better quality information and advice online to help young people, which a new service we suggest could help deliver.

Just as we have core health services delivering interventions such as accident and emergency admissions in hospitals, we argue that our evidence points to the need for a new intervention which will help NHS and local government services act earlier to help tackle the confusing and at times unhelpful information about mental health that many young people are being exposed to online.

One way we propose to achieve this is through an online ‘111’ first port of call service aimed at young people who discuss mental health online. We believe a service, either delivered directly by NHS services, or via trusted third-party agencies, could improve significantly young people’s access to quality and clinically robust information about mental health online. The current research into meaningful online interventions remains sparse, with even organisations such as ‘MQ: Transforming mental health through research’, dedicating minimal resource to this area.

Whilst there are risks online, overall the evidence suggests that the social media environment is as complex as face-to-face interaction. Social media can be a place where there is an ‘ecosystem’ of mental health support for young people, which includes greater use of reputable platforms such as Big White Wall alongside a responsive and proactive service that engages with young people that are talking about mental health online.

We hope this analysis of conversations will help broaden the evidence used to shift the model of operating for services that aim to support young people living with mental illness.

We recommend relevant government departments take a five-stage approach to improving the online environment for young people:

  1. Public Health England, The Department for Health and Social Care and The Department for Digital, Culture, Media and Sport should work together and establish a working group (including mental health service providers, social media companies and the wider health technologies industry) to investigate how to coordinate an effective online response, asking how providers can be incentivised to invest in in a dialogic (rather than broadcast) model of support, which enables children and young people to get digitised practical support to manage emotional distress and mental health conditions.
  2. Engage existing support providers such as Big White Wall, the Samaritans and Childline to assess the efficacy of developing a ‘dialogue’ approach for young people who are raising mental health issues online, involving relevant agencies in the public sector (NHS Trusts, local authorities, schools).
  3. The Government’s Digital Charter should ensure that all digital platform providers meet an accuracy and trust threshold for mental health information and signposting, and require all digital companies to actively report, remove, and challenge harmful misinformation
  4. Pilot a service in selected local authority areas, working across platforms such as Twitter, Facebook, Tumblr and Instagram to respond to concerns about mental health that young people talk about online while also improving the quality and access to information about diagnosis and support services, both locally and nationally. This could pilot techniques such as offering free targeted advertising for a service on their platform.
  5. NHS England (as part of the formulation of a new 10 Year Plan) should stimulate agencies within STP footprints (CCGs, local authorities and providers) to actively promote dialogic, digital mental health information and interventions to children and young people in their area and build this into local Public Health priorities and service pathways.

Sources

Royal Society of Public Health (2018) Status of Mind. Available at: www.rsph.org.uk/our-work/campaigns/status-of-mind.html

Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. (2005). Lifetime Prevalence and Age-of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62 (6) pp. 593–602. Available at: www.mentalhealth.org.uk/statistics/mental-health-statistics-children-and-young-people

Gareth Iacobucci. The BMJ (2018) ‘Social media should take more responsibility for children’s mental health, says NHS boss’ 361:k2634. Available at: www.bmj.com/content/361/bmj.k2634

Education Policy Institute (2017) Social media and children’s mental health: a review of the evidence. Available at: epi.org.uk/publications-and-research/social-media-childrens-mental-health-review-evidence/

Defined as those who spend between one and two hours online on a typical weekday. OECD (2016) PISA 2015 Results Students Well-being Vol III. Available at: www.oecd.org/edu/pisa-2015-results-volume-iii-9789264273856-en.htm

Coates, S. and Lay, K. (2018) ‘Self-harming by teenage girls doubles in 20 years’, The Times, 6 August 2018 www.thetimes.co.uk/article/087be17c-98ff-11e8-be18-9b68e74f878e

www.penninecare.nhs.uk/your-services/manchester-resilience-hub/

The Kerslake Report: An independent review into the preparedness for, and emergency response to, the Manchester Arena attack on 22nd May 2017 (2018) Available at: www.kerslakearenareview.co.uk/media/1022/kerslake_arena_review_printed_final.pdf

Alldritt, C et al. The RSA (2015) Getting the message on mental health. Available at: www.thersa.org/discover/publications-and-articles/reports/getting-the-message-on-mental-health

Department for Health and Social Care (2014) Mental Health: priorities for change. Available at: www.gov.uk/government/publications/mental-health-priorities-for-change

Rawlinson, K. ‘Social media firms ‘must share child mental health costs’ (14 June 2018) The Guardian. Available at: www.theguardian.com/society/2018/jun/14/nhs-child-mental-health-costs-social-media-firms-must-share

See: www.bigwhitewall.com/v2/Home.aspx?ReturnUrl=%2f

See: www.mqmentalhealth.org/research/profiles/testing-online-self-help-for-anxiety

End Notes

[1] Other popular accounts picked up in the exercise were @Rockchichtich_; @davidh7426; @GMPCityCentre; @OVERCOMING_BLOG; @BorwneProject; @ProfCaryCooper; @JessieLingard; @DEpressON2; @HR_Aspects_mag and @J4CKMULL