Linguabrand qualifies emotions through social media

emotional-baggageLinguabrand uses psycho-linguistic algorithms to extract social and emotional metrics from free text.

Recently they were supplied with two sides of an online discourse between those recently diagnosed with multiple sclerosis (MS) and those with the condition existing. Without even reading the discourse, they uncovered and quantified changing emotional needs.

The figures were measured against Linguabrand’s ‘Emotional Background Benchmark’. This provides a comparison of the measurement factors against levels you’d expect to hear in a normal spoken conversation. (In all of the graphs below ‘100’ equals this benchmark. Also note the changing values on the axes).

Although many of the findings seem obvious, quantifying conversations from a forum without even being read is less so. Gut feel is important; but until it’s measured it’s nothing but opinion. Psycho-linguistics opens up a new way of understanding how people are feeling.

MS – Life Transforming

It was immediately apparent that MS is a massive, problem-centric, transformation. Both new diagnoses and existing patients are strongly orientated towards moving away from that problem rather than moving towards answers.

MS1

In this respect we’d guess it’s little different from any other conversation around debilitating or life-threatening conditions. Those newly diagnosed are heavily focused on themselves, but this focus returns to normal levels as they come to terms with the condition.

Family is by far the most important social frame. The impact of MS on family life, present and future is extremely important. This has major implications for communications from drug companies. They need to include the family instead of talking to the patient alone.

A Storm of Emotions

Negative emotions were 50% higher than average levels for both patient sets. Looking at negative emotions in more detail was revealing…

MS2

… Sadness is the overwhelming feeling for new diagnoses, coupled with very high levels of anxiety. As people accept their condition sadness levels more than halve. But anxiety actually increases. This is an important finding, because anxiety is a focus of on-going medication. Understanding and reducing anxiety should be a key metric for medical communications.

Anger levels are half the level of a standard conversation. Perhaps there isn’t room for it amongst the sadness and anxiety? But it suggests that both patient groups are open to a conversation.

Uncovered – The Brave Face?
Existing people encourage the new. They talk much more about success and show higher levels of confidence. However, they are also much less sure of themselves.

MS3

Marry this finding to the increasing levels of anxiety as the illness proceeds and this sounds like putting on ‘the brave face’. Behind all their encouragement lies anxiety and doubt.

These findings, and more, enabled Linguabrand to make specific recommendations on patient communications.
Healthcare professionals are strongly science led. Bringing a metrics-driven approach to emotional data encourages them to understand, measure and manage patient relationships better. This could have a significant impact not only on patients’ quality of life, but potentially on their medical treatment itself.

Big data – intimate insights

Using social media data means is free of many research biases. It’s not question or process led. People take part speaking in their own voices, in their own time and in their own space. They select themselves rather than being offered a proposition to opt in.

Privacy is clearly important. Individuals are never revealed, indeed many don’t use their names anyway. And in big data it’s the collective that is important.

Understanding that language reveals emotions is remarkable in itself. Scaling this using software will change the way research works. This converts social and mobile-based free-text from unwieldy, unstructured data into a quantified emotional insights.

Alastair Herbert (1 Posts)