Academics have used Hawaiian Twitter to analyze anxiety levels in the run-up to and days after the missile strike false alarm – and, unsurprisingly, people were more than a little anxious about the prospect of nuclear annihilation.

Perhaps more surprising, however, is the fact that the event appeared to be more stressful for Hawaiians who are naturally less anxious than those whose anxiety levels typically run high. The results of the study are published in the journal American Psychologist. 

At 8.07am local time on January 13, 2018, people in the Aloha State received a text, saying: "BALLISTIC MISSILE THREAT INBOUND TO HAWAII. SEEK IMMEDIATE SHELTER. THIS IS NOT A DRILL."

This was an error made by an employee at the Hawaii Emergency Management Agency during a preparedness drill and was retracted after 38 minutes. Still, for 38 minutes people really did think they were under attack and that their lives were in danger.

The Centers for Disease Control and Prevention (CDC) published a study in February comparing people's tweets before (8.07-8.45am) and after (8.46-9.24am) the retraction, noticing themes in the messages ranging from examples of information processing and sharing pre-retraction to demonstrations of denunciation and a distrust of authority after the retraction.

Now, researchers at the University of California, Irvine (UCI) have conducted a similar study – again using Twitter but this time extending the time frame to six weeks before and 18 days after the event to measure how anxiety levels were affected by the false alarm.

"Can a false alarm of an impending disaster itself be a form of trauma?" lead author Nickolas M. Jones, now a postdoctoral researcher at Princeton University, asked in a statement. "Our results suggest that the experience may have a lingering impact on some individuals well after the threat is dispelled."

The team used big data (1.2 million posts from 14,830 Hawaiians), scanning weeks' worth of tweets for 114 words associated with anxiety – afraid, scared, and worried etcetera. A score of one was given to any tweet that contained an anxiety-associated word. Others were given a zero. Users were then categorized into three groups, based on how anxious-appearing (low, medium, or high) their tweets were before the false alarm.

Twitter is a popular tool for social scientists, who can analyze users' thoughts and feelings in response to collective trauma (e.g. school shootings and natural disasters) by their wording. 

In this instance, anxiety (or rather, anxiety as expressed on Twitter) increased 3.4 percent every 15 minutes during "the attack" before it was retracted. For many, that anxiety persisted even after they had received the all-clear for at least two days afterward – a finding that surprised the researchers. 

"This suggests that cancellation of a threat doesn’t immediately calm reactions to the situation. Amazingly, some people did not know whether the corrective tweets were believable," senior author Roxane Cohen Silver, UCI professor of psychological science, medicine and public health, said in a statement.

The analysis also revealed – again, somewhat surprisingly – that people categorized as highly anxious calmed down faster than those with low anxiety. Those in the first group displayed a new baseline anxiety level that was 2.5 percent higher than before the message, whereas the baseline for the latter decreased by 10.5 percent.

"While those who before the alert had exhibited the least anxiety took the longest to stabilize, at approximately 41 hours, and the medium-anxiety group took 23 hours, the individuals who had exhibited the greatest anxiety before the alert stabilized almost immediately," said Jones.

Silver suggests it could be that the threat of a strike helped naturally anxious people put day-to-day stress into perspective.

"Anxious individuals may have more to appreciate when they experience a near miss and thus express less anxiety on social media after having ‘survived’ what would have undoubtedly been construed as a deadly situation," she explained.