Recommend Solutions to Aid User Defenses and Reduce Evolving Information Risks

College Students’ Cybersecurity Risk Perceptions, Awareness, and Practices

Dilshani Sarathchandra, Kristin Haltinner, Nicole Lichtenberg Department of Sociology and Anthropology

University of Idaho Moscow, ID, USA

[email protected]

Abstract— In recent decades, considerable scholarly efforts have been invested in exploring college students’ cybersecurity awareness and practices. However, there has been a dearth of research on what specific factors drive students’ perceptions and behavior. In our previous qualitative research [1] we identified four such latent factors: routinization and ritualization of risk, optimistic bias, self-efficacy bias, and the ‘Can-I-Live’ syndrome. In this study, we use an online survey of college students at a large public university in the U.S. Pacific Northwest (N=498) and quantify the prevalence of these latent factors, along with some additional measures of awareness and behavior. We report the findings of the quantitative analysis as well as their implications for cybersecurity scholars and practitioners.

Keywords—cybersecurity; risk perception; college students; survey; optimistic bias

I. INTRODUCTION Recent research on college students' perceptions and

behaviors regarding online security have focused mainly on (1) Internet overuse and addiction, (2) cyber bullying, (3) impacts of Social Network Sites (SNSs), and (4) cybersecurity awareness.

Factors that contribute to Internet addiction or overuse by young adults include: age-related psychological and developmental characteristics, availability of and access to the Internet, and the increased normalization of more frequent online presence of young adults [2]. When college students are cut off from SNSs, researchers find that students report negative emotions such as anxiety, emptiness, and depression [3; 4]. Internet and SNS addiction can be particularly risky for certain students—undergraduates who self-identify as ‘lonely’ tend to spend more time on the Internet as a coping mechanism [5].

Research surrounding cyberbullying has expanded as a result of the increasing prevalence of technology, at least in the context of the more developed countries. Studies find that males report cyberbullying behavior more frequently than females, and that aggression and succorance (i.e., trying to gain sympathy, affection, and emotional support from others) are associated with cyberbullying, while intraception (i.e., engaging in attempts to understand one’s own behavior or the behavior of others) is not [6]. Studies have also established the

negative effects of cyberbullying including depression, anxiety, paranoia and suicidal thoughts and actions [7; 8].

Studies of SNSs focus on the relationship between SNSs and cyberbullying, anxiety, and depression; the positive relationship between SNSs and social capital [9]; and SNS use and quality of life and civic engagement for college students [10]. Another line of SNS studies have explored issues of privacy. These studies suggest that students whose friends have private social media profiles are more likely to have private profiles themselves [11] and that women are more likely to have private profiles than men [11; 12; 13].

Previous research on college students’ cybersecurity awareness indicates that college students are generally unaware of Internet risks and are not given enough information and education regarding potential risks [13]. Additionally, students tend to think that their peers are more likely to be affected by dangers than they themselves are [14]. In general, even when people are aware of Internet dangers, they are unlikely to take steps necessary to protect themselves [15; 16].

In our previous interview-based research [1] we find that college students' perceptions of cybersecurity are shaped mainly by four phenomena: (1) routinization and ritualization of risk (a belief that the behavior is acceptable because others also participate in the same behavior), (2) optimistic bias (the perception and belief that positive events are more likely to happen and negative events are less likely to happen to one's own self), (3) self-efficacy bias (a trust that students have the level of knowledge and intelligence necessary to protect themselves online) and (4) the ‘Can-I-Live’ syndrome (the idea that the steps needed to protect online security and privacy is too overwhelming to think about and could reduce the overall quality of life). This paper builds on this prior research, and contributes quantitative evidence for these four phenomena that shape perceptions and behavior of cybersecurity among college students.

II. METHODOLOGY

A. Data Collection Data for this study comes from a survey conducted among

college students at a large public university and its branch campuses located in the Pacific Northwest of the United States.

2016 Cybersecurity Symposium

978-1-5090-5771-9/16 $31.00 © 2016 IEEE

DOI 10.1109/CYBERSEC.2016.9

69

2016 Cybersecurity Symposium

978-1-5090-5771-9/16 $31.00 © 2016 IEEE

DOI 10.1109/CYBERSEC.2016.9

68

With the assistance of the university’s reporting analysts, we drew a simple random sample of 2000 email addresses from students enrolled in undergraduate courses in Spring 2016. An online survey covering student perceptions, practices, and awareness of cybersecurity issues was developed using findings of reference [1]. Using Qualtrics survey software, the survey was distributed to all 2000 email addresses in the sampling frame. Data collection was conducted between February and April 2016, with two follow-up email reminders. The final sample consisted of 498 valid responses, for an overall response rate of 24.9%, a rate that falls within the typical response rates reported for online surveys among college student populations [17].

B. Participants The final sample consisted of roughly equal proportions of

first to fourth year students (23% each) and a smaller proportion of 5th year undergraduates (8%). A majority of the students identified as white (82%). More respondents identified as female (59%) than male (41%). In terms of political orientation, approximately 25% of the students identified as Republican, Democrat, or Independent, respectively, while the remainder identified with other political party affiliations. As far as religious affiliations are concerned, 23% of students identified as Christian-Protestant, followed by 20% of Christian-Catholic, while 51% of the students in the sample identified as agnostic, atheist, or other.

C. Variables and Analytic Techniques The number of hours students spend online per week was

measured using an ordinal scale, where students reported hours from <10 hours/week to >80 hours/week. We used several dummy variables to measure student awareness (“able to explain the term”=1 and “not able to explain the term”=0) and level of concern about a series of cybersecurity issues (“number one concern”=1 and “not number one concern”=0). A set of statements were used to measure students self-reported behavior (Table 1) with responses ranging from “yes”=1, “no”=2, and “don’t know”=3. Additionally, to test the effects of routinization/ritualization, optimistic bias, self-efficacy, and ‘Can-I-Live’ syndrome, a series of statements were developed using the findings of reference [1]. These were measured using 7-point scales (“strongly disagree”=1 to “strongly agree”=7). Variables were subsequently recoded as “disagree”=1, “neutral”=2, and “agree”=3. Our main predictors included gender (“male”=1 and “female”=0), academic year (“Y1/Y2”=0 and “>Y3”=1), race, academic major, political affiliation, and religious affiliation. Data analysis presented below employs Chi-square tests, correlations, and independent sample T-tests.

D. Limitations Limitations of this methodology include those typical to

online surveys such as recall bias (errors caused by differences in accuracy and/or completeness of the information recalled by the participants), social desirability bias (tendency of interview and survey participants to answer questions in a manner that will be perceived favorably by the researchers), construct validity and reliability (issues related to measurements and

operationalization of variables), and non-response bias. The approximately 25% response rate, while normal for online surveys, may indicate some nonresponse bias if systematic differences existed between those who responded to our survey and those who did not. The sampling technique employed in this study (i.e., simple random sampling) does allow us to generalize to the population of students in the university at which this research was conducted. However, since findings are based on students of one university, they cannot be extrapolated to all college students in the United States.

III. RESULTS

A. Cybersecurity Awareness Approximately 56% of the students who responded to the

survey indicated that they spend less than 30 hours per week online, with the highest percentage of students (26%) reporting that they spend 21-30 hours/week online (Figure 1). In terms of awareness of cybersecurity issues, over 80% of students in each case reported that they can define the terms viruses, cyber bullying, cyber stalking, hacking, identity theft, and overuse/addiction. However, significantly lower numbers of students reported being able to define the terms malware (55%), Trojan horses (52%), phishing (50%), and worms (17%) (Figure 2).

As far as the level of concern about various cybersecurity issues is concerned, 43% of students listed identity theft as their biggest fear, followed by 25% who listed viruses. We found significantly less concern about issues such as phishing (1%), worms (2%), cyber stalking (2%), and cyber bullying (0.6%) (Figure 3). These findings indicate an association between awareness and concern such that awareness seems to increase risk perceptions with regards to some cybersecurity issues. For instance, students report higher awareness of identity theft and viruses while also listing them as their biggest concerns. Findings align with the availability and representativeness heuristics that people employ when making risk decisions (i.e., imaginability, memorability, and similarity serve as cues for probability judgments) [18].

Figure 1: Number of hours spent online per week

7069

Figure 2: Awareness of online vulnerabilities

These findings also shed light on how well student perceptions and concerns align with expert estimates of online vulnerabilities. For instance, expert estimates suggest that one of the biggest security vulnerabilities, at least since 2011, exists in the area of phishing: since 2013, more than 2/3 of incidents that comprise cyber-espionage have featured phishing while reports suggest that nearly 23% of recipients open phishing messages and 11% click on attachments [19]. However, according to the survey results, only 50% of the students reported being able to define phishing, and only a small fraction (1%) listed phishing as their biggest concern, indicating that their vulnerabilities may have increased disproportionately over time.

B. Prior Exposure and Practices A majority of students reported having been directly and

personally affected by a cybersecurity issues (57%) or knowing someone who has been affected (70%). While most students have antivirus software installed on their computers (77%), interestingly, 11% of the students reported not knowing if such software has been installed in their devices. In terms of online behavior, more than half of the students in the sample reported checking their bank account balances or making online purchases while logged into public Wi-Fi services. Additionally, 37% of the students also reported sharing their passwords with another person (Table 1). Chi-square analysis revealed that knowing someone who has been affected by online threats does reduce the likelihood of students checking their bank account balances and making online purchases while logged into public Wi-Fi (p<.05).

Further quantitative analysis supports a number of the findings of reference [1] as seen in results reported in Table 2. A majority of the students reported feeling safe while engaging in online activities (85%). Students also demonstrate optimistic bias: nearly 54% agreed that they are less likely to experience negative online events than the average college student. Self- efficacy is a main strategy used in risk management such that 65% of students believe that they do enough to protect their online security and privacy; 84% report trusting reputable websites; and 59% report being able to fully control the personal information they share online. Additionally, we found partial support for the ‘Can-I-Live’ syndrome: 32% of students

agreed that the steps needed to protect their online security and privacy is too overwhelming to think about.

C. Effects of Sociodemographic Variables As far as our main predictors are concerned, we found that

first and second year students are more likely to report feeling safe when sharing personal information on social networking sites, compared to their more senior counterparts (Y1/Y2 mean= 3.15, > Y3 mean= 2.74; p<.01). Additionally, first and second year students are more likely to report having full control over the information they share online (Y1/Y2 mean= 4.92, > Y3 mean= 4.65; p<.01).

Independent sample T-tests show more optimistic bias among male students than female students (mean for males= 5.01, mean for females= 4.58; p<.01), and more support for the ‘Can-I-Live’ syndrome among female students than male students (variable reverse coded, small values indicating more agreement; mean for females= 4.31, mean for males= 4.71; p<.01). Overall, these findings align with risk perception literature where men tend to have lower risk perceptions than women on a number of different dimensions of risk [20]. We did not observe any statistically significant effects of the other sociodemographic variables used in the study (i.e., race, academic major, political affiliation, and religious affiliation).

Correlation analysis provided further evidence for associations between student perceptions and their self- reported behavior. As such, students who are optimistically biased are also more likely to report that they (1) do enough to protect their privacy online (r=0.33); (2) trust reputable websites (r=0.14); and (3) have full control over what they share (r=0.16). As expected, students who feel overwhelmed by cybersecurity issues (Can-I-Live syndrome) are less likely to report that they do enough to protect their privacy and security (r=0.40) while students who generally feel safe when engaging in online activities report that they do enough to protect their privacy and security (r=0.40).

Figure 3: Level of concern about online vulnerabilities

7170

Statement

% Yes % No % Don’t know

Personally experienced a cyber threat

57.17 42.83 -

Knows someone who experienced a cyber threat

69.78 30.22 -

Has antivirus software installed in the computer

77.14 11.43 11.43

Has checked your bank account balance while logged into a public Wi-Fi network

51.94 43.35 4.70

Has made an online purchase while logged into a public

Wi-Fi network

55.37 39.26 5.37

Has shared a password to an online account with another

person

37.14 61.02 1.84

Table 1: Prior Exposure and Behavior

Risk Management Strategy

Statement % Disagree

% Neutral % Agree

Routinization

I feel pretty safe when engaging in online activities.

9.24 6.09 84.67

I feel safe sharing personal information on social networking sites.

67.51 8.65 23.84

Optimistic Bias

I’m less likely to experience negative online

events than the average college student. 14.05 32.29 53.66

Self-efficacy

On a day-to-day basis, I do enough to protect my

online security and privacy. 12.13 23.01 64.86

I trust ‘big name’/reputable websites such as Amazon and eBay.

10.29 6.09 83.62

I have full control over what types of personal information I share online.

26.95 14.11 58.94

Can-I-Live Syndrome

The steps needed to protect my online security and privacy are too overwhelming to think about.

51.37 16.21 32.42

IV. DISCUSSION AND CONCLUSION

A. Fearing the Wrong Things Our research demonstrates that college students tend to fear

the “wrong” things online. That is, they have greater fears regarding dangers that are less likely to occur and ignore more significant risks. In part, this is due to a tendency towards self- preservation noted in our previous work as the ‘Can-I-Live’ syndrome [1].

This inclination to hold fears around unlikely events rather than those that pose a greater risk to an individual is not unique to college students and online spaces. Glassner explores this phenomenon in his book Fear: Why Americans Fear the Wrong Things [21]. His analysis demonstrates that fears largely arise from media exposure and that journalists, politicians, and pundits often focus on unlikely dangers at the expense of serious risks. For example, Glassner examines the fear Americans have of international terrorism. He demonstrates that, following the events in New York City on September 11, 2001, media attention caused this danger to be overblown in the American psyche. In his book Inside Terrorism, Bruce Hoffman further explores the impact of media attention on fears of terrorism and finds that 14% of Americans think it is probable that they will be involved in a plane hijacking. The actual chance of terrorism affecting an individual is less than 1/100,000. Even with this slim chance, 33.3% of people surveyed refused to travel abroad in the years following September 11, 2001, because of airplane terror threat [22]. Americans are far more likely to be the victims of domestic, right-wing, terrorism [23]. Like Glassner, Hoffman attributes such irrational fears to “the disproportionate coverage accorded to [international] terrorism by the American media” [22]. In the arena of cyber security we see numerous stories and TV shows dedicated to discussing the dangers of identity theft (the greatest fear among college students) and little attention paid to other risks such as phishing.

Table 2: Strategies Employed to Manage Online Vulnerabilities

Glassner suggests that this misrepresentation of danger is partially due to the difficulty in facing perceivably insurmountable large social problems: a society-wide ‘Can-I- Live’ syndrome [21]. Using this theory one could argue that online dangers, such as phishing, are so pervasive and seemingly unsolvable that people focus, instead, on more manageable dangers such as viruses for which they see a clear solution: anti-virus software.

However, this story is not complete as one could argue that identity theft is as difficult to solve as phishing attacks. Gardner uses neuroscience research to explore human reactions to fear [24]. He argues that fear is produced through two brain processes, which he calls the “head” and the “gut”. The head is the rational part of the brain, which measures and evaluates

7271

risk. The gut is a more rudimentary process that makes quick decisions based on environmental information. Gardner argues that the gut is perfect for dealing with the fears imbedded in hunter-gatherer societies, but not so in modern environments where fear-provoking images abound. Thus, Gardner concludes, political pundits, journalists, and businesses activate the “gut” through powerful images and anecdotes, provoking irrational fears in society; fears that the rational brain cannot combat. Thus, it is the interaction of compelling images and anecdotes, from sources viewed as reputable that lead to widespread social fears.

Gardner’s findings suggest that it is not simply media attention that leads to erroneous fear by the public, but also the type of news stories employed. Powerful anecdotes, featuring innocent victims, are particularly influential [21; 24]. In the case of terrorism, the powerful stories of victims contributed to societal fears. Regarding online threats, tales of identity theft dominate news stories and feature innocent victims to whom most people can relate. Phishing attacks, however, tend to affect businesses and are not as easily lent to heart-tugging narratives. As our study demonstrates, people readily fear identity theft but rarely show concern over phishing attacks. Thus, even though less than 1% of people are the victims of identity theft, and around 80% of victims suffer no financial loss [25] and, in contrast, phishing attacks cost businesses an average of $3.7 million a year [26], people continue to fear the less likely and less expensive crime.

B. Costs of Misplaced Fear Obviously misplaced fear leads to a number of problems,

including misdirected resources, poor policy-making, and poor decision-making.

Misdirected resource investment to combat unlikely events is significant. Businesses, politicians, and media conglomerates make money by promoting certain fears in society. For example, in 2006 Jessica’s Law was passed in California after a nine-year old girl was murdered by a paroled sex-offender and the story hit the media waves. Subsequently, the law required that all sex offenders go through psychological evaluations, to the tune of $24 million in 2007. There was no change to the prosecution, punishment, or treatment of sex offenders. The primary change was an increase in revenue for private psychologists. This occurred at a time when California was suffering from a staggering deficit of $42 billion [21].

Fear also has the potential for the passage of poor policies. For instance, due to a number of stories in the media about security and medical records, a minor moral panic arose, resulting in more restrictive policies for sharing such information between medical practices. This has prevented providers’ access to certain information that could have halted 400,000 deaths in the United States last year [27].

Similarly, erroneous fear can lead to people placing themselves in unnecessary danger. Following the terrorist attacks of September 11th, 2001 many Americans chose to drive on their vacations, as opposed to flying. Because driving is more dangerous than air travel, it is estimated than 1,595 people were killed as a result of this fear-inspired behavioral shift [28].

Less is known about the costs associated with misplaced fears in online spaces. However Forbes reports that over $2 trillion has been lost due to cyber crimes [29]. One can assume that better attention to curbing the greatest dangers could save a significant amount of money. This is an area ripe for further research.

C. Change Requires Creative Ways of Disseminating Information Given the current misplaced fear and tendency for people to

ignore present dangers and/or misunderstand their likelihood, it is clear that efforts to combatting online dangers must change. Research on risk indicates the solution may be found in how news about cyber security concerns are presented.

The primary need is to shift public information dissemination to involve compelling stories. In order to work, these stories need to be evocative and feature empathy- provoking characters. Emotions, more than fact, drive the fear reactions of the public [24]. Statistics on threat, risk, and danger appears to have little impact on the degree to which the public develops fear [24].

A second shift needs to be in the frequency of the stories told. The volume of reporting on a particular issue also affects public attention. A wide range of compelling narratives presented by a variety of people (politicians, reporters, etc), frequently, over a period of time will have the greatest effects.

We now know that relatively minor risks (as assessed by technical experts) often elicit strong public concerns, with significant costs to the economy and the society [30]—a phenomenon with notable implications to cybersecurity, as discussed above. Understanding the “human-side” of cybersecurity is essential to tackle the implications of public concern, as the cybersecurity community moves ever closer to engaging fully with artificial intelligence, virtual reality, and their untapped potential for growth.

REFERENCES

[1] Haltinner, Kristin, Dilshani Sarathchandra, and Nicole Lichtenberg. 2016. “Can I Live? College Student Perceptions of Risks, Security, and Privacy in Online Spaces.” Pp. 69-81 in Cyber Security, Second International Symposium, CSS 2015, Revised Selected Papers, edited by K. Haltinner, D. Sarathchandra, J. Alves-Fosse, K. Chang, D. Conte de Leon, and J. Song. New York: Springer. DOI: 10.1007/978-3-319- 28313-5_6.

[2] Kendall, Jonathan J. 1998. “Internet Addiction on Campus: The Vulnerability of College Students.” CyberPsychology and Behavior 1(1): 11-17.

[3] O’Keeffe, Gwenn Schurgin, and Kathleen Clarke-Pearson. 2011. “The Impact of Social Media on Children, Adolescents, and Families.” Pediatrics 127(4): 800-804.

[4] Moreno, Megan A., Mara Stewart, Megan Pumper, Elizabeth Cox, Henry Young, Chong Zhang, and Jens Eickhoff. 2014. “Facebook Use During a Stressful Event: A Pilot Evaluation Investigating Facebook Use Patterns and Biologic Stress Response.” Bulletin of Science, Technology & Society 34(3-4): 94-98.

[5] Morahan-Martin, Janet and Phyllis Schumacher. 2003. “Loneliness and Social Use of the Internet.” Computers in Human Behavior 19: 659-671.

7372

[6] Dilmac, Bulent. 2009. “Psychological Needs as a Predictor of Cyber Bullying: A Preliminary Report on College Students.” Educational Sciences: Theory and Practice 9(3): 1307-1325.

[7] MacDonald, Christine D. and Bridget Roberts-Pittman. 2010. “Cyberbullying Among College Students: Prevalence and Demographic Differences.” Procedia Social and Behavioral Sciences 9(2010): 2003- 2009.

[8] Schenk, Allison M. and William J. Fremouw. 2012. “Prevalence, Psychological Impact, and Coping of Cyberbully Victims Among College Students.” Journal of School Violence 11(1): 21-37.

[9] Ellison, Nicole B., Charles Steinfield, and Cliff Lampe. 2007. “The Benefits of Facebook ‘‘Friends:’’ Social Capital and College Students’ Use of Online Social Network Sites.” Journal of Computer-Mediated Communication 12(2007): 1143-1168.

[10] Valenzuela, Sebastian, Namsu Park, and Kerk F. Kee. 2009. “Is There Social Capital in a Social Network Site?: Facebook Use and College Students’ Life Satisfaction, Trust, and Participation.” Journal of Computer-Mediated Communication 14(2009): 875-901.

[11] Debatin, Bernhard, Jennette P. Lovejoy, Ann-Kathrin Horn, and Brittany N. Hughes. 2009. “Facebook and Online Privacy: Attitudes, Behaviors, and Unintended Consequences.” Journal of Computer-Mediated Communication 15(2009): 83-108.

[12] Lewis, Kevin, Jason Kaufman, and Nicholas Christakis. 2008. “The Taste for Privacy: An Analysis of College Student Privacy Settings in an Online Social Network.” Journal of Computer-Mediated Communication 14(2008): 79-100.

[13] Lawyer, James P. and John C. Molluzzo. 2011. “A Survey of First-Year College Student Perceptions of Privacy in Social Networking.” Consortium for Computing Sciences in Colleges 36-41.

[14] Campbell, Jamonn, Nathan Greenauer, Kristin Macaluso, and Christian End. 2007. “Unrealistic Optimism in Internet Events.” Computers in Human Behavior 23(2007): 1273-1284.

[15] Livingstone, S. 2008. “Taking Risky Opportunities in Youthful Content Creation: Teenagers’ Use of Social Networking Sites for Intimacy, Privacy and Self-expression.” New Media & Society 10(3): 393–411.

[16] Tufekci, Z. 2008. “Can You See Me Now? Audience and Disclosure Regulation in Online Social Network Sites.” Bulletin of Science, Technology & Society 28(1): 20–36.

[17] Sheehan, Kim Bartel. 2001. “E-mail Survey Response Rates: A Review.” Journal of Computer-Mediated Communication 6(2).

[18] Slovic, Paul, Melissa Finucane, Ellen Peters, and Donald G. MacGregor. 2002. “Rational Actors or Rational Fools: Implications of the Affect Heuristic for Behavioral Economics.” The Journal of Socio-Economics 31:329-342.

[19] Data Breach Investigations Report. 2016. Verizon Enterprise Solutions. Retrieved on May 09, 2016. http://www.verizonenterprise.com/verizon- insights-lab/dbir/

[20] Finucane, Melissa L., Paul Slovic, C.K. Mertz, James Flynn, and Theresa A. Satterfield. 2000. “Gender, Race, and Perceived Risk: The ‘White Male’ Effect.” Health Risk & Society 2(2):159-172.

[21] Glassner, Barry. 2010. The Culture of Fear. Basic Books: New York. [22] Hoffman, Bruce. 2006. Inside Terrorism. Columbia University Press:

New York. [23] Piatt, Christian. 2015. “Biggest Threat on U.S. Soil: Domestic Right-

Wing Extremists.” The Huffington Post. January 24. Retrieved on May 29, 2016. http://www.huffingtonpost.com/christian-piatt/biggest-threat- on-us-soil_b_7656346.html

[24] Gardner, Daniel. 2009. The Science of Fear. Penguin: New York. [25] Consumer Reports News. 2010. “New Study Shows ID Theft

Exaggerated.” Consumer Reports. Retrieved on May 27, 2016. http://www.consumerreports.org/cro/news/2010/12/new-study-shows-id- theft-exaggerated/index.htm

[26] Greenberg, Adam. 2015. “Report: Phishing Costs Average Organization $3.7 Million Per Year.” SC Magazine. August 27. Retrieved on May 27, 2016.

[27] Hwang, Joy and Levin, Peter. 2016. “Fear of Lost Privacy Costs Lives.” The Business of Federal Technology. January 28th. Retrieved on May 28, 2016. https://fcw.com/articles/2016/01/28/comment-levin-hwang- ehr-privacy.aspx

[28] Ball, James. 2011. “September 11's indirect toll: road deaths linked to fearful flyers.” The Guardian. September 5. Retrieved on May 29, 2016. http://www.theguardian.com/world/2011/sep/05/september-11-road- deaths

[29] Morgan, Steve. 2016. “Cyber Crime Costs Projected To Reach $2 Trillion by 2019”. Forbes. January 17. Retrieved on May 28, 2016. http://www.forbes.com/sites/stevemorgan/2016/01/17/cyber-crime- costs-projected-to-reach-2-trillion-by-2019/2/#32fa2aba1567

[30] Kasperson, Roger E., Ortwin Renn, Paul Slovic, Halin S. Brown, Jacque Emel, Robert Goble, Jeanne X. Kasperson, and Samuel Ratick. 1988. “The Social Amplification of Risk: A Conceptual Framework.” Risk Analysis 8(2):177-187.

7473