What Nearly 3 Million Tweets Reveal About Americans' Cannabis Attitudes
Analysis of 2.9 million cannabis-related tweets found 23% expressed positive attitudes (mainly about medical value) while 8% were negative (mainly about difficulty quitting), with 42% of users identified as potential cannabis consumers.
Quick Facts
What This Study Found
Among 2.87 million unique cannabis tweets, positive attitudes (22.6%) centered on medical value while negative attitudes (8.2%) focused on difficulty quitting. The 25-34 age group was most represented among cannabis-using tweeters.
Key Numbers
2,865,562 tweets analyzed. 22.62% positive, 8.17% negative, 69.21% neutral. 821,451 unique users, 42.46% potential cannabis users. 25-34 age group most represented.
How They Did This
Content analysis of 2,865,562 unique non-commercial cannabis tweets from February 2022-2023 using deep learning (BERT) for attitude classification and topic modeling (LDA) for theme identification.
Why This Research Matters
Social media reflects and shapes public cannabis attitudes. Understanding what people actually discuss — medical value vs. quitting struggles — helps public health authorities craft messaging that resonates with real concerns.
The Bigger Picture
The cannabis conversation online is overwhelmingly positive or neutral, with negative content focused on dependence rather than health risks. This information asymmetry may influence public perception and policy attitudes.
What This Study Doesn't Tell Us
Twitter/X users don't represent the general population. BERT classification has error rates. Cannot verify actual cannabis use from social media claims. Data from one year only.
Questions This Raises
- ?How do social media cannabis attitudes influence real-world use patterns?
- ?Should public health campaigns target the specific gap between perceived medical value and quitting difficulties?
Trust & Context
- Key Stat:
- Evidence Grade:
- Large-scale social media analysis with advanced NLP methods — strong for understanding online discourse but limited in representing general population attitudes.
- Study Age:
- Recent analysis (2022-2023 data) capturing cannabis discourse in the current social media landscape.
- Original Title:
- Content Analysis of Cannabis Discourses on Twitter/X in the U.S.
- Published In:
- AJPM focus, 4(6), 100408 (2025)
- Authors:
- Xie, Zidian, Zhou, Runtao, Yun, Qihao, Wu, Jianghang, Wang, Zhengyuan, Yu, Mengmeng, Wilson, Karen M, Li, Dongmei
- Database ID:
- RTHC-07984
Evidence Hierarchy
A snapshot of a population at one point in time.
What do these levels mean? →Frequently Asked Questions
What do people say about cannabis online?
Most tweets (69%) were neutral. Among those with clear attitudes, positive tweets focused on cannabis's medical value, while negative tweets centered on difficulty quitting — an interesting split between perceived benefits and real challenges.
Does living in a legal state change what people tweet about cannabis?
States with legal recreational cannabis had more cannabis-related Twitter users per capita (12.2 vs. 7.2 per 10,000), though the difference wasn't statistically significant.
Read More on RethinkTHC
Cite This Study
https://rethinkthc.com/research/RTHC-07984APA
Xie, Zidian; Zhou, Runtao; Yun, Qihao; Wu, Jianghang; Wang, Zhengyuan; Yu, Mengmeng; Wilson, Karen M; Li, Dongmei. (2025). Content Analysis of Cannabis Discourses on Twitter/X in the U.S.. AJPM focus, 4(6), 100408. https://doi.org/10.1016/j.focus.2025.100408
MLA
Xie, Zidian, et al. "Content Analysis of Cannabis Discourses on Twitter/X in the U.S.." AJPM focus, 2025. https://doi.org/10.1016/j.focus.2025.100408
RethinkTHC
RethinkTHC Research Database. "Content Analysis of Cannabis Discourses on Twitter/X in the ..." RTHC-07984. Retrieved from https://rethinkthc.com/research/xie-2025-content-analysis-of-cannabis
Access the Original Study
Study data sourced from PubMed, a service of the U.S. National Library of Medicine, National Institutes of Health.
This study breakdown was produced by the RethinkTHC research team. We analyze and report published research findings without making health recommendations. All interpretations are based solely on the published abstract and study data.