









































Trends in Pro-Gun Control Tweets 
Surrounding a Mass Shooting

Event Suggest Time-Dependent 
Changes in Public Response

Lily Shytle

 
GEORGETOWN SCIENTIFIC
RESEARCH JOURNAL

Volume Three
Edition One
Fall 2022

6



Georgetown Scientific Research Journal 

Trends in Pro-Gun Control Tweets Surrounding a Mass Shooting 
Event Suggest Time-Dependent Changes in Public Response 

Lily Shytle1 

1 Department of Biology, Georgetown University, Washington, D.C. 
Email: eas384@georgetown.edu 
https://doi.org/10.48091/gsr.v2i2.33 

Abstract 

Over the period from May 25, 2015 to May 25, 2018, there were 1,296 mass shootings recorded 
around the United States. This article evaluates overall public support for gun regulation before 
and after a major mass shooting. Big data that accounted for outliers and other potential skewing 
points were included from Twitter.  250 Tweets that occurred immediately before, immediately 
after, or two weeks after a major mass shooting were analyzed. The Tweets were restricted to 
include at least one of the following hashtags (#): secondamendment, neveragain, shooting, NRA, 
guncontrol, gunviolence, and 2A. The Tweets were categorized as pro-gun control, anti-gun 
control, or neutral, defined by operational definitions. It was hypothesized that there would be 
more Tweets that opposed gun control 1-7 days before each mass shooting but fewer Tweets that 
opposed gun control 72 hours after each mass shooting. Three days (72 hours) after a major mass 
shooting, there was a significant decrease in the number of anti-gun control Tweets posted 
compared to 1-7 days prior to the shooting. There was also a significant increase in the number of 
anti-gun control Tweets 14-21 days after the mass shooting compared to 72 hours after the shooting, 
implying that the content of Tweets after two weeks was similar to the content before the shooting 
occurred. 

Keywords: mass shooting, gun violence, Twitter, big data

1. Introduction

Mass shootings are an epidemic in the
United States, taking hundreds of lives each 
year. In the first five months of 2018, gun 
violence killed 5,647 people and wounded 
10,344 people.1 A mass shooting is defined as a 
shooting that claims four or more victims, 
either killed or injured, in a single incident.2 
Many solutions for gun violence have been  

proposed, yet the Center for Disease Control 
and Prevention (CDC) is prohibited from 
researching gun violence and prevention. The 
Dickey Amendment (1996) states that "none of 
the funds made available for injury prevention 
and control at the Centers for Disease Control 
and Prevention may be used to advocate or 
promote gun control.”3 This law has severely 
limited research on gun violence and public 

7



Georgetown Scientific Research Journal 

perception. Although it has been argued that 
some research is still possible, there are no 
training grants or federal funding available to 
researchers, effectively hindering most 
potential research initiatives. Despite federal 
funding blockages, a meta-analysis of data 
from 1999-2013 found that more restrictive 
firearm legislation was associated with lower 
pediatric, unintentional, suicide, and overall 
firearm-related fatality rates.4 

1.1 Legislation Surrounding Gun Violence 

Gun violence accounts for almost half of the 
deaths of female partners in domestic violence, 
making it by far the most lethal method. A 
study conducted by Gollub et al. (2018) found 
that areas with fewer firearm restriction laws 
had a higher number of incidences of murder 
by gun violence compared to those with more 
restrictive gun laws.5 More restrictive gun 
control policies, including more thorough 
background checks and waiting periods longer 
than 72 hours, are associated with a 25% 
decrease in incidences of arrest for firearm 
violence. Additionally, laws requiring a permit 
to purchase firearms have been directly 
correlated with lower firearm homicide rates. 
Background checks have shown to be the most 
effective at decreasing firearm related fatalities, 
with universal background checks facilitating a 
60% decrease in overall fatalities from firearms. 
Despite this staggering statistic, 22% of firearms 
sales in the United States occur without 
background checks, as transactions between 
private parties are exempted from these laws in 
many states.6  

1.2 Utilizing Twitter as an Open Source for Big 
Data 

Social media platforms provide valuable 
insight into public opinion, specifically as it 
relates to news and policy changes. Many 
researchers rely on polls and surveys, but these 
methods are subject to sampling bias and data 
collected directly from the source of public 
information is therefore preferred.7 At the time 
of the data collection, Twitter was both one of 
the largest social media platforms in the world 
and one that had no option to make accounts 
private.8 According to Jashinsky et al. (2016), 
people were approximately 30% likely to blame 
the government for poor gun laws before a 
mass shooting occurred, but up to 66% likely to 
blame the government afterward.9  Therefore, 
the number of Tweets opposing and the 
number of Tweets supporting gun control 
restrictions were expected to show a significant 
difference before and after a major mass 
shooting occurred. 

1.3 Experimental Goals 

The goal of this research was to examine 
how the overall public perception of gun 
regulation changes after a major mass shooting. 

2. Methods

The Tweets analyzed were first filtered
using a web crawler known as Twarc to limit 
them to certain time frames and include specific 
hashtags.10 For all the data collected, the Tweets 
had to include one or more of the following 
hashtags (#): secondamendment, neveragain, 
shooting, NRA, guncontrol, gunviolence, or 
2A. The hashtags analyzed were chosen based 

8



Georgetown Scientific Research Journal 

on their high activity both before and after a 
mass shooting and because they are neutral 
terms that would not unfairly skew the data. 
The Tweets were then restricted based on time 
frames 1-7 days before the chosen mass 
shooting and the three days following the 
event. The mass shootings selected were 
determined by those with the highest number 
of fatalities that occurred from May 31, 2015 - 
May 31, 2018. The chosen time frames are listed 
in Table 1. In addition to this, Tweets were 
collected 2 weeks following the shootings. 
From there, the Tweets that fit the above criteria 
were numbered in the order they appeared on 
Twitter, and a random number generator was 
used to select 250 of the Tweets for each time 
frame. In total, 3750 Tweets were analyzed 
before, immediately after, and 14-21 days after 
a major mass shooting occurred. The Tweets 
that were ultimately selected were then 
classified as being pro-gun control, anti-gun 
control, or neutral. Operational definitions 
were used to classify the Tweets into three 
categories. Tweets were classified as pro-gun 
control if they expressed negative attitudes 
towards the National Rifle Association and 
other gun advocacy groups, encouraged 
Congress to enact stricter laws and background 
checks for purchasing firearms, or expressed 
sentiments that access to guns was not 
restrictive enough in the United States. Tweets 
were classified as anti-gun control if they 
supported the National Rifle Association, 
encouraged a larger number of people to own 
guns to protect themselves, opposed restrictive 
gun laws, and specifically included their access 
to guns given the second amendment. Tweets 
that did not fit these definitions were classified 

as neutral. After collecting the results, a chi-
squared analysis test was run on the data at α = 
0.01.   

Table 1. Dates analyzed for mass shooting data. 
Dates reported as MM/DD/YY. 

Location Before During After Fatalities 

San 
Bernardino 

11/25/15 
– 

12/1/15 

12/2/15 
– 

12/5/15 

12/16/15 
– 

12/22/15 

14 

Orlando 6/05/16 
– 

6/11/16 

6/12/16 
– 

6/15/16 

6/26/15 
– 

7/2/16 

49 

Las Vegas 9/24/17 
– 

9/31/17 

10/1/17 
– 

10/4/17 

10/15/17 
– 

10/21/17 

58 

Sutherland 
Springs 

10/28/17 
– 

11/4/17 

11/5/17 
– 

11/8/17 

11/19/17 
– 

11/25/17 

25 

Parkland 2/06/18 
– 

2/13/18 

2/14/18 
– 

2/17/18 

2/28/18 
– 

3/6/18 

17 

3. Results

Figure 1 shows the results of the analysis of
Twitter data for the mass shootings analyzed. 
The light blue bar represents the number of 
Tweets classified as pro-gun control, the red bar 
represents those classified as anti-gun control, 
and the dark blue bar represents neutral 
Tweets. 

9



Georgetown Scientific Research Journal 

Figure 1. Gun Control Tweet Trends Proximal 
to Mass Shooting Events. For each of the three time 
periods (1-7 days before the event, the 72 hours after 
the event, and days 14-21 after the event), 250 
Tweets were chosen by a random number generator 
from 3750 total Tweets. Subfigures (A)-(E) represent 
an individual event and analysis of the Tweets at 
the indicated intervals surrounding that event: (A) 
San Bernadino, (B) Parkland, (C) Sutherland 
Springs, (D) Orlando, (E) Las Vegas.  

For all the shootings analyzed, there was a 
significant increase in the percentage of pro-
gun control Tweets 72 hours after the shooting, 
which regressed towards the mean after 14-21 
days with a p-value of <0.00001. The data taken 
before the shootings showed that, before a 
major mass shooting, the majority of the Tweets 
were anti-gun control for 4 out of 5 of the 
shootings studied. The relative percentage of 
anti-gun control Tweets were 68.4%, 62.0%, 
60.8%, and 68.4% for San Bernadino, California; 
Parkland, Florida; Sutherland Springs, Texas; 
and Orlando, Florida; respectively. The raw 
data is included in Tables 3-7 below. The only 
shooting that had a larger percentage of pro-
gun control Tweets immediately before the 
event was Las Vegas, which could be due to the 
proximity of the shooting in Orlando. All the 

(B) 

(C) 

(D) 

(E) 

(B) 

(A)

10



Georgetown Scientific Research Journal 

shootings analyzed showed a linear regression 
towards the original classification of Tweets 14-
21 days after the shooting, with some shootings 
having an even larger number of anti-gun 
control Tweets than before. This trend was 
observed in San Bernadino, Sutherland 
Springs, and Orlando. Data can be found in the 
appendix at the end of the document. 

4. Discussion

The purpose of this research was to examine
how the public perception of gun control 
changes after a major mass shooting event. The 
United States of America has the highest rate of 
gun violence among all developed countries, as 
well as some of the least restrictive gun control 
laws.11 All the trials showed a significant 
increase in the number of Tweets expressing 
sentiments arguing for stricter gun control 72 
hours following a mass shooting compared to 
the baseline measured 1-7 days before the 
shooting. As predicted, there was a statistically 
significant decrease in the number of Tweets 
expressing the need for stronger gun control 14-
21 days after the shooting compared to 72 hours 
after for each of the trials in this time frame. 
Although the Tweets classified as pro-gun 
increased immediately after the mass shootings 
analyzed, the percentage decreased two weeks 
after the mass shooting. This suggests a 
regression towards the baseline of a higher 
number of anti-gun control Tweets both 
immediately before and two weeks after a mass 
shooting.  

Many people admit to posting sentiments 
online that they could never or would never say 
aloud, which is known as the online 

disinhibition effect.12 This phenomenon allows 
online or anonymous surveys to often yield 
different results than traditional written or 
identifiable survey methods. In the case of gun 
control, where people often have strong, 
unwavering opinions, it is unlikely that actual 
opinions change significantly after a major 
mass shooting. However, people may feel 
emboldened after a traumatic event and take to 
social media to feel heard. Therefore, people 
who feel strongly about gun control but do not 
usually post or share their opinions may 
influence the percentage of pro- or anti-gun 
control Tweets after a mass shooting, giving the 
false appearance of societal changes in opinion. 
Additionally, introjection refers to the 
subconscious adoption of ideas or attitudes of 
others, and many people may become 
consumed by the overall message dominating 
the hashtags chosen. This phenomenon 
provides further reasoning as to why people do 
not contradict the opinions of the larger group, 
as well as presents an explanation for why 
overall social media message may change 
without massive policy or public opinion 
change. Furthermore, the mass shootings 
included in this project were carefully chosen to 
be in a relevant time frame in which Twitter 
was incredibly popular and to have a high 
enough number of fatalities to reach national 
news sources.  

Other factors that could have contributed to 
the increase in Tweets advocating for stricter 
gun control laws immediately after the 
shootings investigated include people feeling 
more emboldened to speak up about beliefs 
that they already held before a mass shooting 

11



Georgetown Scientific Research Journal 

but were less vocal about. This hypothesis 
cannot be ignored, and further studies 
surveying people throughout the country 
should be done to gather more holistic views, 
rather than just those of the people posting 
online. Although the data suggests that there 
may be a major change in opinion before and 
after gun control, it is also plausible that people 
who are in favor of gun control before a mass 
shooting are not emboldened or vocal about the 
issue until after a shooting. Before the studied 
shootings occurred, people may not have heard 
about shootings that occurred all over the 
country because they were not as lethal as the 
measured ones, resulting in decreased vocality. 
Mass shootings have become commonplace, 
and people typically react less extremely to 
events that happen more frequently.13 

Policy changes regarding mass shootings 
have been much more frequent after increased 
vocal support, but many of these are laws that 
relax pre-existing gun regulation.14 For 
example, the state legislature of Texas passed a 
permit-less carry bill just two years after two 
shootings in El Paso and Odessa claimed 30 
lives.15  

4.1 Previous Research on the Topic 
Few previous studies have conducted a 

meta-data analysis of qualitative Tweet content 
regarding mass shootings before and after a 
mass shooting, as big data is still a relatively 
new science. Therefore, this research does not 
have other experiments to corroborate with, 
though it is expected that repeat experiments 
will yield the same results. It is possible that 
the scanned Tweets may have had a 
larger percentage  of  pro-gun  control  or  anti-gun 

control sentiments than the whole of Twitter 
represented after a mass shooting. However, 
3,250 Tweets were ultimately scanned and 
classified, so any random sampling bias 
should have been eliminated due to sample 
size. In the future, research should expand 
upon the findings of this study to analyze 
and further classify Tweets into categories 
from news outlets, politicians, and general 
social media use. 

4.2 Continuation and Next Steps 
In a continuation of this experiment, a 

larger span of time before and after the 
shootings could be analyzed to establish 
a more comprehensive baseline with which 
to compare the Tweets that occurred 72 hours 
after. Further research could also examine 
less fatal mass shootings to determine if the 
same trends are seen in these shootings, and 
at what point the fatalities are not great 
enough to cause a statistically significant 
difference in the data. Future research 
should analyze statistically significant 
differences in multiple social media interfaces, 
specifically those that include different 
key demographics than Twitter. Future 
experiments could be conducted to 
determine the impact of mass shootings on 
actual policy changes, rather than just the social 
media desire for policy changes to occur. To do 
this, proposed legislation and electoral 
promises would be followed in relation to dates 
of particularly lethal mass shootings. 
Additionally, future research could focus on 
the location from which the Twitter data is 
posted to pinpoint local pockets of data that 
could skew the results. 

12



Georgetown Scientific Research Journal 

Acknowledgments 

At this time, I would like to thank the 
faculty and staff at the Discovery Magnet 
Program at Spring Valley High School, 
specifically Mrs. Michelle Spigner and Dr. 
Michelle Wyatt. Additionally, I would like to 
thank the University of South Carolina and the 
research lab of Dr. Elaine Miller for the summer 
internship opportunity that allowed me to 
analyze and interpret the data in question. 

References 

1. Gun Violence Archive. (n.d.). Mass shootings in
2018. Retrieved from
https://www.gunviolencearchive.org/reports/m
ass-shooting?year=2018

2. Gun Violence Archive. (n.d.). Mass shooting
methodology and reasoning. Retrieved from
https://www.gunviolencearchive.org/explainer
/mass-shooting-methodology-and-reasoning

3. Zhang, S. (2018). Why can’t the US treat gun
violence as a public health problem? The
Atlantic. Retrieved from
https://www.theatlantic.com/health/archive/20
18/02/gun-violence-public-health/553430/

4. Resnick, S., Smith, R. N., Beard, J.H., Holena,
D., Reilly, P. M., Schwab, C. W., & Seamon, M. 
J. (2017). Firearm deaths in America: Can we
learn from 462,000 lives lost? Annals of Surgery, 
266(3), 432-440. 
https://doi.org/10.1097/SLA.0000000000002376 

5. Gollub, E. L. & Gardner, M. (2019). Firearm
legislation and firearm use in female intimate 
partner homicide using National Violent Death 
Reporting System data. Preventive 
Medicine, 118, 216–219. 
https://doi.org/10.1016/j.ypmed.2018.11.007

6. Wintemute, G.J. (2018). How to stop mass
shootings. The New England Journal of Medicine,

379(13), 1193–1196. 
https://doi.org/10.1056/NEJMp1807277 

7. McGregor, S. C. (2019). Social media as public
opinion: How journalists use social media to
represent public opinion. Journalism, 20(8),
1070–1086.
https://doi.org/10.1177/1464884919845458

8. Smith, A. & Anderson, M. (2018). Social media
use in 2018. Pew Research Center. Retrieved
from
https://www.pewresearch.org/internet/2018/03/
01/social-media-use-in-2018/

9. Jashinsky, J. M., Magnusson, B., Hanson, C., &
Barnes, M. (2016). Media agenda setting
regarding gun violence before and after a mass
shooting. Frontiers in Public Health, 4, 291.
http://doi.org/10.3389/fpubh.2016.00291

10. Twarc. (n.d.). Twarc. Retrieved from
https://twarc-project.readthedocs.io/en/latest/

11. Masters, J. (2022). U.S. gun policy: Global
comparisons. Council on Foreign Relations.
Retrieved from
https://www.cfr.org/backgrounder/us-gun-
policy-global-comparisons

12. Suler J. (2004). The online disinhibition
effect. Cyberpsychology & Behavior, 7(3), 321–326.
https://doi.org/10.1089/1094931041291295

13. Moyer, M. W. (2022). The ‘Psychic numbing’ of
mass tragedies. The New York Times. Retrieved
from
https://www.nytimes.com/2022/05/26/well/live/
coping-mass-shooting-
tragedy.html#:~:text=There's%20a%20scientific
%20term%20for,been%20studying%20since%20
the%201990s

14. Luca, M., Malhotra, D., & Poliquin, C. (2020).
The impact of mass shootings on gun policy.
Journal of Public Economics, 181, 104083.
https://doi.org/10.1016/j.jpubeco.2019.104083

15. McCullough, J., & McGee, K. (2022).
Confronted with mass shootings, Texas

13



Georgetown Scientific Research Journal 

Republicans have repeatedly loosened gun 
laws. The Texas Tribune. Retrieved from 
https://www.texastribune.org/2022/05/24/texas-
gun-laws-uvalde-mass-shootings/

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

14



Georgetown Scientific Research Journal 

Appendix 
 
 Table A1. Raw Data of Tweet Classification for Shooting on 2 December 2015 (San 
Bernardino) 
 Pro-Gun 

Control 
Anti-Gun 
Control 

Neutral 

1-7 
Days 
Prior 

42 171 37 

72 
Hours 
After 

144 96 10 

14-21 
Days 
After 

26 208 16 

Table A2. Raw Data of Tweet Classification for Shooting on 14 February 2018 (Parkland) 
 Pro-Gun 

Control 
Anti-Gun 
Control 

Neutral 

1-7 
Days 
Prior 

26 155 69 

72 
Hours 
After 

180 53 17 

14-21 
Days 
After 

80 136 34 

Table A3. Raw Data of Tweet Classification for Shooting on 5 November 2017 (Sutherland 
Springs) 
 Pro-Gun 

Control 
Anti-Gun 
Control 

Neutral 

1-7 
Days 
Prior 

42 152 56 

72 
Hours 
After 

130 111 9 

14-21 
Days 
After 

41 162 47 

 

15



Georgetown Scientific Research Journal 

Table A4. Raw Data of Tweet Classification for Shooting on 12 June 2016 (Orlando) 
 Pro-Gun 

Control 
Anti-Gun 
Control 

Neutral 

1-7 
Days 
Prior 

11 212 27 

72 
Hours 
After 

114 126 10 

14-21 
Days 
After 

24 219 7 

  Table A5. Raw Data of Tweet Classification for Shooting on 1 October 2017 (Las Vegas) 
 Pro-Gun 

Control 
Anti-Gun 
Control 

Neutral 

1-7 
Days 
Prior 

19 189 42 

72 
Hours 
After 

156 86 8 

14-21 
Days 
After 

50 172 28 

 

16




	Table of Contents
	Letter From the Editors
	Trends in Pro-Gun Control Tweets Surrounding a Mass ShootingEvent Suggest Time-Dependent Changes in Public Response
	Human Papillomavirus Vaccination Among Low-Income,Underserved Youth in the Greater Washington, D.C., andHackensack, NJ Metropolitan Areas
	Best Practices for Health Literacy Education for EnglishLanguage Learners
	About the Authors
	Acknowledgments
	GSRJ Volume 3 Edition 1 back.pdf
	Table of Contents
	Letter From the Editors
	Trends in Pro-Gun Control Tweets Surrounding a Mass ShootingEvent Suggest Time-Dependent Changes in Public Response
	Human Papillomavirus Vaccination Among Low-Income,Underserved Youth in the Greater Washington, D.C., andHackensack, NJ Metropolitan Areas
	Best Practices for Health Literacy Education for EnglishLanguage Learners
	About the Authors
	Acknowledgments




