Firearm Violence and Health in Policymaker Discourse: Mixed Methods Social Media Analysis
Bibliographic record
Abstract
Background: Since 2019, firearm violence has remained the leading cause of death for US children and adolescents aged 1-19 years. This crisis has spurred action from policymakers, health professionals, and advocates. However, political polarization has contributed to divergent views on the causes and appropriate responses to firearm violence. Communication by elected officials, especially on social media, plays a critical role in shaping public opinion and policy agendas. Understanding how state policymakers discuss firearm violence, including the use of causal blame, calls to action, and health-related narratives, can inform more effective public health strategies. Objective: This study aimed to examine how Pennsylvania state legislators discuss firearms and firearm violence on social media and assess the extent to which their messaging aligns with public health perspectives. Methods: We conducted a 2-phase mixed methods analysis of X (formerly known as Twitter; X Corp) posts by Pennsylvania state legislators from May 27, 2017, to July 26, 2022. Posts were grouped into 3 time periods surrounding the Tree of Life Synagogue mass shooting in Pittsburgh. Using a Boolean search strategy, we identified 4573 posts related to firearms and firearm violence. After removing reposts and non-English content, we randomly sampled 1491 (32.6%) original posts authored by 152 unique legislators. Posts were coded using a structured codebook based on the Multiple Streams Framework to capture rhetorical framing, causal blame, and policy content. Interrater reliability was high (Holsti coefficient >0.8). We used chi-square tests and multivariable logistic regression to assess associations between rhetorical elements and policy mentions, adjusting for time period. Results: Mass shootings were the most frequently referenced category of firearm violence, peaking after the Tree of Life shooting (22/43, 51% vs 91/118, 77.1% vs 140/220, 63.6%; P=.004), while firearm suicide was rarely discussed. Posts using advocacy frames were nearly 5 times more likely to mention policy (adjusted odds ratio [aOR] 4.67, 95% CI 3.55-6.16), whereas those referencing mass shootings (aOR 0.54, 95% CI 0.37-0.77) or emotional appeals (aOR 0.53, 95% CI 0.40-0.69) were significantly less likely to do so. Most posts used general advocacy (aOR 2.97, 95% CI 2.13-4.13) and vague blame (aOR 8.26, 95% CI 6.02-11.35), resulting in nonspecific policy suggestions. Posts that attributed blame to firearm access were strongly associated with specific policy proposals (aOR 6.37, 95% CI 4.29-9.47) and inversely associated with general policy mentions (aOR 0.26, 95% CI 0.17-0.42). Only 9.4% (133/1422) of posts used health frames; when present, they more often referenced physical consequences (58/133, 43.6% vs 216/1358, 15.9%; P<.001). Conclusions: Pennsylvania legislators primarily focused on mass shootings and relied on emotional or symbolic language without proposing specific policies. Health frames were rare and typically focused on consequences rather than prevention. Findings highlight an opportunity to support policymakers with health-informed messaging strategies to promote actionable firearm violence prevention policies, particularly those addressing prevention.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.044 | 0.110 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.011 | 0.010 |
| Science and technology studies | 0.003 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.002 | 0.005 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".