2023-03-08-兰德-美国空军在社交媒体上的多样性_公平性和包容性价值观-空军附属客户如何在线代表服务的概念验证检查（英）-2023_16页_212kb

KEY FINDINGS

• USAF-affiliated users on public social media platforms rarely used offensive language, slurs, or hateful speech (e.g., racial slurs) targeting others based on identity, occurring at rates between 0.0001% and 0.04%.
• When offensive terms were posted, they often referred to the terms being used against the user (third parties), not the user applying the slur directly.
• Twitter discussions within the USAF community are infrequent regarding leadership, personnel, and policy issues, often focusing on broader American political debates rather than USAF-specific matters, and are politically polarized.
• Online pushback against DEI policies was observed but generally civil and constructive.
• Public SM monitoring can provide relatively low-cost situational awareness regarding adherence to USAF values and DEI principles without monitoring individual actions, respecting privacy and free speech rights.

LIMITATIONS

• Data collection relied on large-scale analytics, unable to definitively verify individual USAF affiliation of accounts or access private digital spaces.
• Twitter's censorship and account deletion practices may limit observed harmful content but could also affect data integrity.

RECOMMENDATIONS

• Evaluate low-cost, publicly-focused SM monitoring for assessing online behavioral consistency with Air Force values.
• Consider machine learning approaches for higher accuracy in identifying authentic accounts and specific problematic behavior.
• Support specific qualitative research to understand DEI-related private SM behavior.
• Continue research on using SM data for detecting extremism and monitoring specific policy reactions.