Time of Day Influences How Most People Think (and Tweet)
Twitter-based study shows how thought processes fluctuate over a 24-hour period.
Posted July 2, 2018
A massive analysis of 800 million anonymous tweets — containing over seven billion words — reveals universal trends in how modes of thinking tend to fluctuate over the course of each daily 24-hour cycle. This paper, “ Diurnal Variations of Psychometric Indicators in Twitter Content, ” was published June 20 in the journal PLOS ONE.
For this pioneering study, researchers from the University of Bristol used artificial intelligence (AI) to track specific word usage on Twitter from a large population sample during multiple 24-hour circadian cycles over a four-year period. Tweets sent during big holidays were excluded, because the language used during Christmas, New Year's, Valentine's Day, etc., was skewed by the nature of each particular holiday.
Word usage during non-holiday times of year was categorized based on 73 psychometric indicators, which other research has correlated with various thinking styles. Then, the Bristol researchers combed through all the Twitter content posted during specific times of day, looking for commonalities that mirrored the 24-hour circadian cycles that are synchronized to Earth's daily rotation on its axis.
Notably, the researchers found that the first peak expression time on Twitter started around 5 a.m. to 6 a.m. Early-morning Twitter language was often associated with drive (e.g., ambition, power, achievement) and personal concerns. In these early morning hours, the mega-analysis also found that tweets contained words correlated with a more analytical and logical way of thinking.
On the opposite end of the psychometric spectrum, the researchers found that anyone using Twitter in the wee hours of the morning (e.g., 3 a.m. to 4 a.m.) tended to use language correlated with a more impulsive, emotional, and existential mode of thinking. Interestingly, these tweets were also anticorrelated with the expression of positive emotions.
The project leader for this research was Nello Cristianini , professor of Artificial Intelligence at the University of Bristol. In a statement, he said: "The analysis of media content, when done correctly, can reveal useful information for both social and biological sciences. We are still trying to learn how to make the most of it."
In many ways, the results of analyzing 800 million tweets reflect common sense. Although the researchers used state-of-the-art technology, simply by observing the day-to-day social media usage of celebrities, politicians, and anyone else on Twitter, one could begin to predict how time of day may have influenced specific Twitter content.
For example, in September 2016, Donald Trump sent out an impulsive and emotional tweetstorm that began at 3:20 in the morning. Jennifer Rubin of the Washington Post reported on the noteworthy aspects of his extremely nocturnal Twitter use in an article: “ Donald Trump’s 3 a.m. Moment .”
After giving some background on the circumstances surrounding this series of tweets, Rubin says, “The question now may be: Do you want Donald Trump up at 3 a.m. tweeting crazy stuff? That’s precisely what he did.” Then, she chronicles Trump’s tweetstorm from the predawn hours that began, “ Anytime you see a story about me or my campaign saying ‘sources said,’ DO NOT believe it. There are no sources, they are just made up lies!" [sic]
From the typical onlooker's perspective, President Trump's daily Twitter habits appear to corroborate some of the recent findings on “Diurnal Variations of Psychometric Indicators in Twitter Content,” by Fabon Dzogang et al.
Like clockwork, @realDonaldTrump usually posts early-morning Twitter content that either reflects "drive and personal concerns," or his language is more "impulsive and emotional," depending on whether he starts tweeting at 3 a.m or some time after 6 a.m.
Anecdotally, most of us know that our modes of thinking and language usage fluctuate throughout the day based on firsthand experience.
Over the past few years, I’ve figured out through trial-and-error that the best time of day for me to do any type of analytical thinking begins around sunrise and ends around noon. As a real-time example, I started writing this post around 5:30 a.m. and felt analytically focused at the time. Unfortunately, July 4th travel plans interrupted my flow. Now, the sun is low on the horizon, and I'm scrambling to connect the dots in new or useful ways. I'm also struggling to finesse the language and polish the content — which is taking me twice as long to write in the early evening as it did this morning.
In an interview with the Paris Review , “ The Art of Fiction, No. 21 ,” Ernest Hemingway corroborates the universal link between time of day, our thinking processes, and written words. When asked by the interviewer, “ When do you work? Do you keep to a strict schedule?" Hemingway responded:
“When I am working on a book or a story I write every morning as soon after first light as possible. There is no one to disturb you and it is cool or cold and you come to your work and warm as you write. You write until you come to a place where you still have your juice and know what will happen next and you stop and try to live through until the next day when you hit it again. You have started at six in the morning, say, and may go on until noon or be through before that. . . It is the wait until the next day that is hard to get through."
Why Are Most of Us More Analytical in the Early Morning?
Although this new study shows that our cognitive and emotional states tend to ebb-and-flow based on a 24-hour cycle, the exact cause of these changes is up for debate. Most likely, the hour-to-hour changes in our modes of thinking and word usage are linked to a fluctuation of hormones and neural activity that is tied to our sleep/wake cycles and circadian rhythms.
As the authors explain, “Indeed, it is clear that circadian genes modify key circuits which impact on mood and reward circuitry. In addition to neural circuits, circulating hormones such as cortisol contribute to optimal neural function and the maintenance of good mental health. Observing temporal variations in the expression of our internal state requires the investigation of large populations of individuals sampled at an adequate rate which has previously proven difficult.”
This study is potentially groundbreaking in that the researchers have culled massive amounts of data from anonymous sources in a way that helps us identify common trends in how people around the globe tend to think and express themselves depending on the time of day.
"Circadian rhythms are a major feature of most systems in the human body, and when these are disrupted they can result in psychiatric, cardiovascular, and metabolic disease. The use of media data allows us to analyze neuropsychological parameters in a large unbiased population and gain insights into how mood-related use of language changes as a function of time of day. This will help us understand the basis of disorders in which this process is disrupted,” co-author Stafford Lightman of the Bristol Medical School concluded in a statement.
Fabon Dzogang, Stafford Lightman, Nello Cristianini. "Diurnal Variations of Psychometric Indicators in Twitter Content." PLOS ONE (Published: June 20, 2018) DOI: 10.1371/journal.pone.0197002