Do your tweets reveal much more about you and the health of those around you than you ever imagined?
What can tweet analysis show and predict?
Over two decades of research suggests that the language we use is significantly related to our personalities, emotional states, social connections and thinking styles. Psychologists at the University of Texas, Austin and Auckland Medical School in New Zealand created the Linguistic Inquiry and Word Count (LIWC) a text analysis program which enables words to be quickly and efficiently analysed. The program forms the Analyse Words website, a site that allows you to input your own tweets (or someone else’s if you feel so inclined) and have them analysed by LIWC.
Who better to run through Analyse Words system than the Dalai Lama?
Analysis of tweets from Dalai Lama
(983 most recent words – 19th March, 2015)
Upbeat (Very high) 100
Worried (Average) 52
Angry (Very high) 89
Depressed (Very low) 13
Plugged In (Average) 59
Personable (Very high) 93
Arrogant/Distant (Very high) 86
Spacy/Valley girl (Average) 59
Analytic (Very high) 91
Sensory (Very low) 20
In-the-moment (Average) 47
How does it work?
LIWC concentrates on ‘junk words’ which include words like I, you they, a, the, an, to, with, for and other small words that pull together more content-heavy nouns and regular verbs.
The Analyse Words website explains that a large number of studies have shown junk words to be powerful indicators of a person’s psychological state. For example, when someone uses the word I, they are momentarily paying attention to themselves. When a person experiences high levels of physical or mental pain they automatically focus on themselves and use more I-words. This focus on self through language can indicate depression, stress or insecurity.
Similarly, there are junk words that can signal deception, leadership and many other psychological states. Because the LIWC research team has collected mountains of language and psychological data, the researchers have a clear idea of which words best link to psychological processes.
Can tweets predict heart disease?
Psychologist, Johannes Eichstaedt and his colleagues at Pennsylvania University believe their research shows that twitter comments are substantial indicators of personality type, optimism and predisposition to depression. More significantly, Eichstaedt and his team asked if the words we use on social media could also be linked to the number of deaths from heart disease in those around us. The team recently published ‘Psychological Language on Twitter Predicts County-Level Heart Disease Mortality’ and discussed their result in the Journal of Psychological Science.
Eichstaedt and his team explain that hostility and chronic stress are known risk factors for heart disease, but they are costly to assess on a large scale. They therefore chose to examine language expressed on Twitter from 26 million tweets across 1400 American counties and used the information they gathered to give each county an emotional profile. The team’s research examined whether a link exists between each county’s emotional profile and the likelihood of death in that county from atherosclerotic heart disease (AHD).
Language patterns reflecting negative social relationships, disengagement, and negative emotions—especially anger—emerged as risk factors. Counties where people tweeted these negatively focused messages had higher rates of strokes, heart attacks and death from AHD.
In contrast, the counties that had a significantly high number of tweets containing positive emotions, reflecting psychological engagement, had a much lower number of heart attacks, strokes and fewer deaths from AHD. Positive emotions displayed in tweets emerged as protective factors.
Even when the researchers factored income and education into their data, the correlations remained significant. The team found that their twitter language model predicted heart disease mortality even more effectively than a model combining 10 common demographic, socioeconomic, and health risk factors, which included smoking, diabetes, hypertension, and obesity.
What can we learn from the study?
Research shows that less than twenty percent of Americans use twitter and within that percentage, the vast majority of twitter users are aged below fifty. Heart attacks, on the other hand, are statistically more likely to occur in the over fifties. We have to ask how the tweets from the younger generation of a county can predict heart health in the county’s older generation? Any causal link requires further investigation but Eichstaedt and his colleagues suggest that “the language of Twitter may be a window into the aggregated and powerful effects of the community context.” In other words, a young person living in a poorer county with fewer opportunities and a greater occurrence of social problems might be more likely to vent their stress and anger in negative tweets and the same environment may have a negative impact on the health of the county’s older generation.
The importance of tackling stress and managing negative emotions
For each and every one of us, this research illustrates the importance of managing our emotions and cultivating the skills that allow us to effectively deal with stress. Regardless of where we begin our stress management journey, we are all able to practice the techniques that enable us to enhance a feeling of calm and quickly move away from negative emotions rather than dwelling on whatever has caused us to feel that way.
Looking for some quick tips on where to make this positive change?
Try mindfulness. Andy Puddicombe’s short video, it takes is 10 mindful minutes is a great place to start.
Hear ‘the happiest man in the world’, Matthieu Ricard, share his habits of happiness for tips on developing a positive outlook.
Look out for Positive Change Guru’s forthcoming series of blogs covering how to develop positive emotions.