A study of U.S. Air Force social media behavior found a nearly undetectable number of instances of offensive speech, according to a RAND Corporation research report published Wednesday.
RAND researchers conducted an analysis of social media posts on Twitter and other websites from accounts that self-identify as members of the Air Force “through a diversity, equity, and inclusion (DEI) lens” the study states.
They found that across multiple platforms Air Force members, dependents and veterans avoided speech considered “hateful” and communicated in a respectful manner over the entire duration of the study, undercutting plans to track the effects of Air Force DEI policies on member conduct.
“We were able to find evidence that, at least in public, airmen only rarely use slurs or other distinct hate speech terms,” the researchers wrote.
To conduct the study, experts pulled communications from Twitter, a popular Air Force community thread on Reddit and an internal forum for enlisted members, all of which were identified as online communities with high traffic.
Researchers searched for keywords including “racism, affirmative action” and “bigot,” instances of “BLM” near the word “riot,” and racial slurs, as well as words pertaining to gender relations and posts mentioning race theory and similar concepts.
Out of the 53.9 million tweets included in the survey, only 2,371, or 0.0031%, contained a certain racial slur, the researchers found.
“Racist, homophobic, or xenophobic talk did not break thresholds of significance or effect size in standard text analyses,” the researchers wrote, while keyword searches yielded just 0.04% of tweets containing the offensive language.
Further analysis found that, in most cases, social media users did not always use those words in hostile ways.
They also found that users posted opinions critical of the Air Force’s DEI policies on Reddit and the internal forum, but those posts tended to be “civil and constructive.”
“Concerns were voiced about the amount and efficacy of new DEI-related training, with users expressing concern that DEI training was politically motivated,” they wrote.
Twitter data was collected going back to 2018, while the other datasets went back as far as 2011, the study says.
RAND had hoped to use the results for a before-and-after study of the effects of DEI policies on discourse.
“We were unable to find any individual DEI policy changes since 2016 or 2018 … that had a clear and major impact on USAF online community discourse,” they wrote.
The proof-of-concept analysis had several limitations; for example, they could not determine whether accounts were falsely representing closeness to the Air Force, and the use of keywords to harvest data meant more subtle digs were not included, the authors wrote.
Twitter members were identified using keywords, which means they could include retired airmen or family members of airmen. However, every account identifying with the Air Force could potentially influence public perception of the service, the authors noted.
They also acknowledged the potential effects of moderators and self-censorship. “We believe oversight directly contributed to the civility seen on these platforms,” they wrote.
The Air Force did not immediately respond to the Daily Caller News Foundation’s request for comment.
Leave a Comment