Abstract

Little is known about the content of conspiratorial posts on social media, specifically regarding active shooter events. This study explores the content of conspiracy theory tweets during the six months following the active shooter event at Robb Elementary School in Uvalde, Texas. 194 tweets were analyzed. The analysis resulted in two main themes: False Flag and Hypothesized Reasoning. These themes represent the similarities and variations in content among the tweets. In line with research suggesting that people are most vulnerable to conspiratorial thinking when trying to explain unusual and unsettling events, results indicate that many posts were attempting to provide some kind of causal explanation for the event. Additional findings, implications, and directions for future research are also discussed. 

On Tuesday, May 24, 2022, at 11:33 am an armed man entered Robb Elementary School in Uvalde, Texas. Shortly after entering and firing in the hallway, the man entered one of the classrooms and opened fire. This attack caused 21 fatalities (19 students and 2 teachers) and 17 injuries (ALERRT, 2022). The victim characteristics (i.e., elementary-aged children and teachers) and the inadequate police response resulted in national attention across all forms of media (Kellner, 2022). While there have been active shooter events with elementary-aged victims (e.g., the event at Sandy Hook Elementary) and long police response times (e.g., police response to the event at Columbine High School), never has there been an attack such as that at Robb Elementary School. Previous research assessing aspects of social media use around active shooter events is instrumental to understanding what and how much is being shared during these events (e.g., Mazer et al., 2015) and the short-term effects on users (e.g., Jones et al., 2017). The literature specific to social media responses to active shooter events has not explicitly explored the content of conspiracy theory posts regarding those events. The broader conspiracy theory literature provides researchers explanations for why individuals may believe various theories, including differences across racial groups (Goertzel, 1994), educational attainment (van Prooijen, 2017), social media use (Enders et al., 2023), and political ideology (e.g., Enders et al., 2022; van der Linden et al., 2021). Trends in conspiracy theory belief over time are also explored (Uscinski et al., 2022). However, like social media response research, existing conspiracy theory research does not include the exploration of the content of conspiracy theories regarding active shooter events. Addressing this gap in the extant literature is necessary. To be able to combat the negative effects of conspiracy theories, we must first know what they believe. We examine the content of conspiracy theory tweets regarding the Robb Elementary School shooting, bringing to light the negative impacts this content could have on survivors, families of victims, perceptions of polarization in the U.S., and perceptions of police legitimacy.

Background and Current Purpose 

Active Shooter Events

An active shooter is defined as “an individual actively engaged in killing or attempting to kill people in a confined and populated area; in most cases, active shooters use firearm(s) and there is no pattern or method to their selection of victims” (U.S. Department of Homeland Security, 2008, p. 2).  Note there is not a minimum required number of casualties (i.e., injuries or deaths) for an incident to be classified as an active shooter event (Sandel & Martaindale, 2022). Active shooter events should not be mistaken for mass murder or mass shootings. Mass murder requires three or more deaths in a single incident (Investigative Assistance for Violent Crimes Act of 2012, 2013), and though mass shooting is not federally defined, researchers have indicated that it should similarly require three or more people be shot in a single incident (Sandel & Martaindale, 2022). Given these definitions, it is important to iterate that mass murder and mass shooting incidents are not necessarily active shooter events, and vice versa. It is possible for an active shooter event to also be deemed a mass murder and/ or mass shooting, but it all depends on the situational characteristics and outcome of the incident.

 Active shooter events are multifaceted with some ending quickly due to the actions of the attacker, citizens, or law enforcement, while some are more complex and involve barricaded doors and mobile attackers targeting more than one location (U.S. Department of Homeland Security, 2008). While these are rare, their frequency and severity in the U.S. is increasing. From 2017 to 2021, active shooter events have been trending upward with 61 events occurring in 2021, a 52.5% increase from 2020 and 96.8% increase from 2017 (U.S. Department of Justice, 2022). In 2022 there were 50 events with 313 total casualties (100 killed and 213 wounded) (U.S. Department of Justice, 2023). The 2022 casualty count was higher than the 2018-2021 period average (222.50) (U.S. Department of Justice, 2023).

As people try to understand how and why these events occur, they may be left feeling powerless and fearful for their own safety. The combination of feeling fearful and powerless may make people more susceptible to believing and sharing conspiracy theories about these events (Franks et al., 2013; Uscinski & Parent, 2014). This might include believing explanations like the government orchestrated the event as a means to some political end. In other words, as they try to make sense of the event, some may cope by embracing conspiracy theories.  Ultimately, the belief that the government was responsible for the event to bolster a particular agenda at least provides an identifiable and confrontable villain, creating a false sense of power and control. Acknowledging that these events do not necessarily have simple causal explanations means there is no obvious way to identify or prevent potential active shooters. So, one must accept that these events can occur anywhere and claim anyone as a victim. Unfortunately, this heavy truth provides little comfort, making the explanations provided in conspiracy theories more appealing.

Understanding Conspiracy Theories

Conspiracy theories are any “belief that two or more actors have coordinated in secret to achieve an outcome and their conspiracy is of public interest but not public knowledge” (Douglas & Sutton, 2023, p. 282). These theories are characterized by their opposition to the generally accepted understanding of an event, descriptions of malevolent acts, and assignment of agency to individuals and groups rather than to impersonal or systemic situations (Douglas & Sutton, 2023).  Though they are not necessarily false or implausible, they are more likely to be false than other beliefs (Douglas & Sutton, 2023). These theories are often not falsifiable. They assume many people coordinate and cover up the events and argue that anyone who tries to disprove the theory is involved in the conspiracy (Lewandowsky et al., 2015). These theories persist despite a lack of reliable evidence to support their claims (Lewandowsky & Cook, 2020). Relatedly, a false flag is “a hostile or harmful action (such as an attack) that is designed to look like it was perpetrated by someone other than the person or group responsible for it” (Merriam-Webster Dictionary, n.d., 1st listed on webpage). For example, in 1939, German SS soldiers posed as Polish soldiers and stormed the Gleiwitz radio tower on the German side of the Germany-Poland border (Pope, 2018). They transmitted a broadcast saying that the tower was under Polish control. The next day, Adolf Hitler delivered a speech, citing the Gleiwitz attack and other orchestrated events to justify the invasion of Poland (Pope, 2018). There was ample evidence to indicate that the Gleiwitz event was a false flag attack (Pope, 2018).  If there is no substantiated evidence that an event was a false flag operation, that explanation would then be characterized as a conspiracy theory.

Why People Believe Conspiracy Theories

Some research has reported individual differences in conspiracy theory beliefs based on race and education. One study found that minority status (i.e., being Hispanic or Black) was significantly and positively correlated with belief in eight of ten conspiracy theories being studied (Goertzel, 1994). Another found that people with higher education are less likely to believe conspiracy theories, compared to those with low education (Douglas et al., 2016; van Pooijen, 2017; van Prooijen et al., 2015). Although these individual-level differences emerge in research, belief in conspiracy theories generally is quite common. For example, over half of Americans believe that the government was involved in covering up the truth about the JFK assassination (Jensen, 2013). Likewise, many Americans believe that the attacks on 9/11 were an inside job orchestrated to justify a war on terror (Laine & Parakkal, 2017).

 A key feature of conspiracy theories is the providing of causal explanations for upsetting and often unusual events (van Prooijen & Jostmann, 2013), making them more appealing to various audiences. There are several reasons someone may believe a conspiracy theory including feeling powerless (Uscinski & Parent, 2014), coping with threats (Franks et al., 2013), explaining unlikely events (Kovic & Füchslin, 2018), and disputing mainstream politics (Sapountzis & Condor, 2013). Overall, people are more likely to believe and endorse conspiracy theories when they are coping with a threat or uncertainty (Lewandowsky & Cook, 2020).  Research suggests that believing conspiracy theories is a simple cognitive response to a search for meaning and patterns where there likely are none (e.g., Whitson & Galinsky, 2008). There is discomfort around rare crisis events, especially those that result in seemingly indiscriminate casualties. Conspiracy theories are designed to make those events more easily digestible whether that be by blaming an outgroup (e.g., an opposing political party) or denying that the event occurred at all.

Political Ideology and Conspiracy Theory Belief

Research is mixed regarding whether people identifying as either liberal/democrat (i.e., the left) or conservative/republican (i.e., the right) are more likely to believe conspiracy theories (Enders et al., 2022). Some argue that individuals on the political right more often use a bottom-up psychological process in which sensory information is received from the environment, resulting in perceptions that are based on the immediate sensory input (Gibson, 1966). Alternatively, top-down psychological processing filters incoming information through the lens of prior knowledge, experiences, and expectations (Gregory, 1970). In other words, those who engage in bottom-up processing may form perceptions that rely on sensory information, rather than existing knowledge and experiences, which could make them more apt to believe conspiracy theories (van der Linden et al., 2021). However, recent research suggests that the observed differences are more likely artifacts of conceptualization and operationalization of conspiracy theories (Enders et al., 2022; Enders & Uscinski, 2021). This argument also exists regarding the effects of political extremism (i.e., those on either the far right or left) on conspiracy theory beliefs. Some researchers indicate that far-right or -left extremists are more likely to believe conspiracy theories than moderates (Imhoff et al., 2022). Others suggest those findings depend on how ideologies are operationalized and the effects of other variables (Enders et al., 2022; Enders & Uscinski, 2021). For example, findings indicate that knowledge of, and engagement in, politics result in less conspiracy theory buy-in. This is particularly notable for those theories challenging the political order and not a particular political out-group (Enders & Uscinski, 2021).  Thus, the effects of political extremism may be moderated by one’s political involvement. Although findings are generally mixed, one study recently found that identifying as conservative/republican was significantly related to believing that Sandy Hook was faked and belief in government false flags (Enders et al., 2022). Both beliefs are relevant and provide context for data interpretation in the current study.

Impacts of Social Media

A lack of gatekeeping online allows for the spread of misinformation, including conspiracy theories, sometimes at rates faster than those of correct information (Vosoughi et al., 2018). During active shooter events in particular students, parents, and community members seek continuously updated information as the situation develops and may turn to social media platforms for information (Lachlan et al., 2014). Students and parents share a considerable amount of information through social media during these events, this information may be inaccurate, increasing emotional trauma after the event (Suomalainen et al., 2011).  The surge in attention around the event may lead to the spread of information that conflicts with that coming from official sources and may be difficult to disentangle.

Conspiracy theories often emerge following crisis events (e.g., an active shooter event) (van Prooijen & Douglas, 2017), and online spaces like X (formerly Twitter) increase people’s exposure to conspiracy theories, false information, and rumors during and following crises. Previous research examining the effects of social media use on conspiracy theory belief suggests that online exposure to conspiracy theories is not enough to make someone believe them (Enders et al., 2023). Rather, one must exhibit a tendency toward conspiratorial thinking (Enders et al., 2023). Broadly speaking, conspiracy theory beliefs following online misinformation and conspiracy theory exposure is prevalent among groups with particular characteristics (Bail et al., 2019; Benkler et al., 2020; Enders et al., 2023; Guess et al., 2019; 2020; Nyhan, 2020).

Current Study

The active shooter event at Robb Elementary School colloquially referred to as “Uvalde”, was chosen for the current study for three notable reasons. First, the shooting at Robb Elementary was the deadliest attack at an elementary school since the Sandy Hook Elementary School shooting in 2012. Second, the police response received national attention. Third, there continued to be revelations regarding the police response with new information being shared with the public over time resulting in continued public engagement and reaction to the events that transpired that day. These factors make this active shooter event susceptible to the effects of rumors and ultimately conspiracy theories, as individuals attempt to understand the behaviors and events that led to the loss of 21 lives. This study explores the content of “#FalseFlag” tweets during the six months following the active shooter event. We believe that understanding the content of conspiracy theory tweets around these types of events may aid researchers in combating the negative effects conspiracy theories.

 We took a generic qualitative inquiry approach (e.g., Caelli et al., 2003; McLeod, 2001; Merriam 1997; Patton, 2015). This approach includes “qualitative methods - in-depth interviewing, fieldwork observation, and document analysis - to answer straightforward questions without framing the inquiry within an explicit theoretical, philosophical, epistemological, or ontological tradition” (Patton, 2015, p. 155). This approach is, at its core, a pragmatic one, in which “qualitative methods now stand on their own as reasonable ways to find out what is happening in programs and other human settings” (Patton, 2015, p. 154). As such, the following research question was addressed. 

RQ: What themes are present in the conspiracy theory tweets regarding the Robb Elementary School active shooter event? 

Method

Data Collection

After creating an X (formerly Twitter) developer account, we applied for access to the X application programming interface (API). An API “is a set of rules or protocols that let software applications communicate with each other to exchange data, features and functionality” (IBM, n.d., First para. on webpage). Recent changes in the X developer platform limit the extent of access to data depending on the type of API one subscribes to: X API v2 Free, Basic, or Pro. To be able to retrieve tweets, one must subscribe to at least the $100 per month, API v2 Basic subscription, allowing for the retrieval of up to ten thousand tweets per month. For this data collection effort, we used the $5000 per month, API v2 Pro subscription. This allowed for the retrieval of up to one million tweets per month. We used the API v2 Pro subscription for one month to collect the data for this and other related projects. After obtaining access to the X API and determining which subscription is best suited for their purposes, researchers are able to use the application of their choice (e.g., python, R, Postman) to collect data from X based on researcher-identified query terms. 

Data for the current study were collected in the Postman application. Within the global environment, we found “Twitter API v2”. We then created a fork of this collection from the global environment to our personal workspace. This is necessary, as it will not allow one to conduct a full archive search in the global environment. Once this collection was forked to our personal workspace we went to “Full-archive search” in the “Search Tweets” folder.  Researchers entered their bearer token into the “Authorization” section of the Full-archive search tab.

Each time a query is used, it returns 500 tweets, beginning with the most recent date and working backward. The “next_token” was used to collect the next batch of 500 tweets, continuing backward until the “start_time” was reached. Each batch of 500 tweets is saved as a JSON file that must be converted by researchers. The collected JSON files were converted to csv files and subsequently combined into a single sheet using Gigasheet. This sheet was saved and converted to Excel format for the data analysis procedures. 

All original, English tweets posted between May 24, 2022, and November 24, 2022, containing #Uvalde were queried (n = 91,289). Researchers then extracted all tweets containing “false,” “staged,” and/ or “hoax” (n = 361). These terms were chosen as a proxy for “conspiracy theory.” This was done because although users would say that the event was a false flag, staged, and/ or a hoax, they were not as prone to using the term “conspiracy” in their posts. These tweets were reviewed for relevance, and those that used the queried terms but were not conspiracy theory-related were dropped from the analysis, providing a final sample of 194 tweets.  

Data Analysis

The researchers used an inductive approach to the data analysis. To facilitate the data analysis, the tweets were added to an Excel spreadsheet. The authors then conducted separate analyses of the data sets to identify main and sub-categories. Data were analyzed using thematic content analysis (Kuckartz, 2014). First, tweets were read holistically. The initial read provided both authors the opportunity to become familiar with all tweet information. This first step also allowed for the observation of any patterns or relationships in the tweets prior to developing the main categories for analysis. This information was used in the next step to create the initial coding structure (i.e., main categories). Second, the authors met to create a coding scheme based on information gathered during the initial read. Main categories were created and applied later in the analysis (Kuckartz, 2014). The main categories included “belief” and “blame” 

Third, data were initially coded using the main categories. Data included both the text content of the tweets and the content of any links that the tweet included, as this information would be visible within the tweet. Tweets were reread, and relevant tweets were assigned to the main categories. Fourth, each main category would be used to generate subcategories that break down each code further. The initial category of “belief” was restructured into subcategories, “denial,” “hoax,” and “orchestrated.” “Blame” was separated into the subcategories, “who” and “why”. The second major phase of the coding process followed (Kuckartz, 2014). During this coding phase, no new coded segments were created, but each segment coded into a main category was then reassigned to the appropriate subcategory.  These categories and subcategories were then used in the development of the two major themes: False Flag and Hypothesized Reasoning.

Results

The current study's purpose was to explore themes in tweets about the Robb Elementary School active shooter event in Uvalde, Texas. Tweets containing conspiracy theory-related content were most common immediately after the event occurred, with 140 tweets shared in May, followed by 40 in June 12 in July, one in both August and September, and none in October or November. The extent to which other users engaged with the tweets in the sample varied (see Table 1). Of all tweets in the sample, about 65% were liked, 13% were quoted, 42% were replied to, and 37% were retweeted by at least one other user. 

Table 1. Engagement with Conspiracy Theory-Related Tweets 

Pro- Versus Anti- Conspiracy Posts 

Upon completing the initial read through of all relevant conspiracy theory- related tweets, posts were sorted into either the “pro-conspiracy” or “anti-conspiracy” groups. “Pro-conspiracy” tweets (n = 110) included content that indicated the user believed in some conspiracy theory related to the event. “Anti-conspiracy” tweets (n = 84) mentioned conspiracies with the intention of ridiculing people who believe in those theories. Those who posted anti-conspiracy content differed in whether their posts were proactive or reactive. That is, some users were sharing posts in which they anticipated what conspiracy theory believers might say or do. Others posted in reaction to what they had already seen, heard, or read from the conspiracy theorists. However, they remained constant in their antagonizing, insulting, and calling out people who believe and share conspiracy theories.

Both pro- and anti- conspiracy tweets were most common immediately after the event, with pro-conspiracy content persisting beyond that of anti-conspiracy content (see Figure 1).  

Figure 1. Persistence of Pro- and Anti- Conspiracy Tweets over Time 

After we identified differences in users’ orientation to conspiracy theories, we decided to explore the level of engagement with both pro- and anti- conspiracy tweets. It is worth noting these analyses are exploratory and were conducted after we determined there were two distinct groups of conspiracy-related posts. Having access to the engagement metrics data for all posts in the sample allowed for further exploration of this information. First, we examined the percent of tweets in each group that received at least one like, quote, reply, and/ or retweet (see Table 2). Pro-conspiracy tweets had higher percentages of tweets receiving at least one like, quote, and/or retweet. Anti-conspiracy tweets had a higher percentage of posts receiving at least one reply.  

Table 2. Percent of Tweets with at Least One Like, Quote, Reply, and/ or Retweet 

Second, four separate independent-sample t-Tests were conducted to assess whether the pro- and anti- conspiracy tweets varied significantly on any of the engagement metrics. Table 3 includes descriptive information for the engagement metrics for each group.  

Table 3. Descriptive Statistics for Engagement with Pro- and Anti- Conspiracy- Related Tweets 

There was no significant difference in the number of retweets (t₍₁₉₂₎ = 1.80, p = 0.07) received by pro-conspiracy tweets (M = 1.95, SD = 4.86) and anti-conspiracy tweets (M = 0.89, SD = 2.61). There was no significant difference in likes received, t₍₁₉₂₎ = 1.36, p = 0.18, despite pro-conspiracy tweets (M = 6.54, SD = 22.66) receiving over twice as many likes on average than anti-conspiracy tweets (M = 3.08, SD = 6.40). This is likely due to the large standard deviation present in the pro-conspiracy tweets. There was also no significant difference in the number of quotes (t₍₁₉₂₎ = 0.38, p = 0.71) and replies (t₍₁₉₂₎ = -0.23, p = 0.82) received by either group. Although we include exploratory findings comparing the volume and engagement of posts in both groups, the qualitative analysis and results are focused on the pro-conspiracy tweets. 

Pro-Conspiracy Tweet Themes 

Among the pro-conspiracy tweets, there were two themes- defining “false flag” and hypothesized reasoning. The way users defined “false flag "included three distinct belief types: denial, a hoax, and orchestrated event. Denial refers to those tweets that indicate that the poster does not believe the event occurred at all.  About 6% of all posts denied that the event occurred. For example, an individual posted the following tweet:

This is one of the handful of tweets that implies that the shooting in Uvalde did not occur at all. In this tweet, the user quotes the word gunman, implying there was no gunman at all. They also question the legitimacy of a school classroom containing students from more than one grade. However, that information is not confirmed, all the student victims’ ages ranged from 9 to 11 years old with most of them being 10 years old, which is consistent with that of a typical fourth grader considering early and late birthdays.   

Hoax refers to posts in which the individual believes the event occurred but was staged with crisis actors, and no lives were lost. About 24% of posts suggested that this event was a hoax. For example, a poster wrote:

This sentiment appeared 26 times in the sample. These posts indicated that some people thought the whole attack was staged using crisis actors, and subsequently broadcasted to the public. The logic being that the goal is to further the rhetoric promoting gun control in the U.S. 

Orchestrated refers to those who believe that the event occurred, and innocent lives were lost, but that it was coordinated by the government to bolster a political agenda. This subcategory contained the greatest number of tweets (n = 38 or 35% of all pro-conspiracy tweets) referencing a certain type of false flag. For example, a user tweeted the following:

Others, like the above writer, note their belief that the government and law enforcement officials are corrupt. For example, another individual stated:

This post differs from the one preceding it though, including the “#2A” as a reference to the 2nd Amendment. This was a commonly used hashtag in tweets discussing gun control topics. While the previous tweet mentioned the government's “many evil agendas” and the one referenced after it included only the “#2A,” others were more explicit in the political agenda that they believed the government would murder children and teachers for. For example, a user stated,

This tweet identified the gun control agenda as being the primary motivation for orchestrating this event. These users are among several that indicated they believe this event occurred, and that it was planned and executed by the government in the interest of pushing a certain political agenda. 

The second theme was “hypothesized reasoning.” Tweet content that contributed to this theme was that which suggested some motivation for the alleged false flag operation. Suggested motivations were primarily related to ideas about gun control. These posts explicitly mention gun laws or the Second Amendment.  Most often, tweets would accuse the government of being involved in the planning and execution of the active shooter event- orchestrated or otherwise-with the goal being to take guns. For example, a poster said,

Relatedly, in their response to a post made by Daily Mail, a poster wrote,

These examples highlight the lengths to which some people believe government and law enforcement officials would go to push a particular political agenda. Note that these accusations were also present in the tweets informing the False Flag theme. Recall, for example, the tweets shared by the poster for “hoax” and two of those for “orchestrated” all also referred to the gun control agenda to some extent.

Discussion

The combination of lapses in school security and subsequent police response to this active shooter event made it susceptible to the effects of rumors and conspiracy theories, as individuals attempted to understand the behaviors and events that led to the loss of 21 lives. This study explored the content of conspiracy-related tweets during the six months following the active shooter event. Interestingly, users posted both pro-conspiracy (belief in conspiratorial aspects) and anti-conspiracy (mocking conspiratorial beliefs) tweets immediately following the attack. No significant differences in engagement metrics (i.e., likes, quotes, replied, retweets) were found between the two groups. One key interpretation of this finding is that pro-conspiracy tweets suggesting that the attack was staged in some manner receive the same amount of engagement as the anti-conspiracy tweets. This suggests there is an equally engaged audience supporting conspiratorial beliefs that the Uvalde attack was in fact a false flag.  

The most notable finding among the pro-conspiracy tweets was the variation in how users defined “false flag.” Definitions included complete denial of the event occurring, belief that the event occurred and included real victims but was orchestrated by the government, and belief that the event was staged and employed crisis actors. This finding highlights the importance of conceptualization and operationalization in future research. As researchers continue working on this topic, they should be aware of and account for definitional differences between research participants.  

Relatedly, most of those who defined “false flag” indicated that they believed the event did occur and included real victims but was orchestrated by the government. This increased level of distrust in the government may negatively impact individuals’ perceptions of police legitimacy. Although these tweets represent a small proportion of the total X response to this event, they did have views, likes, and retweets. The frequency with which the posts in this sample placed blame on both the government and various law enforcement agencies is alarming. As public perceptions of police legitimacy decrease so does trust and obligation to obey authorities (Gau, 2011, 2014; Sunshine & Tyler, 2003). In other words, folks who do not view the police as a legitimate source of authority are less likely to trust and/ or comply with police officers, having negative consequences for public and officer safety. 

Similarly, many users either denied that the event occurred or indicated that they thought it was a hoax. Suggesting that this event and others like it were staged with crisis actors or denying they occurred at all, may cause additional emotional harm to the survivors, families, and friends of the victims. For instance, following Alex Jones’ continued conspiracy theories regarding the Sandy Hook Elementary School active shooter events, victims’ families were harassed by those who believed the conspiracy theories. Families of the victims were sent photos of dead children, accosted in person, being told that their deceased loved one never existed, and some even received death threats (Collins & Eaton-Robb, 2022).  

Overall, the pro-conspiracy content indicated that many users were trying to provide a causal explanation for the event. Most explanations placed blame on the government or police agencies and stated that gun control was the primary motive for the event. This finding is in line with research suggesting that people are most vulnerable to conspiratorial thinking when trying to explain unusual and unsettling events (Kovic & Füchslin, 2018; Lewandowsky & Cook, 2020; van Prooijen & Jostmann, 2013).   

For those interested in getting through to conspiracy theorists, Lewandowsky and Cook (2020) suggest four research-backed considerations. First, use trusted messengers to deliver information (Schmitt et al., 2018). Second, show empathy and build understanding with the person (Ponsot et al., 2018). Third, be sure to affirm their critical thinking (Voogt, 2017). Lastly, avoid aggressively ridiculing or attacking a conspiracy theory, as the conspiracy theorist will likely immediately reject your arguments (Schmitt et al., 2018).

Researchers and laypeople can and do use these strategies to combat the negative effects of conspiracy theories online. For example, two anti-QAnon subreddits provide support to former and current QAnon conspiracy believers as well as families and friends of those who believe. These subreddits highlight the importance of education and support resources in preventing and discontinuing conspiracy theory beliefs. Policymakers and researchers may be informed by these types of online communities and resources as they continue to work to prevent and counter the impacts of conspiracy theories.

Limitations

The current study only focuses on the conspiracy theories posted on a single, mainstream social media platform. Though this sampling frame is narrow, it provides an initial understanding of the conspiracy theories present on a platform that is frequented by people who may not see these arguments or discussions otherwise, as they likely are not using sites focused on these types of theories. Future research should examine the content on other and more niche sites such as reddit and parler. The content of these discussions across platforms could be assessed for similarities and differences.

Similarly, the decision to focus solely on the tweets containing “#Uvalde” did not allow us to include other conspiratorial tweets that did not contain that hashtag. However, given that the original data set contained over one million observations, limiting the initial examination of the content of conspiratorial tweets to those with the #Uvalde was necessary. Future research could examine whether the content of those conspiratorial tweets without the hashtag differs substantially from those containing the hashtag.

Additionally, when the data were collected, the user information was not able to be collected with the tweets to determine how many unique users were posting these ideas. This may be viewed as a limitation in the generalizability of these findings. However, we provided descriptive statistics for the engagement metrics (likes, retweets, quotes, and replies) for these tweets, which does provide some idea of how many different users were engaging with and reacting to these tweets. While this proxy for the volume of independent users who believe and identify with these ideas is suitable for the purposes of this study, future research may aim to collect user information to identify any trends in the number of pro-conspiracy posters and their popularity. 

Finally, the focus on responses to a single event may be viewed as a limitation. The conditions during this event, including the failure of the lock on the exterior door and the police response, lend themselves to speculation and formulation of alternative explanations. Future research should include responses to several active shooter events to examine whether themes in the content of these theories are consistent across events, and if there are certain event characteristics that make an event more likely to be subject to conspiracy theories.

References

ALERRT. (2022). Robb Elementary School Attack Response Assessment and Recommendations. 

Bail, C., Guay, B., Maloney, E., Combs, A., Hillygus, D. S., Merhout, F., Freelon, D., & Volfovsky, A. (2019). Assessing the Russian internet research agency’s impact on the political attitudes and behaviors of American Twitter users in late 2017. Proceedings of the National Academy of Sciences, 117(1), 243-250. 

Benkler, Y., Casey, T., Bruce E., Hal R., Justin C., Robert F., Jonas K., & Carolyn S. (2020). Mail-in voter fraud: Anatomy of a disinformation campaign. The Berkman Klein Center for Internet & Society Research Publication Series. 

Caelli, K., Ray, L., & Mill, J. (2003). Clear as mud: Towards a greater clarity in generic qualitative research. International Journal of Qualitative Methods, 2(2), 1–23. 

 Collins, D., & Eaton-Robb, P. (2022, September 29). Father: Jones Sandy Hook lies forced move from Connecticut. Associated Press.

Douglas, K. M., & Sutton, R. M. (2023). What are conspiracy theories? A definitional approach to their correlates, consequences, and communication. Annual Review of Psychology, 74, 271-298.  

Douglas, K. M., Sutton, R. M., Callan, M. J., Dawtry, R. J., & Harvey, A. J. (2016). Someone is pulling the strings: Hypersensitive agency detection and belief in conspiracy theories. Thinking and Reasoning, 22, 57-77. 

Enders A. M., Uscinski, J. E., Seelig, M. I., Klofstad, C. A., Wuchty, S., Funchion, J. R., Murthi, M. N., Premaratne, K., & Stoler, J. (2023). The relationship between social media use and beliefs in conspiracy theories and misinformation. Political Behavior, 45, 781-804. 

Enders, A. M., & Uscinski, J. E. (2021). Are misinformation, antiscientific claims, and conspiracy theories for political extremists? Group Processes and Intergroup Relations, 24(4), 583-605.  

Enders, A., Farhart, C., Miller, J., Uscinski, J., Saunders, K., & Drochon, H. (2022). Are republicans and conservatives more likely to believe conspiracy theories? Political Behavior.

Franks, B., Bangerter, A., & Bauer, M. W. (2013). Conspiracy theories as quasi-religious mentality: An integrated account from cognitive science, social representations theory, and frame theory. Frontiers in Psychology, 4(424), 1-12. 

Gau, J. M. (2011). The convergent and discriminant validity of procedural justice and police legitimacy: An empirical test of core theoretical propositions. Journal of Criminal Justice, 39(6), 489-498. 

Gibson, J. J. (1966). The senses considered as perceptual systems. Boston: Houghton Mifflin. 

Goertzel, T. (1994). Belief in conspiracy theories. Political Psychology, 15(4), 731-742. 

Gregory, R. (1970). The intelligent eye. London: Weidenfeld and Nicolson. 

Guess, A., Nagler, J., & Tucker, J. (2019). Less than you think: Prevalence and predictors of fake news dissemination on Facebook. Science Advances, 5, 1-8. 

Imhoff, R., Zimmer, F., Klein, O., António, J. H., Babinska, M., Bangerter, A., ... & Van Prooijen, J. W. (2022). Conspiracy mentality and political orientation across 26 countries. Nature Human Behaviour, 6(3), 392-403. 

Investigative Assistance for Violent Crimes Act of 2012, Pub. L. No. 112-265, § 1, 126 Stat. 2435 (2013). 

Jensen, T. (2013). Democrats and Republicans differ on conspiracy theory beliefs.

Jones, N. M., Thompson, R. R., Schetter, C. D., & Silver, R. C. (2017). Distress and rumor exposure on social media during a campus lockdown. Proceedings of the National Academy of Sciences of the United States of America, 114(44), 11663-11668.  

Kellner, D. (2022). The Uvalde, Texas school shooting massacre. In Educational Philosophy and Theory.  

Kovic, M., & Füchslin, T. (2018). Probability and conspiratorial thinking. Applied Cognitive Psychology, 32, 390-400. 

Kuckartz, U. (2019). Qualitative text analysis: A systematic approach. In G. Kaiser & N. Presmeg (Eds.). Compendium for early career researchers in mathematics education. (pp. 181-197). Springer.

Lachlan, K. A., Spence, P. R., Lin, X., & Del Greco, M. (2014). Screaming into the wind: Examining the volume and content of tweets associated with Hurricane Sandy. Communication Studies, 65(5), 500-518. 

Laine, E. E., & Parakkal, R. (2017). National security, personal insecurity, and political conspiracies: The persistence of Americans’ beliefs in 9/11 conspiracy theories. IUP Journal of International Relations, 11(3), 16-41. 

Lewandowsky, S., & Cook, J. (2020). The conspiracy theory handbook.

Lewandowsky, S., Cook, J.,Oberauer, K., Brophy, S., Lloyd, E. A., & Marriott, M. (2015). Recurrent fury: Conspiratorial discourse in the blogosphere triggered by research on the role of conspiracist ideation in climate denial. Journal of Social and Political Psychology, 3, 142-178. 

Mazer, J. P., Thompson, B., Cherry, J., Russell, M., Payne, H. J., Gail Kirby, E., & Pfohl, W. (2015). Communication in the face of a school crisis: Examining the volume and content of social media mentions during active shooter incidents. Computers in Human Behavior, 53, 238-248.  

McLeod, J. (2001). Qualitative research in counseling and psychotherapy. Sage. 

Merriam, S. (1997). Qualitative research and case study applications in education. Jossey-Bass.  

Merriam-Webster (n.d.). False flag. In Merriam-Webster.com dictionary.  

Nyhan, B. (2020). Facts and myths about misperceptions. Journal of Economic Perspectives, 34(3), 220-236. 

Orosz, G., Krekó, P., Paskuj, B., Tóth-Király, I., Böthe, B., & Roland-Lèvy, C. (2016). Changing conspiracy beliefs through rationality and ridiculing. Frontiers in Psychology, 7(1525), 1-9. 

Patton, M. Q. (2015). Qualitative Research & Evaluation Methods (4th ed.). SAGE. 

Ponsot, A. S., Autixier, C., & Madriaza, P. (2018). Factors facilitating the successful implementation of a prevention of violent radicalization intervention as identified by front-line practitioners. Journal for Deradicalization, 16, 1-33.

Pope, C. (2018, August 31). How a False Flag Sparked World War Two: The Gleiwitz Incident Explained. History Hit.  

Sandel, W. L., & Martaindale, M. H. (2022). What are we talking about? Definitional confusion within active and mass shooting research. Journal of Mass Violence Research, 1(2), 4-16. 

Sapountzis, A., & Condor, S. (2013). Conspiracy accounts as intergroup theories: Challenging dominant understandings of social power and political legitimacy. Political Psychology, 34(5), 731-752. 

Schmitt, J. B., Rieger, D., Ernst, J., & Roth, H.-J. (2018). Critical media literacy and islamist online propaganda: The feasibility, applicability and impact of three learning arrangements. International Journal of Conflict and Violence, 12, 1-19.

Sunshine, J., & Tyler, T. R. (2003). The role of procedural justice and legitimacy in shaping public support for policing. Law & Society Review, 37(3), 513-548. 

Suomalainen, L., Haravuori, H., Berg, N., Kiviruusu, O., & Marttunen, M. (2011). A controlled follow-up study of adolescents exposed to a school shooting–Psychological consequences after four months. European Psychiatry, 26(8), 490-497. 

U.S. Department of Homeland Security (2008). Active shooter: How to respond.  

U.S. Department of Justice: Federal Bureau of Investigation (2022). Active shooter incidents in the United States in 2021.  

U.S. Department of Justice: Federal Bureau of Investigation (2023). Active shooter incidents in the United States in 2022.  

Uscinski, J. E., & Parent, J. M. (2014). American conspiracy theories. Oxford University Press.  

Uscinski, J., Enders, A., Klofstad, C., Seelig, M., Drochon, H., Premaratne, K., & Murthi, M. (2022). Have beliefs in conspiracy theories increased over time? PLoS ONE, 17(7), 1-19. 

van der Linden, S., Panagopoulos, C., Azevedo, F., & Jost, J. T. (2021). The paranoid style in American politics revisited: An ideological asymmetry in conspiratorial thinking. Political Psychology,42(1), 23-51. 

van  Prooijen, J. W., & Acker, M. (2015). The influence of control on belief in conspiracy theories: Conceptual and applied extensions. Applied Cognitive Psychology, 29, 753-761. 

van Prooijen, J. W. (2017). Why education predicts decreased belief in conspiracy theories. Applied Cognitive Psychology, 31(1), 50-58.  

van Prooijen, J. W., & Douglas, K. M. (2017). Conspiracy theories as part of history: The role of societal crisis situations. Memory Studies, 10(3), 323-333.  

van Prooijen, J. W., & Jostmann, N. B. (2013). Belief in conspiracy theories: The influence of uncertainty and perceived morality. European Journal of Social Psychology, 43(1), 109-115.  

van Prooijen, J. W., Krouwel, A. P. M., & Pollet, T. v. (2015). Political extremism predicts belief in conspiracy theories. Social Psychological and Personality Science, 6(5), 570-578.  

Voogt, S. (2017). Countering far-right recruitment online: CAPE’s practitioner experience. Journal of Policing, Intelligence and Counter Terrorism, 12(1), 34-46.

Vosoughi, S., Roy, D., & Aral, S. (2018). The spread of true and false news online. Science, 359(6380), 1146-1151.

Whitson, J. A., & Galinsky, A. D. (2008). Lacking control increases illusory pattern perception. Science, 322, 115-117.

Contributors

Madison K. Doyle is a Doctoral Candidate in the School of Criminal Justice and Criminology at Texas State University. Her research interests include public perceptions of police, individual and organizational health and wellbeing among policing agencies, and police training and evaluation.  

 M. Hunter Martaindale is the Director of Research at the ALERRT Center and an Associate Research Professor within the School of Criminal Justice and Criminology, both at Texas State University. His research interests include improving law enforcement performance/ decision-making, non-law enforcement perceptions of law enforcement functions, and active shooter events.