Saturday, November 16, 2019

Crisis Opportunity Model


Abstract
The crisis opportunity model is demonstrated by the Sandy Hook school shooting, which was a crisis that advocates sought to leverage to spur additional gun control legislation.  This study explores the relationship between the Sandy Hook mass shooting, its media coverage, between media coverage, and public opinion regarding gun control.  The study explores the duration in which people were actively information-seeking to fulfill their need for orientation, including information-seeking of previous shootings.  This study resulted in three original contributions to the communication field.  It was determined that there is a relationship between information-seeking of previous shootings, information-seeking of the Sandy Hook school shooting and support for gun control.  Additionally, it was determined that information-seeking peaked immediately after the event with sustaining intensity for about two weeks, then drastically reduced to pre-crisis levels.  Therefore, the study culminates by presenting a new model incorporating the various theories, called the crisis opportunity model.
Keywords:  crisis opportunity model, framing, shootings, agenda-setting, crisis opportunities, media effects



Introduction
            The crisis opportunity model is a unified graphical depiction of the convergence of crisis communication, media effects, and uncertainty theories in this study. The bottom portion represents the effect of uncertainty and how it drives information-seeking behaviors triggered by the crisis.  The top portion represents the media effects and the historical effects of prior crises, known in media effects as priming.  The primed events become a pattern to which future events will be compared.  When an event occurs, people may think to themselves “I have seen this before.”  The opportunity is the convergence of media coverage and people’s information-seeking behaviors in time and place, finding the information they are looking for, when and where they are looking for it.     
            Figure 1: Crisis Opportunity Model
Information-seeking provides the path to change, while the media represents what people find along that path.  Event-related information-seeking correlated with support for gun control, and the media attention and coverage correlated with those information-seeking events.  This model also aids in explaining Tucson, how leveraging a prior event outside of the optimal time of information-seeking is ineffective as information-seeking and media coverage do not converge in time, and therefore, opportunity was lost.  One element in the model that is not specifically supported in the findings is the aspect of frame acceptance and socialization.  To achieve this, horizontal media would have to be analyzed and this was outside the scope of this study but warrants further study.  This study into the Sandy Hook school shooting supports the belief that a crisis provides an opportunity for social changes and traces a crisis from the event through changes in policy support.
            First, this study sought to define the duration during which people face uncertainty following a crisis, specifically a school shooting.  No previous studies identified the duration of information-seeking following a crisis.  The uncertainty drives information-seeking behaviour; research reveals that the height of information-seeking occurs in the first two weeks and then drastically tapers off, nearly returning to normal in 30 days.  These findings are similar to what Uecker (2011) found with increased church attendance following September 11, 2001, as people were attempting to make sense of the terrorist attack, though not specifically through information-seeking.
The information-seeking behaviors are important not only to uncertainty communication but also for media effects. In addition to identifying the duration of information-seeking specific to the Sandy Hook crisis, this study uncovered information-seeking of prior events spurred by the crisis and potentially the role of the media in activating memory tags of prior events.  This was the second original contribution to the communication field.  While previous research discussed the role of memory tags, this study added to those findings by discovering active information seeking of those prior events to fulfill the need for orientation.
This study resulted in a holistic model unifying numerous theories in understanding how a crisis can be leveraged for change; considered as the third original contribution to the communication field.  There may be a window of opportunity to leverage people’s information-seeking to fulfill their need for orientation.  Tucson may represent an attempt to connect crises after the window of opportunity and failed to spur information-seeking of that prior event regardless of the media coverage.  The media began covering Tucson after many people already processed the Sandy Hook shooting and what it meant to them.  With the ability to monitor real-time information-seeking on Google Trends and socialization of the crisis on social media, special interests can monitor the effectiveness of issue and issue attribute coverage, leveraging data to aid the population in making sense of the world around them.  Such poses ethical challenges and certain responsibilities for the media, the government, and social structures.   
Since this was a single case study, the implications may provide theoretical propositions but do not necessarily extend beyond this case study. In future studies, it would be important to expand the research of other shooting crises, such as the four previous shootings from this case study, to build upon the findings.  The previous shootings may have aided the public in making judgments about the Sandy Hook shooting, and comprehensive research and analysis of those events may yield important insights.  Another limitation of the study was that it included only four vertical media outlets. Additional media outlets, including horizontal media, could yield even more insights.  While the study included major news outlets, it was not inclusive of horizontal media, more personalized media that reflect people’s social network.  Information is a social process and people’s networks aid in emotionally and cognitively processing the information.  People participate in the process on Twitter, Facebook, and a multitude of other sources.  Future research should analyze these other information sources that fall outside of the traditional media, which was the scope of this study.  Even with carefully selected traditional media outlets, it is not representative of all news sources available.
 The crisis opportunity model can be used to detect the manifestation of crises before they happen.  The adoption of such a model is necessary as it sets the pace for defensive mechanisms that can be used to quell issues before they are experienced. Although the model is one of the stances that can be used to improve the state of affairs in any given institution or state, it cannot be effective. Some of the dominant laws and rules that are applicable in the United States have rendered the model ineffective. For instance, the law in gun rights has made it impossible for culprits that quell such stances to be apprehended. The issue of gun rights in the United States has increased in prominence over recent years, especially following mass shootings.  The Democratic Party generally focuses on one aspect of the gun debate, gun control, while the Republican Party focuses on gun rights.  Gun control was once a taboo issue for the Democratic Party but now is a central political issue as a result of the increase in issue salience.  Gun control includes measures such as increasing bans, magazine capacity limitations, weapon registration, background checks, assault weapon bans, and increasing waiting periods for gun sales. 
The crisis opportunity model emphasizes the need to identify the stances that propagate the manifestation of crises in different situations. A crisis is important to this study as it is the focusing event that brings attention to issues and can become an integral force in the political and social factions (Birkland, 1968; Sellnow & Seeger, 2013).  Perry (2007) generally identifies a crisis as “some breakdown in a system that creates shared stress” (p. 17).  Coombs (2007) describes a crisis as the change and adjustments that are anticipated in the case of some random violation of the existing policies and stances. Sellnow, Seeger, and Ulmer (2003) define a crisis as a specific, unexpected, non-routine event or series of events that can create high levels of uncertainty and a significant or perceived threat to high priority goals.  While numerous definitions exist, the definition by Seeger, Sellnow, and Ulmer (2015) represents the meaning for the role of the study.  The crisis can then serve as a focusing event. 
To effectively understand the crisis opportunity model, one can use mass shooting as a focusing event.  A focusing event may spur a need for orientation, whereby people seek information to understand unfolding events and the cause of the crisis.  It is the desirability for effective orientation as one contingent and effective factor for agenda-setting effects by causing the population to seek information to reduce uncertainty (Camaj, 2014).  People with the requirement for desirable orientation are likely to face stances where they are sufficiently exposed to the media, would end up adapting to media (Camaj, 2014; Matthes, 2008; Weaver, 1977).  People select their information source to satisfy their information need based on beliefs relevant to the issue and their predispositions (Camaj, 2014; Mutz & Martin, 2001; Zillmann & Bryant, 1985).  The media selected have the opportunity to begin agenda building, priming and setting through their framing efforts.  Agenda building is the initial step in developing a recognized issue.  Priming includes the ideas that have been left in the memory of the public that will guide future information processing.  Agenda setting refers to the transference of what issues are important in the media’s agenda to the public.
Framing is one of the ways in which the crisis opportunity model can be estimated and evaluated. It refers to how a story is structured and organized in such a way that it tells us how to understand an issue and information (Coombs, 2007).  The policy goals or social change framed within the story connects the crisis with the desired goal or solution (Coombs, 2007; Wolfe, Jones, & Baumgartner, 2013).  As new selected information is available, people can reframe an issue since it joins existing knowledge, providing deeper understanding (Brashers, 2001; Hale, Dulek, & Hale, 2005).  The framing of new information may lead people to interpret an issue or event through a different lens, changing their previous opinion or belief. 
A crisis raises issue salience with a population, but how long does it cause a need for orientation?  The Democratic Party and President Barack Obama sought to capitalize on issue salience following mass shootings, regardless of whether at a school, theatre, or after a terrorist attack.  Mass shooting crises serve as a focusing event, but how long does the population focus on the issue and seek orientation?  What is the relationship of a need for orientation, use of Internet search engines, changes in public opinion on issue attributes, and public policy support?  Understanding such relationships provides a greater understanding of agenda-melding specific to mass shootings. 
            The study on the crisis opportunity model will identify the relationship between media coverage, people’s information seeking, changes in public opinion regarding issue attributes, and policy support following the Sandy Hook mass shooting.  The significance of the study is to understand further how crises can be leveraged to bring about social change, and it begins with understanding how people’s need for orientation may heighten their susceptibility to framing by the media and issue advocates.  The stage in which people have a need for orientation may lead to the first exposure to information (Chernov, Valenzuela, & McCombs, 2011). The orientation from first exposure may influence all future information processing on an issue, a significant advantage toward the represented issue advocate’s perspective.  Horizontal media is an under-researched area but is necessary as individuals are increasingly free to choose their own agendas via the internet. Consequently, fewer people pay regular attention to newspapers and TV news; hence this traditional function of media agenda setting seems likely to decline (Weaver, Wojdynski, Mckeever, & Shaw, 2010, p. 3). 
Literature Review
The crisis opportunity model presupposes that events happen and manifest as prompted by the societal concerns that characterize the given place or region. A crisis and uncertainty are linked, but before their additional theories are explored, definitions of both are necessary.  According to the definition by Sellnow, Seeger, and Ulmer (2003), a crisis creates high levels of uncertainty and a significant or perceived threat to high priority goals.  Therefore, when people believe there is a crisis, they are likely to behave differently than they would in normal situations (Sellnow et al., p. 5).  Crises come in many forms such as a natural disaster, a human-made disaster, terrorist attack, organizational crisis, and economic crisis among others.  The type of crises may help determine the level of uncertainty and the population facing uncertainty.  This study looks at a human-made crisis, a school shooting.  A crisis changes behavior, and uncertainty theories can describe the behavior.  The more people experience high levels of uncertainty caused by a crisis, the more people are potentially seeking information and are susceptible to agenda-setting and framing efforts proceeding from the focusing event to bring about political change.
To perfectly understand the operationalization of the crisis opportunity model, it is necessary to make due regards and reference to the manifestation of stances of uncertainty. Berger and Bradac (1982) identified two types of uncertainty within the uncertainty reduction theory: cognitive and behavioral uncertainties.  Cognitive uncertainties are the questions that people encounter about their beliefs (Knobloch & McAninch, 2014, p. 298).  Behavioral uncertainties are the questions people face about their actions.  Both types of uncertainty can occur simultaneously but not necessarily equally, depending on whether the uncertainty questions beliefs, actions or both.  For example, during and shortly after a mass shooting such as the Sandy Hook incident, people may face more behavioral uncertainty, unsure about what actions they should take.  As people take actions to reduce uncertainty, it allows them to focus on cognitively and emotionally processing the event.  People attempt to make sense of the mass shooting relative to their beliefs.
The problematic integration theory identified two meanings of uncertainty: epistemological and ontological (Babrow, 2001).  Ontological uncertainty refers to the state of being unclear about the nature of the world, concerning the causes and consequences of events (Knobloch & McAninch, 2014, p. 302).  On the other hand, epistemological uncertainty entails questions about the nature of information like the sufficiency, reliability, credibility and structure of knowledge (Knobloch & McAninch, 2014, p. 302).  The two uncertainty types, when applied to the context of a mass shooting, begin with a heavier focus on ontological uncertainty.  However, cognitive processing of information causes the balance to shift toward epistemological uncertainty since people interpret available information based on how they construct reality.  Consequently, both are “interrelated and interconnected in people’s experiences” and a part of a person’s sense-making (Knobloch & McAninch, 2014, p. 302).  The theory of Problematic Integration explores human nature and their desire to understand social circumstances.  Often, people apply knowledge and value judgments in the deriving of meaning, which addresses both behavioral and cognitive uncertainty.
           
The present study employed a single case study approach leveraging mixed methods involving examination of historical polling data, Google Trends data, and traditional news reporting following a mass shooting.  Historical polling data from respective polling companies provides insight into the public’s issue attribute salience on matters surrounding the Sandy Hook school shooting, gun control, and gun rights.  Google Trends provides insights into Google search intensities representing the public’s active information-seeking.  Traditional news reports on the Sandy Hook mass shooting provide insight into how the media reports interpreted the crisis and issue attribute salience for the population.  Each type of data analyzed individually and with the other variables for a period following the crisis provides insight regarding the relationship of the public’s information seeking, media reporting, and public opinion on issue attributes, providing an opportunity to analyze these data to answer the research questions.    
The crisis opportunity model presupposes that once a crisis has been detected, it can be quelled with effectiveness. The Yin (2009) explained that a single case study approach is beneficial if the desire is to understand a complex social phenomenon of a real-life event traced over time.  This study sought to understand the potential agenda-setting effects after the Newtown, Connecticut, school shooting.  This study was designed to explore the phenomenon in depth by incorporating multiple forms of data with a need for triangulation (Yin, 2009).  The timeline provided an opportunity to observe converging lines of inquiry consistent with theoretically guided propositions to understand the impacts of a crisis on a population, impacts of media effects, and the role these have on public opinion.  The ability to trace change over time is a major benefit of a case study (Yin, 2009).  The case study approach is explanatory and exploratory, which may identify correlations over time (Yin, 2009).  A single case study was selected, as the research was testing the relationship of well-established theories within a defined period.  As part of the case study, statistical analysis was employed to answer the research questions.  This was accomplished by developing an intensive timeline incorporating all four elements of data collected to determine correlations in the unfolding events following the Newtown school shooting between people’s need for orientation, issue attribute salience, and
Media attention included the frequency of media articles published regardless of issue attribute, which was also a variable within the Scheufele (2000) study.  The correlation test explored whether relationships existed between the variables.  Additional testing of relationships was needed, including multiple regressions, to determine if a variable predicted the outcome of support for gun control.  The public’s information-seeking used Google Trends data, with each week providing a data point.  The public support for policy action leveraged polling on legislative actions and polls on gun control generally.  The polls were averaged and reported with a single percentile as the data point for the week.  In answering RQ1, that more media attention and increased public information-seeking would be related to support for gun control, the relationship was explored by conducting correlations.  The robustness of the relationship amid the working independent variables was indicated by the fact that the dependent variable happened to be reflected in the existing r value, in such a case, 7 outlining a strong working relationship and .25 highlighting a weak relationship.  As significant relationships were found across the three variables, a multiple regression was done in the effective interest and desirability of getting to terms with the accurate highlights that denote the existing working relationship.








Methodology and Research Design
Introduction
Research design
This research used a case study approach in which historical polling data, Google Trends data, and traditional news reporting following a mass shooting were examined.  The Sandy Hook case study was used to illustrate the outcomes of the present study such that it was possible to come up with working inferences and viewpoints on the point in issue.  Information-seeking scaled data from Google Trends were collected, with each week serving as a data point for Sandy Hook and four previous mass shootings at Aurora, Columbine, Tucson, and Virginia Tech.  The Aurora shooting occurred six months before the Sandy Hook shooting on July 20th, 2012.  The Virginia Tech shooting occurred on April 16th, 2007.  The third shooting occurred less than a year before Sandy Hook in Tucson, Arizona, where Arizona Congressional Representative Gabby Giffords was shot.  The fourth shooting was the Columbine High School shooting that occurred on April 20th, 1999, nearly 13 years before Sandy Hook.
Even though the case study can be used to build theoretical propositions, the research design and findings do not extend beyond the case study.  A timeframe was used to identify the period in which people have a need for orientation following a crisis while monitoring traditional media, polling, and Google Trends for correlations.  One hundred days was the study’s timeframe. The data was validated through corroboration with other sources like peer-reviewed journals whose authors are people of no mean repute. Besides, reports that were used were mainly from government agencies and thus, the information was reliable.
Data Sources and Collection Methods
Data were retrieved from different sources using keywords for instance, gun control. The main source of data used in the research was news analysis, polling data and data from various websites. The information has been used by various scholars and policymakers in different fields and thus reliable.
Variables
The researcher aimed to determine the relationship between the Sandy Hook mass shooting, media coverage, and public opinion regarding gun control.  Good trends, polling data and news reporting are some of the data sources that were used in the research.
Google Trends
Google trends were utilized to explore the intensity of Google search, a representation of information seeking by the public. News reports on the Sandy Hook mass shooting provided insight into how the media reports interpreted the crisis.  Each type of data analyzed individually and with the other variables for a period following the crisis provided insights in regards to the relationship of the public’s information seeking, media reporting, and public opinion on mass shooting.
Polling data
This method was used to determine the acceptance of the public regarding the issue of mass shooting. The polls provided insights into what issue attributes the population accepted.  The polls are organized in a chronological manner depending on how events unfolded. The poll percentage and the questions related to the poll question are included within the timeline.  Polls that surveyed similar questions and issue attributes were averaged. 

News reporting
Different news outlets were selected for information in regards to mass shooting. The media outlets selected include two major cable news outlets and two major newspaper outlets.  The newspapers selected were the highest in circulation at the time of the events.   The first newspaper on the list was omitted, as it had the same parent company as one of the cable news outlets selected, Fox News.  The two newspaper outlets selected as the major source of information were the New York Times and USA Today. Although this approach may not be fully representative of all media, it does remove selection bias.  It also provides an opportunity to trace events or actions post-crisis while monitoring issue attribute reporting.  All relevant articles that met the search criteria of the date range and search string were included in the database and coded, except blogs. 
Data analysis
The data collected from Google Trends were correlated to determine if significant relationships existed between information-seeking of other mass shootings.  The period examined began a week before the crisis when all four search levels were reported as zero, and the period continued until all three previous events returned to zero reported Google Trends search level, December 2, 2012, through February 9, 2013.  A correlation was done to identify the link amid the variables to determine if further testing was necessary.  The correlation indicated whether the relationship was significant, the direction of the relationship (positive or negative), and the strength of the relationship.  The strength of the relationship was reported in the correlation coefficient, r value, with .25 representing a weak relationship and .7 representing a strong relationship. 
The coding included nine separate issue attributes: background checks, assault weapon ban, ammunition ban, expanded background checks, mental health checks, and waiting periods for gun purchase, gun show loopholes, online gun sales, and the Second Amendment.  Each article segment was coded for the issue attributes and then coded whether the article segment included support for, support against, mixed, or neutral.  Each article’s title and first two paragraphs were coded separately and as separate data points. 
The public’s issue attribute salience was identified from polling data with polling percentiles as the data point.  Multiple polls in a week were averaged and reported as a single, averaged data point.  To answer the question on the relationship of information-seeking and the transfer of issue attributes was determined by conducting correlations of information-seeking search terms, including Sandy Hook, NRA, gun rights and gun control; coded issue attribute media coverage in the articles; and related issue attribute polling. 

             









Research Findings
            The role of the study was to explore the crisis opportunity model as illustrated by the use of spearman correlation, Pearson correlation, and chi-square on the Sandy Hook school shooting.  In the previous chapters, literature suggested the exploitable vulnerability of the population following a crisis, especially while people are faced with uncertainty and require orientation.  The media can fulfill the need for orientation and aid people to make sense of reality, both cognitively and emotionally.  Exploring the timeline of people’s information-seeking following a crisis may aid in understanding the time necessary for a population to obtain information, emotionally and cognitively process that information, and prepare to take or support action. The findings represent data from 219 articles from four sources: Fox News, CNN, USA Today, and The New York Times.  Findings also include the analysis of 20 polls from Rasmussen, Gallup, and Pew, in addition to Google Trends scaled data for searches of key terms in the United States on Google.  The data for this study were queried and collected for all data types from December 10, 2016, through January 28, 2017.  To cut down the propensity for type 1 error due to the number of comparisons, the findings for correlations will report results significant at p < .01 with a few exceptions.  The composite variables were excluded and three additional exceptions with significant results reported at p < .05 (see tables 10, 14, and 17).  The three exceptions were between issue attribute polling and other variables.  Without the exceptions, no additional significant variables were found to explore issue attributes.  The composite variables were also exceptions, providing an expanded opportunity to explore relationships of inter-correlated variables with other variables.  
Table 1
Chi-square analysis. Google Trends Search Intensity Data on Four Search Terms
Week
Sandy Hook
Gun rights
Gun control
NRA
0
0
2
1
Dec. 9-15
100
0
3
3
Dec. 16-22
91
0
8
13
Dec. 23-29
12
0
3
4
Dec. 30 – Jan. 5, 2013
8
0
2
2
Jan. 6-12, 2013
7
0
5
4
Jan. 13-19
19
1
10
9
Jan. 20-26
8
0
4
3
Jan. 27 – Feb. 2
5
0
4
3
Feb. 3-9
4
0
3
2
Feb. 10-16
3
0
3
2
Feb. 17-23
3
0
2
2
Feb. 24 – Mar. 2
2
0
2
2
Mar. 3-9
2
0
2
2
Mar. 10-16
2
0
2
2
Mar. 17-23
2
0
2
2
Mar. 24-30
3
0
2
2
Mar. 31 – Apr. 6
2
0
2
2
Apr. 7-13
2
0
3
2
Apr. 14-20
9
0
6
2
Apr. 21-27
2
0
3
2
Apr. 28 – May 4
1
0
2
2
Note. Search intensity data scaled 0–100, with 100 as highest intensity.
The results of the information-seeking terms revealed two peaks, one immediately on the December 14, 2012, Sandy Hook school shooting, and declining significantly a week later. The second peak was identified one month after the incident then declined back to pre-crisis levels.  It appears the duration is consistent with Uecker’s (2011) study.  Uecker found that following September 11, 2001, people attended churches above historical norms to alleviate the uncertainty that peaked two weeks following the attack and returned to historical norms within two months. 
Multiple correlations were found between information-seeking terms.  The National Rifle Association, or NRA, had a significant, positive relationship with two other search terms: Sandy Hook (r = .63, p < .01), and gun control (r = .83, p < .001).  The Sandy Hook search term only had a significant correlation with the NRA search term.  Gun rights and gun control information-seeking correlated with a moderately strong, positive relationship (r = .69, p < .001).  After identifying the populations’ information-seeking, news reporting provided insights to the media coverage.
Table 2
Pearson Correlation Matrix between News Outlets
News outlet
Fox News

CNN

New York Times
R
p
R
p
R
P
CNN
  .67**
.001






New York Times
  .59**
.004

.78***
<.001



USA Today
       .50
.017

.73***
<.001

.73***
<.001
Note. N = 22.
**p < .01 (2-tailed). ***p < .001 (2-tailed).

Table 4
Data Points for Polling Issue Attributes by Week
Week
Support for gun control
For background checks
For assault weapon ban
For high-capacity magazine ban
For gun rights
Mental health
Dec. 2-8, 2012






Dec. 9-15
47





Dec. 16-22
48
92
55
53
42
84
Dec. 23-29
51





Dec. 30 – Jan. 5, 2013
51





Jan. 6-12, 2013
53
86
55
55
45
80
Jan. 13-19
52
85
60


82
Jan. 20-26

91

54


Jan. 27 – Feb. 2






Feb. 3-9
50

55



Feb. 10-16
50
83
56
53
50

Feb. 17-23






Feb. 24 – Mar. 2
52

58



Mar. 3-9






Mar. 10-16






Mar. 17-23






Mar. 24-30






Mar. 31 – Apr. 6






Apr. 7-13






Apr. 14-20
49





Apr. 21-27

78




Apr. 28 – May 4
50

54




Table 5
Pearson Correlations Among Information Seeking, Media Outlet, and Support for Gun Control
Variable
Fox News
CNN
New York Times
USA Today
Composite Media Attention
Support gun controla
Sandy Hook IS






Pearson r
.28
.46
   .60**
  .53
.51*
 -.74**
P
.211
.031
 .003
 .012
.015
 .010
Gun rights IS






Pearson r
 .53
.20
 .33
 .28
.41
 .32
P
.011
.379
 .130
 .206
.060
 .338
Gun control IS






Pearson r
  .68**
.68***
   .80***
   .54**
.79***
 .00
P
.001
<.001
 <.001
 .010
<.001
 .991
NRA IS






Pearson r
  .72***
.83***
   .89***
   .82***
.93***
-.14
P
<.001
<.001
 <.001
 <.001
<.001
 .675
Support gun controla






Pearson r
.22
-.02
-.34
-.20
-.05

P
.519
.956
 .313
 .563
.877

Note. N = 22 except where indicated. IS = information-seeking; NRA = National Rifle Association.
aN = 11.
**p < .01 (2-tailed). ***p < .001 (2-tailed).
Results for RQ2
            RQ2 discusses how the crisis opportunity model can be tethered to meet the requisite outcomes that include estimating how effective the model is as far as the pragmatic functioning of the respective elements is placed. The events that were considered in this case include the Columbine, Aurora, Tucson, and the Virginia Tech shootings, might have been leveraged by people information-seeking to make evaluations and appraisals of the Sandy Hook shooting.  If previous events were used to evaluate the Sandy Hook shooting, did it influence support for gun control?  As part of determining if previous events were leveraged to make sense of the crisis, a correlation was conducted between Sandy Hook and information-seeking regarding previous shootings, using Google Trends data.  The previous shootings might have served as cues and memory tags to process the Sandy Hook shooting even before all the information was available.  Google Trends was queried on January 19, 2017, for the prior shooting incidents.  The data were examined to determine if there existed a link amid information-seeking for Sandy Hook and the three shooting events.  Table 6 presents the Google Trends data returned from the query, with scaled data points from 0-100, with 100 representing the highest level of information seeking for the previous shooting terms.
Table 6
Google Trends Search Intensity Data on Three Previous Mass Shootings
Week
Aurora
Virginia Tech
Columbine
Dec. 2-8, 2012
0
0
0
Dec. 9-15
1
1
4
Dec. 16-22
2
1
5
Dec. 23-29
0
0
1
Dec. 30 – Jan. 5, 2013
1
0
1
Jan. 6-12, 2013
1
0
1
Jan. 13-19
1
0
1
Jan. 20-26
0
0
1
Jan. 27 – Feb. 2
0
0
1
Feb. 3-9
0
0
0
Note. Search intensity comparatively to Sandy Hook information-seeking.  Data scaled 0–100, with 100 as the highest intensity.
Previously in the chapter, information-seeking was explored for gun control, NRA, gun rights, and Sandy Hook. To better appreciate the influence of information seeking, previous shooting information-seeking was explored with the additional variables of coded articles referencing prior shootings and support for gun control polling. 
Table 7
Data Points for Articles Mentioning Previous Shooting Incidents, by Week and News Outlet
Week
New York Times
USA Today
CNN
Fox
News
Total with Tucson
Total without Tucson
Dec. 2-8, 2012
0
0
0
0
  0
  0
Dec. 9-15
0
0
2
6
  8
  8
Dec. 16-22
0
6
8
6
20
20
Dec. 23-29
0
0
0
1
  1
  1
Dec. 30 – Jan. 5, 2013
0
0
0
0
  0
  0
Jan. 6-12, 2013
4
6
6
3
19
  7
Jan. 13-19
6
0
0
0
  6
  6
Jan. 20-26
0
0
0
2
  2
  2
Jan. 27 – Feb. 2
0
0
6
0
  6
  0
Feb. 3-9
0
0
0
2
  2
  2
Feb. 10-16
0
0
4
0
  4
  0
Feb. 17-23
0
0
0
0
  0
  0
Feb. 24 – Mar. 2
0
0
0
0
  0
  0
Mar. 3-9
0
0
0
5
  5
  0
Mar. 10-16
0
0
0
6
  6
  4
Mar. 17-23
0
0
0
2
  2
  2
Mar. 24-30
0
0
0
2
  2
  0
Mar. 31 – Apr. 6
0
0
0
0
  0
  0
Apr. 7-13
0
0
5
4
  9
  2
Apr. 14-20
2
0
0
0
  2
  2
Apr. 21-27
0
0
0
0
  0
  0
Apr. 28 – May 4
0
0
0
0
  0
  0
Total
12
12
31
39
94
56

Table 8
Pearson Correlation Matrix Between Searches for Previous Mass Shooting
Search term
Aurora

Virginia Tech
R
p
r
p
Virginia Tech
.68
.031



Columbine
  .77**
.009

.96***
.000
Note. N = 10.
**p < .01 (2-tailed). ***p < .001 (2-tailed).
Additional correlations were conducted between the two groups of information-seeking, with three significant findings (see Table 9).  Sandy Hook information-seeking significantly correlated (p <.001) with Columbine and Virginia Tech shooting events and the composite variable, including the three previous shootings combined into a single variable.  The composite variable was also found to have significant relationships with gun control and the NRA, which represent all the issues that relate to the gun debate. The gun debate is one aspect that can be used to illustrate the effectiveness and operationalization of the crisis opportunity model. Given the significant relationship between the information-seeking of similar events, those previous events might have aided in making judgments of the Sandy Hook crisis. 
Table 9
Pearson Correlations between Search Terms and Previous Shooting Searches
Search term
            Aurora

Virginia Tech

Columbine
Composite IS
R
p

R
p

r
p
r            p

Sandy Hook
 .70
.025

 .99***
<.001

 .96***
<.001
 .95***  <.001

Gun rights
.20
.578

-.17
.645

-.11
.770
.10         .650

Gun control
.58
.079

 .22
.541

 .36
.310
.52*       .013

National Rifle Association (NRA)
.75
.012

 .51
.132

 .65
.041
.76***  <.001

















Note. N = 10.
*p < .05, **p < .01 (2-tailed). ***p < .001 (2-tailed).
Earlier, we explored the relationships between media attention, public information-seeking, and support for gun control.  The testing was repeated using previous events’ information seeking.  Two direct correlations with support for gun control were found, including Virginia Tech and Columbine (see Table 10).  No significant relationships were found between Virginia Tech and Columbine information-seeking and media attention of the media outlets in the study.  Sandy Hook information-seeking was previously found to be linked with The New York Times and the composite media attention variable (see Table 5).  Sandy Hook information-seeking correlated with the previous shooting composite variable strongly (see Table 9).  The media attention composite variable ties in with the previous shooting information-seeking composite variable and with each of the three previous shootings (see Table 10). 
Table 10
Pearson Correlations Among Previous Shooting Info Seeking, Media, and Support Gun Control
Fox News
CNN
New York Times
USA Today
Composite Media
Support gun controlb
Composite ISa






Pearson r
   .44*
    .63**
  .71**
     .61**
    .67**
-.55
p
.038
 .002
 <.001
 .002
.001
.078
Aurora






Pearson r
.62
 .50
 .76
  .69
    .73**
-.31
p
.055
 .145
 .011
 .027
 <.001
 .504
Virginia Tech






Pearson r
.15
 .34
 .57
 .57
    .69**
   -.89**
p
.680
 .342
 .089
 .084
.006
 .007
Columbine






Pearson r
.30
 .53
 .71
  .71
    .81**
 -.78*
p
.404
 .119
 .022
 .021
  <.001
 .038
Note. N = 20 except where indicated. aN = 11. bN = 7. *p < .05 (2-tailed). **p < .01 (2-tailed).
The analysis explored whether information-seeking surrounding previous shooting events was linked to the functionality factions that denote the operationalization of gun control. A strong, negative relationship was found between Virginia Tech information-seeking and the dependent variable of support for gun control, r =
-.89, p < .01, meaning that as information-seeking for Virginia Tech decreased, support for gun control increased.  A strong, negative relationship was also found between Columbine information seeking and support for gun control, r = -.78, p < .05.  As people reduced their information-seeking and fulfilled their need for orientation, some people supported additional gun control.  Given the relationship of Columbine with support for gun control, this relationship was further explored as an exception to requiring a significance level greater than or equal to .01. 
Table 11
Regression Analysis of Information-Seeking Terms Sandy Hook and Composite of Previous Events on Dependent Variable of Support for Gun Control
Variable
Unstandardized coefficients

Standardized coefficients
t
p



B
SE
Beta
F
p
Adj R2
Overall Model






10.604
.006**
.658
Sandy Hook IS
  -0.096
0.027

 -1.940
-3.509
.008**



Composite IS
0.848
0.367

1.277
 2.309
  .05*



Note: Linear Regression with aid for Gun control stances and attributes as the working dependent variable.
            The previous events information-seeking and media attention relationship with support for gun control were analyzed, with several significant findings.  Next, correlations were conducted between articles coded that included references of previous shooting events by CNN, Fox News, and in aggregate of all four media outlets.  USA Today and The New York Times yielded insufficient data points to be analyzed independently.  Additionally, information-seeking for Sandy Hook and support for gun control were included in the correlations.  Previous shootings included four events: Columbine, Virginia Tech, Aurora, and Tucson.  Tucson was excluded in previous testing because of no Google Trends data resulted from the query, precluding it from other testing.
Table 12
Pearson Correlations Among Previous Shooting Information-Seeking, Media Outlet, and Support for Gun Control
Variable
With all four shootings (with Tucson)

Without Tucson
Fox News
CNN
Media events aggregate
Media events aggregate
Sandy Hook IS





Pearson r
.63**
  .46
.55**

.81***
p
.002
 .032
.007

.000
Support for gun controlb





Pearson r
-.58
-.31
-.10

-.41
p
.059
 .349
.774

.210
Aurora IS





Pearson r
.61
 .53
.74

.86**
p
.063
 .113
.014

.002
Virginia Tech IS





Pearson r
.89***
 .53
.54

.80**
p
.000
 .115
.108

.006
Columbine IS





Pearson r
.86**
 .63
.64

.86**
p
.002
 .050
.049

.002
Composite IS





Pearson r
.59**
.57**
.68**

.89**
p
.004
.006
<.001

<.001
Note. N = 22 except where indicated. IS = information-seeking.
bN = 11.
**p < .01 (2-tailed). ***p < .001 (2-tailed).
Given the significant relationships between media coverage and information-seeking terms, Columbine, Virginia Tech, Sandy Hook, and the composite variable, regressions were conducted to determine the effects on support for gun control.  The regressions were not significant (p > .05).
The research sought to determine if the public’s information-seeking (Independent Variable 1) associates with the transfer of issue attributes from the media (Independent Variable 2) to the public (the dependent variable).  Nine issue attributes were initially coded: weapon, ammunition and magazine bans, background checks, mental health, waiting periods, gun show loopholes, online gun sales, and Second Amendment.  Consistent polling resulted in the ability to test only five of the gun debate issues: assault defense mechanisms as well as the high-capacity magazine bans, mental health, background checks, and the Second Amendment.  Excluded issues of waiting periods, gun show loopholes, ammunition bans, and online gun sales each had two or fewer polling data points for the period covered in this study.  Of the issues explored, specific issue attributes were further limited due to data limitations, such as lack of coded or polling data.  The included issue attributes coding was tested for inter-coder reliability using Krippendorff’s Alpha with the following results: for assault weapon ban .926; against assault weapon ban .97; for the second amendment .937; for high volume magazine ban .985; for intermediary checks .985; for mental adjustments and checks.
Additional significant correlations were found between the independent variables, including articles coded in support of background checks and information-seeking for gun rights, r = .56, p < .01; gun control, r = .75, p < .001; and NRA, r = .74, p < .001 (see Table 13).  The coded articles and the polling for background checks did not return any significant relationships with previous event information-seeking.
Table 13
Pearson Correlations Among Background Check Data
Data
Background
poll

Background checks aggregate

Against background checks

For background checks
R
p
r
p
r
p
r
p
Sandy Hook IS
 .63a
.181

.13
.554

 -.16
.489

.26
.241
Gun rights IS
-.08a
.882

.44
.040

-.11
.620

.56**
.007
Gun control IS
  .41a
.422

.64**
.001

 .05
.814

.75***
.000
NRA IS
  .58a
.231

.55**
.008

-.21
.348

.74***
.000
Background polla



.21
.685

       


.21
.685
Aurorab
.12
.884

.56
.095

     -.30
   .397

.60
.065
Virginia Techb
.66
.336

.15
.688

     -.17
.645

.17
.638
Columbineb
.66
.336

.29
.489

     -.32
.368

.30
.407
Note. N = 22 except where indicated. IS = information-seeking; NRA = National Rifle Association.
aN = 6. bN = 10.
**p < .01 (2-tailed). ***p < .001 (2-tailed).
            Correlations between issue attributes were also explored with one significant finding.  Background checks articles in aggregate correlated significantly with articles in support of mental health background checks, r = .72, p < .001.  Articles in support of background checks correlated with articles in support for a high-capacity magazine ban, r = .86, p < .001; and articles in support for a ban on assault weapons, r = .75, p < .001.  Assault weapons are explored next.  
Table 14
Pearson Correlations among Assault Weapon Ban Data
Data
Assault weapon ban poll

Assault weapon ban aggregate

For assault weapon ban

Against assault weapon ban
r
p
r
P
r
p
r
p
Sandy Hook IS
-.10a
.829

.32
.144

.35
.106

 .13
.574
Gun rights IS
   .80*a
.029

.46
.031

.59**
.004

 .19
.392
Gun control IS
  .48a
.273

.71***
.000

.76***
.000

 .45
.036
NRA IS
  .22a
.641

.79***
.000

.82***
.000

 .49
.022
Assault weapon ban poll



.34a
.458

.48a
.279

-.05a
.910
Aurora



.60b
.069

.54b
.136

 .20b
.579
Virginia Tech
-.33c
.667

.14b
.711

.06b
.871

-.05b
.897
Columbine
-.23c
.775

.36b
.309

.23b
.554

 .15b
.676
Note. N = 22 except where indicated. aN = 7.  bN = 10. cN = 4.
IS = information seeking; NRA = National Rifle Association.
*p<.05,  **p < .01 (2-tailed). ***p < .001 (2-tailed).
            Assault weapon ban coded articles in aggregate, in support of, and against had significant correlations with numerous other issues’ attributes not previously reported.  The assault weapon issue in aggregate correlated significantly with the Second Amendment in aggregate, r = .69, p <.001; articles in support for the Second Amendment, r = .67, p = .001; and articles on high-capacity magazines in aggregate, r = .77, p < .001.  Articles coded in support of the assault weapon ban correlated with the Second Amendment articles in aggregate, r = .55, p < .01; articles on high-capacity magazines in aggregate r = .75, p < .001; and articles support for high-capacity magazines, r = .84, p < .001.  Articles coded against the ban also correlated significantly with articles in support for the Second Amendment, r = .64, p = .001; and articles in support of high-capacity magazines in aggregate, r = .60, p < .01.
Table 15
Pearson Correlations among Data Regarding High-Capacity Magazines
Data
High-capacity magazine ban poll

High-capacity magazine aggregate

For high-capacity magazine ban

Against high-capacity magazine ban
r
p
r
P
r
p
r
p
Sandy Hook IS
-.49a
.511

.25
.267

.65*e
.011

  .05
.820
Gun rights ISi
-
-

.43
.049

.69**e
.007

-.14
.529
Gun control IS
-.16a
.839

.65**
.001

.89***e
.000

-.13
.571
NRA IS
-.38a
.622

.70***
.000

.86***e
.000

  .12
.603
Aurora
-.50f
.667

.63g
.051

.75h
.050

  .52g
.125
Virginia Tech
-.87f
.333

.14g
.705

.46h
.294

  .55g
.102
Columbine
-.87f
.333

.29g
.416

.52h
.237

  .55g
.100
Note. N = 22 except where indicated. IS = information-seeking; NRA = National Rifle Association.
aN = 4. dN = 21. eN = 14. fN = 3. gN = 10. hN = 7.
iGun Rights IS and polling data was insufficient for testing.
**p < .01 (2-tailed). ***p < .001 (2-tailed).
            The articles coded for the high-capacity magazine issue attribute in aggregate and in support of the ban had significant correlations with numerous other issue attributes not previously reported.  The issue attribute of high-capacity magazines articles correlated significantly with the Second Amendment articles in aggregate, r = .61, p < .01; support for the Second Amendment articles, r = .59, p < .01; and support for mental health background checks articles, r = .58, p < .005.  Articles in support for the high-capacity magazine ban correlated significantly with support for mental health background checks articles, r = .66, p < .01.  
Table 16
Pearson Correlations Among Data Regarding the Second Amendment
Data
Second Amendment poll

Second Amendment aggregate

For Second Amendment

Against Second Amendment
r
p
r
P
r
p
r
p
Sandy Hook IS
-.81a
.399

.11b
.630

.11b
.646

  .05b
.815
Gun rights IS



.36
.096

.39
.073

-.09
.688
Gun control IS
-.97a
.164

.50
.018

.51
.016

 .07
.745
NRA IS
-.88a
.315

.57**
.005

.57**
.005

 .13
.581
Aurora



.56c
.090

.53c
.119

  .30c
.397
Virginia Tech



-.02c
.959

.01c
.986

-.15c
.671
Columbine



.14c
.703

.15c
.689

-.03c
.946
Note. N = 22 except where indicated. IS = information-seeking; NRA = National Rifle Association.
aN = 3. bN = 21. cN = 10.
**p < .01 (2-tailed).
            Mental Health.  Only one significant correlation was found between the independent variable of NRA information-seeking and the dependent variable of mental health checks polling, p < .05, r = .998 (see Table 17).  While the significance level is below the p < .01 threshold, it is one of the limited relationships between information-seeking and issue attribute polling that is discussed further in Chapter 5.  There was not a significant relationship between polling and articles in support of mental health checks; therefore, a multiple regression was not conducted.  Gun control, r = .57, p < .01, information-seeking had significant relationships with articles coded in support of mental health checks.
Table 17
Pearson Correlations among Data Regarding Mental Health Checks
Data
Mental health poll

For mental health checks
ra
P
R
p
Sandy Hook IS
.92
.249

.47b
.034
Gun rights IS



.22
.323
Gun control IS
.60
.593

.57**
.006
NRA IS
  .998*
.041

.74***
.000
Aurora
.87
.333

.63c
.049
Virginia Tech
.87
.333

.53c
.118
Columbine
.87
.333

.58c
.080
Note. N = 22 except where indicated. IS = information-seeking; NRA =
National Rifle Association.
aN = 3. bN = 21. cN = 10.
*p < .05 (2-tailed). **p < .01 (2-tailed). ***p < .001 (2-tailed).
            The analysis also explored relationships between mental health polling, articles coded on the issue attribute of mental health, and previous shootings.  From the correlations, no additional significant relationship emerged when expanded to include previous shootings (see Table 17).    Mental health issue attribute correlations with other issue attributes were covered in the previous attributes. 
Throughout analysis of the data addressing RQ3, only two issue attributes’ polling had a significant relationship with one of the independent variables and both were an exception to the p < .01 threshold: support for mental health background checks polling and NRA information seeking (see Table 17), and the support for assault weapon ban poll and gun rights information seeking (see Table 14).  The assault weapon and mental health checks polling both also significantly correlated with the articles coded for that issue attribute.  These relationships could indicate second level agenda setting.  As the issue attributes were examined together more closely, a pattern emerged of articles coded in support of issue attributes (for assault weapon ban, for high capacity magazine ban, for background checks, for 2nd Amendment, and for mental health checks) significantly correlated with NRA information-seeking.  The NRA was the only information-seeking term that was not a previous shooting event to correlate with Sandy Hook information-seeking, which had significant relationship with support for gun control.  As a result, the four issue attributes aligned with the themes in support for more gun control were combined into a composite variable with a Cronbach’s Alpha .855.  The composite variable included for assault weapon ban, for high capacity magazine ban, for background checks, and for mental health checks.  The issue attribute for Second Amendment was excluded since it was not an additional gun control being sought.
Table 18
Pearson Correlations among Composite Media Issue Attribute Data in Support of Gun Control 
Variable
Sandy Hook IS
VT IS
Columbine IS
Previous IS Composite
Composite Media Attention
Support for Gun Control
Composite Media Issue Attribute






Pearson r
   .80***
    .69**
  .81***
      .83***
    .51*
-.31
P
.001
 .006
 <.001
<.001
  .015
  .502

Discussion
            School shootings have been central in the gun debate over the past two decades, with crises refocusing attention to issues surrounding this controversial topic.  A crisis introduces uncertainty that drives information-seeking behaviors.  People are more susceptible and vulnerable following a crisis, as it forces people to seek information, a story to cognitively and emotionally process what happened.  A crisis is important as it is a focusing event that brings attention to issues that can drive social and political change (Birkland, 1968; Sellnow & Seeger, 2013).  However, a holistic examination of how and why a crisis can be used to achieve change is necessary.  This study built upon an extensive body of knowledge spanning multiple disciplines and theories reviewed in Chapter 2.  It considered uncertainty, crisis communication, and media effects theories.  The Sandy Hook school shooting provided an opportunity to understand how a crisis can be used to drive change.  The study was designed to fill current gaps of knowledge to understand how crises are leveraged to bring about change and has implications for numerous theories.       
The literature review explored uncertainty and how it influences people’s information-seeking, requiring the least cognitive effort that alleviates the stress associated with uncertainty (Berger, 1997; Knobloch & McAninch, 2014).  Also, media effects theories were discussed, including agenda setting and framing.  These theories provided the framework of how and what information from the media fulfilled people’s information-seeking requirements that may influence which issues and issue attributes are important to people.  In Chapter 3, the study was carefully designed to understand people’s information-seeking behaviors and the media coverage of the Sandy Hook crisis to understand the influence the media and issue advocates can have on policy support following a crisis.  Chapter 4 presented the findings of the three research questions, including various types of data such as Google Trends, polling data, and media coverage.
“Priming occurs when news content suggests to new audiences that they ought to use specific issues as benchmarks for evaluating the performance of leaders and governments. The success of priming is dependent upon how easily an issue or association comes to mind to evaluate like or related issues. The ease of issues coming to mind is why priming is considered an accessibility model as is agenda setting.  Framing, an applicability model, may assist in drawing associations and linkages to previous issues. 
A focusing event such as a crisis can refocus attention to an issue, providing an opportunity to introduce or amplify frames by the media, issue advocates, or other social structures. A crisis typically provides this opportunity because the event needs interpretation, and framing has an important role in developing associations. Research considered a framing effect as the organization, selection of content, and thematic structure rendering particular thoughts that activate a knowledge structure applicable to the desired schema for evaluations.  Under this model, it is difficult to distinguish framing effects from accessibility counterparts within agenda setting.  As a result of these challenges, Cacciatore provided a new perspective, the preference-based effects models.  Priming is the outcome of the message and whether the audiences accepts the connection between two or more concepts. This is achieved by people receiving information from their social networks or from a fragmented news environment representing a person’s predisposition and cognitive schemas. Within the larger agenda setting construct, story selection increases the public’s perception of issue salience while framing is a particular way the issue is presented. Framing is considered an extension of agenda setting when selecting and focusing on a restricted number of thematically related attributes when discussing an object within the media agenda to increase the salience of issue attributes.
Framing is a story structured and organized that tells us how to understand an issue, events, and information. It is how we think about an issue though interpretive schemas and descriptions of an issue. Framing policy goals or social change within the story connects the crisis with the desired goal or solution. The issue advocate’s goal within a social structure and the crisis in which people are seeking information must be related and connectable even though an actual connection is not necessary.  Cues and mental tags established in agenda setting and priming allow and increase the opportunity that the framing efforts will meet the demands for coherence by activating previous information that guides how the new information is processed.  Emotional responses can shift over time when new information is available, which allows people to reframe an issue as new information joins existing knowledge, providing a more complete understanding.
In 1968, Birkland considered a crisis as a way to bring attention to issues and to set the larger public policy agenda.  Individuals select frames to interpret an issue that may influence them to support movements, reach a consensus, and lead to mobilization for action and social change. Frames tell people what issues are important (Littlejohn & Foss, 2011).  The opportunity is the result of an uncertain population that is susceptible to framing since fewer channels of information are available or were at that time.  The new information can be framed within the agenda to influence the population to interpret the events through a specific lens.  Uncertainty can be manipulated in a desired direction through the information provided to the population in order to influence them. As long as the frame allows people to develop meaning for an event, comes from a credible source, and holds together coherently, the crisis may be framed in a way that may serve a policy objective.  Framing a crisis must both meet the needs of and be relevant to the public. A story defines the important issues and highlights the need for change.  The crisis provides the opportunity to define or redefine an issue during a time the population is vulnerable and susceptible to information. This study explores the relationships between the Sandy Hook mass shooting, its media coverage, related Google searches, public opinion, and policy actions.  Additionally, it seeks to identify a time period in which people nationally had a need for orientation and actively sought information following the Sandy Hook school shooting.  This and the literature reviewed lead to three research questions.  This study seeks to extend our understanding of agenda setting following the crisis of the Newtown, Connecticut, school shooting on December 14, 2012.  Gun control has been an agenda of the Democrat Party for more than 20 years.  A crisis presents a unique opportunity to achieve policy actions, given the population’s uncertainty and information-seeking tendencies.
            In this final chapter, we explore the findings from Chapter 4 of this study and the implications for communication scholars and our understanding of uncertainty, media effects, and crisis communication theories.  This was achieved by exploring relationships between the Sandy Hook mass shooting, media attention, media coverage of related issue, public opinion, and the public’s information-seeking following the crisis.  This chapter concludes with discussing the limitations of this study and considerations for future research. 
Numerous significant relationships were found between information-seeking terms.  While looking at only the scaled numbers, the complete scope of the number of people information-seeking can be lost. It is advisable to look out for the stances that characterize the crisis opportunity model (Tversky & Kahneman, 1973).  The span of the information-seeking is an important consideration, as the scaled number represents information search intensity of a term or incident over the lifetime of Google Trends or since the event occurrence, though without the raw data of actual unique searches, the full scope of the meaning is limited.  Gun rights, gun control, and NRA correlated with each other though were not statistically related to eventual support for gun control.  Information-seeking of shooting events correlated with each other and corresponded with increased support for gun control. Only information-seeking of the NRA correlated significantly for both sets of information-seeking terms, potentially representing the role the NRA plays as an issue advocate, a social structure people rely on to understand the implications of a crisis. 
While gaining a better appreciation of information-seeking, only three of the seven individual information-seeking terms following the Sandy Hook crisis have a significant relationship with support for gun control, with the composite variable making a fourth; Sandy Hook (see Table 5), composite previous shooting information seeking variable (see Table 11), Virginia Tech and Columbine (see Table 10).  As we discuss further how crises are leveraged for policy, political, and social change, it is evident that prior events play a role in influencing support for change.  In part, this can be explained by priming, a component of agenda setting.  Camaj’s (2014) findings that the requirement for orientation is a key factor that is often required in the case of agenda setting. This is further supported and entrenched by these findings, as information-seeking behaviors can influence policy support.  This in turn supports the theory that people are more susceptible to change following a crisis because it spurs the need for orientation to fill the information void. 
After focusing on the population following a crisis, another important element in the equation is the media that aid the population in understanding and interpreting the events around them.  In exploring the media’s role in driving change as a result of a crisis, media effects theories aid in exploring the relationship.  The population has a need for orientation, seeking information about the crisis and previous crises.  Next, we discuss the role the media played and what people found from four major news outlets as they sought information.    
According to Eagly and Chaiken (1993), media effects in this study mostly considered the agenda setting theory and models of information processing. Media generated symbols of a focus event, in this case previous shootings, may serve as evidence for gun control advocates for a dramatized need for change (Birkland, 1998).  Scheufele (2000) considered priming to influence future agenda setting efforts by influencing how people make evaluations.  Several findings on priming have already been reported in this chapter, specifically how prior events are used to judge and evaluate future information in addition to spurring information-seeking of those prior events. The initial lack of coverage of Tucson also supports the role of need for orientation as a contingent factor.  The Tucson coverage began after people no longer had a need for orientation, a month after the shooting.  Both Virginia Tech and Columbine were covered during the height of information-seeking, with media coverage focused on those two events during the same period.  In examining Tucson specifically, the lack of relationship with linking stances that are associated with the issue of gun control was potentially the result of media coverage not occurring until after the need for orientation had passed, as observed in the information-seeking data.  The media covered Tucson as people were tuning out.  Sandy Hook was on the minds of the American people during this time with numerous significant relationships between information-seeking, and media coverage. 
Caccitore (2016) found that media attention enables the information to be more available and accessible for decision making. Of the four event-related information-seeking terms that had a significant relationship with support for gun control, all had significant relationships also with the composite media coverage, the composite media coverage of prior shootings, and the composite media gun control issue attributes variables (see Table 18).  The crisis information seeking has the more significant relationship with gun control than other information-seeking terms and all other variables in this study.  This study included examining media attention of the crisis, and the articles were then coded to understand the related issue attributes. According to Vu & McCombs (2014),  it is important to point out that the crisis opportunity model serves to show the strategies that can be inculcated to create a stance where crises and other manifesting concerns are resolved as soon as they happen.
Framing provides people a schema with which to interpret and classify information encountered in life and to provide meanings of events (Goffman, 1974).  In this study, issues represent the themes in the ongoing gun debate’s competing narratives, where each side of the debate attempts to interpret meaning of recent events.  Each side of those issues are issue attributes that advocates are attempting to advance as part of their narrative.  Gun control and the NRA represent the two broader, competing narratives in the gun debate.   In this study, issue attributes for background checks, for assault weapon ban, for high capacity magazine ban, and for mental health background checks correlated with both information-seeking terms NRA and gun control.  The issue attribute for the Second Amendment correlated only with the NRA information-seeking activity.
The term NRA was found to have a significant relationship to Sandy Hook and with each of the five-issue attribute coded articles mentioned above.  As a result of the linkage, a composite variable was created with the four issue attributes in support of additional gun control.  The composite issue attributes variable significantly correlated with the Sandy Hook, Virginia Tech, Columbine, and the previous shooting composite variable for information-seeking (see Table 18).  While linkages were found between issue attributes and information-seeking, there was no direct relationship between issue attributes and support for gun control.  The timing of information-seeking surrounding related prior shootings correlated with media articles on the prior shootings (see Table 12), potentially demonstrating the activation of memory tags by the media of primed events, both peaking within the first two weeks of the crisis.  
There was one issue attribute, for mental health checks, with a significant relationship between issue attribute polling and articles coded in support of that attribute, possibly reflecting second level agenda setting.  Mental health background checks were an area that both NRA and gun control advocates supported during the period of this study, which may be why this issue attribute may have achieved second level agenda setting.  No legislation of mental health background checks passed during the period covered in this study.  

Conclusion
In understanding what people think about, we looked toward the issue attributes polling and where people stand relative to the articles that were published by the four media outlets.  Issue attributes were examined to see the influence that may exist with support for gun control more broadly. When looking at those variables more closely, a pattern emerged.  Each of those variables had significant relationships with coded articles of previous events and the composite variables of media attention, media issue attributes for gun control, and the previous shooting information-seeking.  The relationships show the convergence of information-seeking and media coverage, a necessary condition to achieve change.  When people looked to the media, they found numerous issues covered with most issue attributes for additional gun control.
It is important to maintain perspective on the change that took place, which were minor changes in the overall support for gun control. What made it more significant was exceeding the 50% threshold - a major narrative win.  The increase in support of gun control would not likely have occurred without the Sandy Hook school shooting crisis to refocus the attention of the population and require people to reevaluate their positions.  Crises have been long believed to be able to influence change, and this study looked at the Sandy Hook school shooting to understand it. Specific to this study, the opportunities for change were realized with increased support for gun control and mental health background checks though gun control advocates were unable to achieve legislative success at the federal level.  The model reflects the relationships found in this study.  As people were seeking information about the crisis and related crises to make sense of the world, the media was providing information to meet their needs.  The information people encountered were references to previous shootings and issue attributes that mostly supported more gun control.  As people concluded their information-seeking, some people shifted to support gun control.





















References
Babrow, A. (2001). Uncertainty, value, communication, and problematic integration. Journal of Communication, 51, 553-573. doi:0.1111/j.1460-2466.2001.tb02896.x
Berger, C., & Bradac, J. (1982). Language and social knowledge: Uncertainty in interpersonal relationships. London: Edward Arnold.
Birkland, T. (1998). Focusing event, mobilization, and agenda-setting. Journal of Public Policy, 18(1), 53-74.
Brashers, D. E., & Hogan, T. P. (2013). The appraisal and management of uncertainty: Implications for information-retrieval systems. Information Processing and Management(49), 1241-1249.
Cacciatore, M., Scheufele, D., & Iyengar, S. (2016). The end of framing as we know it . . . and the future of media effects. Mass Communication and Society, 19, 7-23. doi:10.1080/15205436.2015.1068811
Camaj, L. (2014). Need for orientation, selective exposure, and attribute agenda-setting effects. Mass Communication and Society, 17, 689-712. doi:10.1080/15205436.2013.835424
Chernov, G., Valenzuela, S., & McCombs, M. (2011). An experimental comparison of the two perspectives on the concept of need for orientation in agenda-setting effects. Journalism and Mass Communication Quarterly, 88, 142-155.
Coombs, W. T. (2007). Attribution Theory as a guide for post-crisis communication research. Public Relations Review, 33, 135-139.
Eagly, A., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth: Harcourt Brace.
Goffman, E. (1974). Frame analysis: An essay on the organization of experience. Cambridge: Harvard University Press.
Hale, J., Dulek, R., & Hale, D. (2005, April). Crisis response communication challenges. Journal of Business Communication, 42(2), 112-134. doi:10.1177/0021943605274751
Knobloch, L., & McAninch, K. (2014). Uncertainty management. In P. Schultz, & P. Cobley, Handbooks of Communication Science: Vol. 6. Interpersonal Communication (pp. 297-319). Berlin: De Gruyter Mouton.
Matthes, J. (2008). Need for orientation as a predictor of agenda-setting effects: Causal evidence from a two-wave panel study. International Journal of Public Opinion Reseach, 20, 440-453.
Mutz, D., & Martin, P. (2001). Facilitating communication across lines of political differences. American Political Science Review, 95, 97-114.
Perry, R. (2007). What is a crisis? Management Communication Quarterly, 206-232.
. Communication Research, 24, 481-506. doi:10.1177/009365097024005002
Scheufele, D. (2000). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication and Society, 3, 297-316.
Seeger, M., Sellnow, T., & Ulmer, R. (2003). Communication and organizational crisis. Westport, CT: Praeger.
Tversky, A., & Kahneman, D. (1973). Availability-Heuristic for judging frequency and probability. Cognitive Psychology, 5, 207-232.
Uecker, J. (2011). Religious and spiritual responses to 9/11. Austin: National Institutes of Health.
Ulmer, R., Sellnow, T., & Seeger, M. (2015). Effective crisis communication: Moving from crisis to opportunity. Washington, DC: Sage.
Vu, H. T., Guo, L., & McCombs, M. (2014). Exploring "the world outside and the pictures in our heads": A network agenda-setting study. Journalism & Mass Communication Quarterly, 91(4), 669-686. doi:10.1177/1077699014550090
Weaver, D. (1977). Political issues and voter need for orientation. In D. Shaw, & M. McCombs, The emergence of American political issues: The agenda-setting function of the press. (pp. 107-119). St Paul. MN: West.
Weaver, D., Wojdynski, B. M., & Shaw, D. (2010). Vertical and or versus? Horizontal communities: Need for orientation, media use and agenda melding. World Association for Public Opinion Research Annual Conference, (pp. 1-20). Chicago. doi:10.13140/RG.2.1.3481.4881
Yin, R. K. (2009). Case study research: Design and research methods (4th ed., Vol. 5). Washington DC: Sage.

No comments:

Post a Comment