Could Facebook be the next Freud?

Camilla Penney 20 November 2016

The interests of Dr Becky Inkster are wide-ranging. A brain-imaging geneticist by training, in 2014 she attracted viral media attention for her work using hip-hop to raise awareness of mental health issues. She is now senior manager of the Neuroscience in Psychiatry Network (NSPN), which researches brain development in adolescents and young adults. In last week’s print edition TCS Science reported on Dr Inkster’s latest paper, “A decade into Facebook: where is psychiatry in the digital age?”, published in The Lancet Psychiatry, which advocates the incorporation of social media and social networking data into mental health research and treatment. The Cambridge Student spoke to Dr Inkster to get a more detailed understanding of this work.

The growing popularity of social media, particularly amongst young people, offers vast quantities and new types of information, which could provide insights into users’ mental health. In “A decade into Facebook”, the authors argue that “data from social networking sites should become a high priority for psychiatry research and mental healthcare delivery.” Inkster expands on this; “We must tap into the enormous possibilities that social media could offer, and specifically social networking… to find innovative ways to improve the detection, treatment and prevention of mental health problems. We need to start asking questions such as, ‘Do the things that people put on social networking sites tell us something about their mental health, and can we use this information to help identify warning signs, or find new treatments?’“ 

Understanding what data collected from social media actually means requires online information to be linked to offline mental health information. The researchers “propose a novel online-offline framework, which combines Facebook data collection … with mental health information already provided by… participants.” Offline information will help the researchers to understand the meaning of the new types of data being shared on Facebook. For example, particular patterns of friendship groups, or changes in communication could be indicative of social isolation. Previous studies have already shown “that positive things that are written on the newsfeeds tend to influence positively what you write on your own profile and vice versa”. Inkster hopes that by combining online and offline datasets it will be possible to identify relevant Facebook measures, which can be used effectively in treatment. However, “we are only at the very beginning of this extraordinary challenge and there is a strong need for evidence-driven exploration.” At the moment no studies have been carried out involving mental health patients so “we really don’t know how best to supplement or compliment offline information with appropriate online data in order to refine or improve mental health treatments. This is the ambitious goal we must carefully steer toward.”

The lack of psychiatric studies is partly due to emphasis of possible adverse consequences in both the scientific and media narrative. “The paper was essentially an outline, or almost a manifesto, trying to illustrate that there are positives,” says Inkster, “We often hear about the risks and potential negatives – and rightfully so – [but] we might also be able to use online social networking data to help improved our understanding and support of mental health problems.” A recent Guardian article exemplified this, worrying that “the idea of being neatly categorised as “mentally ill” or “mentally well” simply because of the things we choose to share online is both unethical and potentially dangerous.” Inkster says that this misunderstands the research, “We’re not trying to come up with some ‘black box’ algorithm where you or ‘Big Brother’ can go online and assess or survey whether or not you have a mental health problem based on your momentary thoughts and weekly data entries etc. This is very much still a conversation between the patient and the healthcare provider. We are suggesting that… online social media data might help to refine or supplement existing psychotherapies…  Facebook can be a very personal thing given that you decide what to share …so this might help to personalise [patients’] therapeutic intervention.”


Many of the concerns highlighted by the Guardian are related to who can access this online data. Inkster stresses that obtaining informed consent is paramount. For her, one of the key questions is “would this be acceptable offline?” As a result, the consent process for patients participating in the studies is very detailed, “if the [patients] are happy with all of the ethically-approved information… and everything we’ve put online then they click opt-in consent buttons. There are multiple consent buttons, it’s not a case of ‘can we have all of your Facebook data?’… Each thing is separate [so patients] think about each thing individually. They also have a very clear and easy way to remove themselves from the study and to have all of the data removed too – offline and online. Participants share very intimate things about their offline lives with us already, so we have very well defined standard operating procedures and safeguarding mechanisms in place for those circumstances. We need to expand and adapt [these mechanisms] for online circumstances and this will be very important.”

Although her Lancet paper doesn’t present any clinical results, in initial workshops with young mental health service users “The feedback had been very positive. I’ve worked with intelligent and socially conscious young people. I think sometimes people can underestimate just how aware they are about how they might like [Facebook] to be implemented into their care… they’re really in touch with the language used in Facebook and the digital social space. I can’t say that about all clinicians.” The patients have even suggested particular things that the researchers might wish to investigate. This knowledge of the medium, particularly amongst young patients, makes social media “one… area where [patients] really can lead and voice their views and concerns into the conversation” which can be “really empowering and relevant for the patient.” These patient and public involvement workshops are a continuing method of research, and one that Inkster is very keen for students to get involved in.

The authors also have an interesting cohort to begin their research with. “The very first dataset we will be analysing is a Cambridgeshire dataset… called the ROOTS cohort and our study is called “My Online Life”. It’s an adolescent/young adult cohort, [with data collected from the] early 2000s onwards… What’s really interesting about collecting cohort data around that time… is that Facebook exploded in popularity then, with the creation of ‘Likes’ in 2009, for example.“ Inkster says she and her colleagues “have collected… over 100 Facebook datasets from ROOTS participants [with] paired offline information …Participants have previously completed a wide range of offline questionnaires, such as about childhood experiences (e.g., whether they experienced maltreatment) and for young adults, what types of relationships they have, how they perceive their friendships in the offline sense, their family settings… [These] offline social factors … will compare really nicely with what we’re able to obtain through Facebook… data such as 'Likes', status updates, comments and so on.”  

With such enriched and socially interesting initial data sets, Inkster is focused on the future, “There is a lot of work to be done by many people no doubt, yet so much potential to revolutionise mental healthcare.” Watch this space.