Master Thesis Crisis and Security Management
The return of foreign fighters and
the level of fear of terrorism
A survey among Dutch society
Sophie de Jong (s2111705) University of Leiden
04-08-2019 1st reader: Tahir Abbas 2nd reader: Eamon Aloyo
Table of content
1. INTRODUCTION 1.1 Research purpose
1.1.1 Research questions
1.2 Added value
1.3 Structure of the research
2. THEORETICAL FRAMEWORK
2.1 Dutch foreign fighters in the Islamic State
2.1.1 Terrorism in general 2.1.2 Foreign fighters 2.1.3 Islamic State
2.1.4 Dutch foreign fighters 2.1.5 Return to the Netherlands
2.2 Fearful society
2.2.1 Levels of security
2.2.2 Components of attitude against fear
2.3 Descriptive statistics 3. METHODOLOGY 3.1 Kind of research 3.2 The survey
3.3 Respondents and ethical aspects
3.3.1 Respondent recruitment 3.3.2 Ethical aspects 3.4 Variables 3.4.1 Demographic variables 3.4.2 Independent variables 3.4.3 Dependent variables 3.5 Data analysis 3.5.1 Descriptive statistics 3.5.2 One-way ANOVA 3.5.3 Chi-square test 3.5.4 Correlations
3.6.1 Reliability 3.6.2 Validity 3.6.2.1 Internal validity 3.6.2.2 External validity 3.6.3 Limitations 4. RESULTS 4.1 Population 4.2 Variables
4.2.1 Attitude towards returnees 4.2.2 Fear of terrorism 4.3 Exploratory analysis 4.4 Hypotheses testing 4.4.1 Main hypothesis 4.4.2 Other hypotheses 4.5 Demographics 4.5.1 Gender 4.5.2 Age 4.5.3 Urban or rural 4.5.4 Children 5. CONCLUSION 6. DISCUSSION BIBLIOGRAPHY APPENDICES
1 Introduction
Almost everybody remembers the attacks in Paris in November 2015. Eight men attacked the Bataclan and a terrace of a restaurant in the center of Paris. A total of 130 people died and over 360 people got injured. RTL Nieuws (2015) elaborates on the backgrounds of the known attackers. They say that at least four of the eight known attackers had travelled to Syria to join Islamic State. When they returned, they started to plan the devastating attacks. Foreign
fighters who went to Syria or Iraq to fight beside their brothers and sisters of the Islamic State and who are now returning to their homeland are a dangerous threat for the Dutch society as well. This is said by the Dutch General Intelligence and Security Service (AIVD, 2018) and the National Coordinator of Counterterrorism and Security (NCTV, 2016). According to the AIVD (2019) approximately 300 adults travelled to Syria and Iraq until 20th of March 2019 to
join Islamic State or an alliance of Al-Qaida, Ha’yat Tahrir al-Sham. Over 50 people already returned and a number of approximately 180 adults are yet to return (AIVD, 2019). The number of foreign fighters who went to Syria and actually will engage in terrorism upon their return is not determined, however, the threat remains serious (De Roy Van Zuijdewijn, 2014). How does the Dutch society feel about the return of foreign fighters? Is this opinion
influenced by the fear of an attack? Could this opinion differ between men and women, or age, or level of education? These questions form the principal focus of this research.
The introduction continues to discuss the exact focus of the study and will present an overview of the goal of this research, followed by the research questions. Next, the added value for science and society will be explained. Finally, it presents an overview of the structure of this research.
1.1 Research purpose
The purpose of this research was in first instance to shine more light on the motivations and push or trigger factors for potential jihadists leaving their homes in order to discover more information on the process and to be more able to prevent these kinds of situations in the future. To accomplish this, the focus was on people who travelled to Syria or Iraq to join the Islamic State in the period of 1st of July 2012, this is when the first people left the Netherlands
to go to Syria or Iraq (AIVD, 2017), until the present. As these people are not easy to encounter, due to detention or unknown whereabouts, this research focused solely on close relatives to these foreign fighters. Interviews would have been conducted among these relatives to comprehend the biography of the foreign fighters and to see if there are any similarities in sociodemographic factors between the foreign fighters.
However, the family members of foreign fighters were too hard to reach. Therefore, the strategic and evaluative decision was made to redefine the research purpose into a survey research to find out what the Dutch society feels in relation to the return of the foreign fighters. With the switch to a survey research, the chances of collecting high quality data are optimized.
The opinion of the Dutch people about the returnees is tested in relation to their level of fear of terrorism. This is done to find out if there is a connection when someone’s opinion is in disadvantage of returning and his or her level of fear of terrorism. The focus remains on the foreign fighters who travelled to Syria or Iraq to join the Islamic State in the period of 1st
of July 2012 until now (AIVD, 2017). As stated above, about 300 adult people in total
travelled from the Netherlands to the Islamic State. Except for the people who died in combat, these people already did come back to the Netherlands, are currently staying in refugee camps or are in unknown places likely to come back in the future.
1.1.1 Research questions
The main question of this research is:
“Are Dutch adults who are more afraid of terrorism, more against the return of foreign fighters than people who are less afraid of terrorism?”
Furthermore, there are no questions but hypotheses that will be constructed in chapter 2.
1.2 Added value
The results of this research will contribute to the understanding of the feelings and opinions of Dutch citizens and could cause a better insight on how to handle these kinds of issues in the future. Therefore, this research is societal relevant. It is important for the Dutch government, for example, to understand how Dutch citizens feel about their decisions in policies around foreign fighters and terrorism. Moreover, it will be interesting to know how fearful the Dutch citizens are in this day and age, when it comes to terrorism.
Although there are multiple researches conducted about this subject, these researches did not yet include the return of the foreign fighters in their surveys/interviews. Probably because this is relatively new as a public concern and the wave of foreign fighters in the Islamic State is currently coming back. Therefore, this is a relative new issue. The researcher will also look at the differences between age, gender, education, religion, living situation,
working situation, political preference and whether or not participants have kids. By looking at differences in age it is possible to look at the intergenerational gap. The intergenerational gap is the difference between generations in perspective of, in this research, foreign fighters, their return and the level of fear of terrorism. The intergenerational gap is not often researched in previous works. Together with the new concept of returnees, this research has a great scientific value.
1.3 Structure of the research
Several building blocks bring answers to the research questions. This introduction is followed by the first chapter. This is a theoretical framework to conceptualize the terms used in this study and the previous articles written about these terms. Moreover, previous relevant researches are discussed to find the theoretical fundament that is already existing about this subject. Several relationships have been tested before and some are to be solved with literature. These items are aligned in the theoretical framework and from there, the assumptions and expectations are developed.
The next chapter will be the methodology chapter to explain how the research question is to be answered. It will include a clear methodological justification, the survey development and the operationalization. The chapter includes a description of the research design, the hypotheses, the variables, the survey, the respondents, the data-analysis, the ethical aspects of the research, and validity, reliability and the limitations of this research.
Chapter 4 will include the results of the survey. In this chapter, the results of the survey are being tested within the program IBM SPSS Statistics 24. The results from those tests are elaborated in the Results-chapter, this will be done per variable to paint a clear picture of the results in order to be able to make conclusions about them later.
Finally, the conclusions present the answer to the research questions, as well as a substantiated discussion of the limitations of the research and possible avenues for future research. Hopefully, this chapter results in a concrete and convincing practical
recommendation.
2 Theoretical framework
In this chapter, the positioning of this research project in the body of knowledge is shown. It will be a clear identification of the gap in knowledge this research is addressing and the potential added value of this research. Moreover, this chapter includes the conceptualization of the terms used and display possible answers to the research questions, in the form of the
expectations and assumptions. To begin with the explanation of the foreign fighter phenomenon a few researches are explained.
Some researchers state that leaving for Syria or Iraq is due to some sort of frustration about their societal position in the Netherlands (Weggemans, Bakker & Grol, 2014; Dawson & Amarasingam, 2016). Because of this frustration they might feel lonely or in some sort of identity crisis or uncertainty, which could be a reason for these people to join the Caliphate (Curtis et al, 2016). Other researchers, e.g. Perliger and Milton (2016), say that
sociodemographic factors such as age, income, education and social class might cause the foreign fighter phenomenon. The first section elaborates on the conceptualization of foreign fighters who are returning. First, the term terrorism is defined, then ‘foreign fighters’ is being conceptualized, a little bit of information about the Islamic State is given, the foreign fighters from the Netherlands are enlightened and lastly, the return to the Netherlands is explained.
2.1 Dutch foreign fighters in the Islamic State 2.1.1 Terrorism in general
To explain the concept of foreign fighters, it would be convenient to look at the concept of terrorism. First of all, terrorism does not have a worldwide accepted definition. However, for a lot of governments it is top of mind these days. Therefore, a few definitions will be given to make sure that for this research it is clear what is seen as terrorism. The definition that
immediately comes up when you search the term ‘terrorism definition’ in Google (www.google.nl) is “the unlawful use of violence and intimidation, especially against
civilians, in the pursuit of political aims”. This definition seems right, but when you look into the literature you will see a lot of additions and contradictions. According to Cooper (2001), the definition of terrorism has undergone multiple refinements and suggestions over the years. He tries to consider how to define terrorism or at least to be able to know it. Cooper (2001) gives a basis of the definition of terrorism where he would like to build on further in his article. This definition sounds as following: “Terrorism is the intentional generation of massive fear by human beings for the purpose of securing or maintaining control over other human beings” (p. 883). He already stated that this definition already has gone through a lot of small refinements over the last decades, but it is a good starting point. He ends his article with a quote of Wittgenstein (as quoted in Cooper, 2001, p. 892): “There are indeed, things that cannot be put into words. They make themselves manifest. They are what is mystical”. Cooper says that terrorism one of these things is.
Although Cooper (2001) did not reach a definition of terrorism, other researchers did. For example, Shanahan (2016). In this research, the definition is as follows: “Terrorism is the strategically indiscriminate harming of threat of harming members of a target group in order to influence the psychological states of an audience group in ways the perpetrators anticipate may be beneficial to the advancement of their agenda” (p. 110). As you can see, this
definition has a lot more aspects than the basic definition Cooper (2001) has given. This will therefore be the definition that is leading in this research. However, terrorism in general is not solely focused on Islamist extremism, what this research is about because of the Islamic State. In this research, therefore, Islamist extremism is the main form of terrorism in this research and is always meant when the term terrorism is used.
2.1.2 Foreign fighters
The phenomenon of foreign fighters is not clearly defined yet. According to Fritz and Young (2017), after comparing multiple definitions, the right definition is that foreign fighters are individuals who leave the comfort of their homes to engage in fighting for a cause and/or people with whom they identify. By all means, according to Malet (as cited in: Fritz and Young, 2017) the definition should contain the following elements: the person has to be foreign from the conflict zone, has to come to the conflict zone to join an insurgency and is a volunteer without the expectation of payments in any kind. Borum and Fein (2016) are
researching the psychology of foreign fighters. Before they do that, they are trying to establish a definition. Borum and Fein (2016) state that Hegghammer’s (as quoted in Borum & Fein, 2016, p. 215) definition comes close: “An agent who has joined, and operates within the confines of, an insurgency, lacks citizenship of the conflict state of kinship links to its warring functions, lacks affiliation to an official military organization, and is unpaid”. Borum and Fein (2016) do not agree on the activity of the foreign fighters as just joining the insurgency. They do not find a commonly accepted typology of actions in support of an insurgency, but name four basic categories that should be considered: direct action, operational support, movement support and logistical support.
According to Schmid and Tinnes (2015), European citizens who become foreign fighters often have a second-generation immigration background. As Abbas states in his research about foreign fighters from Britain (2001, p. 247): “Given that the socio-economic and demographic nature of Muslims and that the type of Islam practiced is differentiated, it is important to state that a further disparity exists between generations of Muslims in the West. As the generations evolve, younger individuals increasingly disconnect from their parental
perceptions of Islam”. Therefore, it could be important to keep in mind, in relation to
identification with Islam. According to Borum and Fein (2017), the seeking for identity could be a reason to join the brotherhood in Syria or Iraq, to feel that they belong.
Whenever foreign fighters are mentioned in this research, jihadist foreign fighters are meant by that, because this research is solely focused on Dutch foreign fighters who are going to join the Islamic State.
2.1.3 The Islamic State
In this section, the concept ‘the Islamic State’ is elaborated to explain this group, to show what it does and also how this group is recruiting so many people from the Western world, and therefore, also from the Netherlands, to go all the way to Syria or Iraq to join them.
According to Laub and Masters (2016), the Islamic State is a movement with militant Sunni backgrounds that has gained ground in western Iraq, eastern Syria and Libya, where the movement has tried to start the Caliphate from. The Caliphate is, according to Laub and Masters (2016), meant to exert exclusive political and theological power over all Muslims all over the world. Their state-building process has not been labeled as institution building, but rather as extremely violent. Laub and Masters (2016) state that ever since the widely
publicized successes of the movement have attracted thousands of foreign recruits and also led to groups or terrorists attacking other countries in their name. The movement is led by the self-proclaimed leader or caliph Abu Bakr al-Baghdadi. It is uncertain where he is and even whether or not he is still alive (Rosman, 2019).
Gates and Podder (2015) researched, among other subjects, the way the Islamic State recruits so many foreign fighters. They state that the Islamic State heavily relies on the recruitment of foreign fighters. After analyzing the different ways that the Islamic State uses to recruit foreign fighters, they conclude that Islamic State needs all of these recruits to keep their gained territory. This is in contrast with most rebel military organizations, who normally use a hit-and-run technique (p. 113). Gates and Podder (2015) say that for this extensive recruitment, the Islamic State has built an enormous social media recruitment machine. They use linguistic and technological talents to reach people all around the world. The widely broadcasted successes of the Islamic State only enhanced the effectiveness of these recruitment techniques.
2.1.4 Dutch foreign fighters
According to Weggemans, Bakker and Grol (2014), the Dutch authorities observed, at the end of 2012 and the beginning of 2013, the rapidly growing number of Dutch citizens and
residents going to Syria to join jihadists groups. They say that the first movement of jihadist travelers to Syria were members of activist radical Islamist movements, such as
Shariah4Holland. They also explain how the growth of interest for the conflict in Syria, the line between radicalism and jihadism became blurred, as radical quotes and thoughts were transformed into jihadist views. In the time that Weggemans, Bakker and Grol (2014) wrote this article the terrorism threat level in the Netherlands was since 2013 ‘substantial’, given the on-going threat posed by above mentioned groups and increasing number of people returning from Syria. The National Coordinator of Counterterrorism and Security (NCTV, 2019) explains how this threat level remains ‘substantial’ in the Netherlands. Since there have been a few attacks and attempts to attack in the Netherlands for the last couple of months, they do not see the threat level lowering any time soon. The NCTV (2019) also states that the Islamic State is losing more and more ground in Syria and Iraq, which causes most foreign fighters to be forced to go to their home country.
The Dutch General Intelligence and Security Service (AIVD, 2019) state in their report that approximately 300 adult people from the Netherlands went to Syria or Iraq to join the Islamic State. One third of those people is female and two third is male. 55 people have returned to the Netherlands and 90 people were killed in combat. That still leaves about 165 people who are somewhere between the Islamic State and the borders of the Netherlands. Next to that, approximately 200 children with a link to the Netherlands are still in Syria or Iraq. The AIVD (2019) states that about 25 per cent was taken by their parents and 75 per cent was born there. However, it is extremely difficult to estimate the newborns in the conflict zone, partly by foreign fighters who married a local woman in the conflict zone. The AIVD (2019) thinks that over half of the children are younger than four years old and somewhat 10 per cent is over nine years old. In the Islamic State is an age of nine the age you start learning how to use weapons and how to plan attacks.
As stated above, a lot of Dutch people went to Syria or Iraq. Therefore, this is seen as a significant threat for the Dutch society by the AIVD and the NCTV. Consequently, it is important to know why these people left for Syria in order to be able to prevent it in the future.
2.1.5. Return to the Netherlands
As stated above, according to the NCTV (2019), the Islamic State is losing more ground in Syria and Iraq. Therefore, the foreign fighters are forced to leave and most likely will return to their home countries. According to the AIVD (2019), a total of 300 adult people left from the Netherlands to the Islamic State. 55 already returned, 90 killed in combat and that leaves 165 people to be in unknown places and these people are likely to return to the Netherlands. This information, together with the information about the many children who went or were born in the Islamic state, caused the political agenda to focus on the return of these foreign fighters and these children.
The Dutch newspaper the Volkskrant (2019) wrote an article about the different opinions by the political parties in the Netherlands. Some parties say that the foreign fighters should stay there and should be tried in Iraq and Syria. In Iraq and Syria, it is still possible to be sentenced to death. Others say it is not morally and ethically possible to leave the Dutch foreign fighters there and we have to bring them back. Another reason that is given to bring them back to the Netherlands, is that they cannot run from there punishment. Therefore, we need to repatriate and bring them back, with all the safety issues and financial burdens. In the article of the Volkskrant, Bart Schuurman is quoted. He is a terrorism expert of University of Leiden. He says: It seems now as if we are returning them to give them a second chance. We need to reform the debate into how we can make sure that these people are tried correctly and will form no danger to society in a later stadium (as quoted in Volkskrant, 2019). The article ends with the statement that there is enough professional infrastructure to shelter foreign fighters in the Netherlands, if this will be used is a political issue.
A report of the researchers Bakker, Scarione and de Roy van Zuijdewijn (2019) about the return of foreign fighters explains the fact that, in December 2018, the Ministry of Justice and Safety sent a letter to the House of Representatives to convince them not to help foreign fighters return. The Ministry based this letter on an analysis of the NCTV about the safety issues around the return of foreign fighters. On the one hand, leaving them there and offer no help with their return and on the other hand, offer the help and help the foreign fighters return safely to the Netherlands. Both scenarios come with safety issues according to the NCTV (Bakker, Scarione and de Roy van Zuijdewijn, 2019).
2.2 Fearful society
The term security has been defined many times (Zedner, 2009). Zedner (2009) says that the term security contains two aspects: objective and subjective security. Objective security is,
according to Zedner (2009), the attempt to create a situation where there is absolutely no possibility of danger. This situation could only be reached when threats are being neutralized or avoided. There will always occur new threats, so it remains an attempt and will make the end goal impossible to reach. Subjective security is the attempt to reach an absolute condition where everybody feels safe and secure. In this condition, the perspective is on how the people experience security (Zedner, 2009). The experience derives from the mind, the perceptions and the emotions people have with respect to security. With subjective security, everything could be interpreted as a threat and for every person, threats are something different, therefore subjective security remains an attempt as well.
For this research, it is important to look at the subjective security when testing how fearful the Dutch citizens are for terrorism and the return of foreign fighters. In this way, a clearer view of the security experience and the opinion about returnees can be developed.
2.2.1 Levels of security
The experience of security is influenced by multiple factors. These factors are being divided into three levels (Oppelaar & Wittebrood, 2006). These levels are: factors on individual level, situational level and cultural level. These factors of the three levels influence the experience of security and could lead to feelings of fear. The individual level is the level where the personal traits are deciding in the level of fear of a person. The most important trait is the dispositional fear, this is the tendency to react fearful in certain situations (as cited in
Oppelaar & Wittebrood, 2006). Feelings of fear could be embedded in the personal traits and are coherent to the subjective estimation of the own vulnerability and helplessness. When people are more convinced that they are helpless, the stronger the feelings of fear will be (Oppelaar & Wittebrood, 2006). Moreover, previously being victimized could contribute to the estimation of vulnerability.
Moreover, individual traits are being tested in this research. This is because a lot of research has been done on, for example, the difference between women and men and age differences in the level of fear of terrorism. Lane (2013) states that gender is one of the better indicators of the difference in fear for crime. Women are often more afraid of being a victim than men. Lane (2013) says that women are therefore more likely to avoid certain places in order to be safer. Age is an important trait in the scientific research on security. Most
researches state that elder people are more afraid of crime than younger people. Although, the age is very difficult to define. When is someone older and when is someone younger?
varies in kind of crime. For instance, people under the age of 25 are more afraid of vandalism, murder and rape while older people are more worried about street robberies or getting hurt by a burglar.
The situational level is the situation where traits of the environment could contribute to the feelings of fear. These traits could be the design of a public place, the social
composition, the level of nuisance and deterioration and the social cohesion (Oppelaar & Wittebrood, 2006). The cultural level is in comparison to the contextual level more broadly. The cultural factors are not tested in this research but will be named shortly. These factors are developments in the media, individualization and changing social proportions and the
globalization and internationalization. In this research, the perspective of Oppelaar & Wittebrood (2006) about the experience of security is being used to interpret the results and draw a conclusion about the results of the survey.
2.2.2 Components of attitude against fear
Rosenberg and Hovland (1960) performed a research on the attitude of people against fear. They described three components that influence this attitude within the three levels described above. These components are the cognitive component, the affective component and the conative component. The cognitive component is being measured through the perception of risk of people. It contains the estimation of the risk of becoming a victim or a relative becoming a victim (Rosenberg & Hovland, 1960).
The affective component is in relation to the feelings someone has when thinking about security. In research about the affective component, multiple feelings are being tested such as anger, fear, sadness, stress and concern. The feeling a person has varies. Also, the intensity of these feelings varies among people (Rosenberg & Hovland, 1960).
The conative component refers to the behavior and the probability of executing certain actions due to security issues. This behavior is derived from the previous two components, the cognitive and the affective component. The perception of risk and the feeling someone has determine the level of tendency of someone to act. In this research, avoiding certain places is an example.
Based on the above, the first hypothesis of this research is:
H1: Someone who does not want the Netherlands to help the foreign fighters return, show a higher level of fear of terrorism than people who do think the Netherlands should help bring the foreign fighters back.
2.3 Descriptive statistics
The descriptive statistics in this research is the information about the living situation of the respondent and about the respondent him- or herself. In this research, age, gender, level of education, working situation, religion, ethnicity and political preference is tested. Previous researches (eg. Lane, 2013; Nellis & Savage, 2012, Nellis, 2009) report about the difference in gender and level of fear of crime. According to these researches, females are more likely to have a higher level of fear of crime in general. Wilcox, Ozer, Gunbeyi and Gundogdu (2009) conducted a research about gender differences in the level of fear of terrorism. They found that the level of fear of terrorism differentiated significantly between the genders, with females to be more likely to report higher on feelings of fear.
In addition to the first hypothesis, where it is believed that people who have a higher level of fear of terrorism will be more against the return of foreign fighters, the second part of the hypotheses is:
H2: Females report more often against the return of foreign fighters than men. H2a Females report a higher level of fear of terrorism than men.
H2b: Females report a higher perception of risk of becoming a victim themselves than men.
The second descriptive statistic is the level of education. Multiple researches show a
difference in fear of terrorism and level of education. For instance, Balnaves and Aly (2008) report a higher level of fear of terrorism among lower educated people than the level of fear of terrorism among higher education people. As well as Sjöberg (2004), who says that lower educated people gave larger perceived risk ratings than higher educated people. Therefore, the hypotheses are:
H3: Lower educated people report more often against the return of foreign fighters than people who are higher educated.
H3a: Lower educated people report a higher level of fear of terrorism than people who are higher educated.
H3b: Lower educated people report a higher risk of becoming a victim of terrorism than people who are higher educated.
Another descriptive statistic is religion. The Islamic State is obviously Islamic. As Borum and Fein (2017, p. 262) state in their research: “We know that it is as misleading to assume
ideology is the root of all foreign fighting as it is to ignore ideology entirely”. Chassman (2017) also tries to explain an ideological commonality as an explanation to who joins the Islamic State. She quotes Malet (as quoted in Chassman, 2016, p. 229): “In order to make sense of foreign fighter volunteering, one has to assume the existence of subjective grievances linked to an expanded notion of nationhood or nonmaterial selective incentives (in this case afterlife rewards), or a collective belief or ideology”. Because of this Islamic root, the next hypothesis is:
H4: Muslims show a more positive attitude towards the return of foreign fighters than non-Muslims.
Furthermore, the researcher will look at the other demographic variables as well to see if there are any particularities in the results. The researcher will look at, for instance, age, because Sjöberg (2004) found that older people reported higher on the perceived risk of becoming victim of terrorism. The researcher wants to know if this also results out of this research. Moreover, whether or not the respondents have children will be tested in order to see if having children influences your opinion about returning foreign fighters and terrorism. Does it make you more concerned about an acquaintance becoming a victim of an attack or not? These things would be interesting to look at, so that will be done at the end of the results-chapter, in chapter 4.5.
3 Methodology
In this chapter, the methodology of this research is explained. The first part will explain on what kind of research this one is. The second part is about the data collection. Next, the
surveys are described, what kind of questions, who, where, what? The inclusion and exclusion factors are described as well. Moreover, the literature used for the interviews and the method is briefly elaborated. The next part is about how the surveys are conducted. Furthermore, there is a part about the data analysis. And lastly, the validity and reliability of the method is described. The surveys were conducted in the period from the 8th of June 2019 until the 31st of
3.1 Kind of research
In this quantitative exploratory research, a survey study will be performed. The study is meant to answer the question: Are Dutch adults who are more afraid of terrorism, more against the
return of foreign fighters than people who are less afraid of terrorism?
The main hypothesis of this research is:
H1: Someone who does not want the Netherlands to help the foreign fighters return, show a higher level of fear of terrorism than people who do think the Netherlands should help bring the foreign fighters back.
H1a: Someone who shows a higher level of fear of terrorism, has a more negative attitude towards returnees than someone who shows a lower level of fear of terrorism.
The other hypotheses of this research are:
H2: Females report more often against the return of foreign fighters than men. H2a Females report a higher level of fear of terrorism than men.
H2b: Females report a higher perception of risk of becoming a victim themselves than men.
H3: Lower educated people report more often against the return of foreign fighters than people who are higher educated.
H3a: Lower educated people report a higher level of fear of terrorism than people who are higher educated.
H3b: Lower educated people report a higher risk of becoming a victim of terrorism than people who are higher educated.
H4: Muslims show a more positive attitude towards the return of foreign fighters than non-Muslims.
In order to test these hypotheses, a minimum of 200 respondents was needed to draw a conclusion which could be related to the population of the Dutch society. The total of
3.2 The survey
In this section, an elaboration is given on the survey. The researcher chose to use surveys because this is a suitable manner to explore in views, experiences, and beliefs of individual respondents on specific matters. It is also a suitable method for a quantitative research. In order to get the respondents, the survey, which was designed with Google Forms
(www.forms.google.com), was posted on Facebook and Instagram with the request to fill it in. The survey took about 10 to 15 minutes and was conducted in Dutch. 66 people responded in the first two days, 212 people responded in total. The English translation of the survey can be found in Appendix I.
The survey started with a short introduction to inform the respondents about this research. The goal of the research was explained and the reason behind the research as well. The respondents were informed about the participation and how it was voluntarily, they could stop any moment if they wanted to and the survey and the answers were being used for this research. Furthermore, the respondents could read about the confidentiality and anonymity of their participation. For the full text, see Appendix I.
The survey was conducted by the researcher in order to answer the research question. The survey contained 52 questions. Most questions were closed, some questions were open, such as: origin of the respondent, their father and their mother and the question about a different opinion about the policy around returnees. The standard questions about
demographics were closed. These questions asked for the gender, education, job situation, residence, religion, whether they have children, and political preference.
Most other questions, which are explained below, had to be answered on a Likert scale. Jamieson (2004) explains that the Likert scale is a tool to measure the intensity of a feeling or opinion, rather than a simple yes or no question. A Likert scale is most often used as a 5-point scale, for example 1 (totally disagree) to 5 (totally agree). In this research, this was the most suitable way to measure the attitude of the respondents against terrorism and returnees and their level of fear of terrorism.
The other questions started with seven questions to measure the opinion about the returning foreign fighters, such as: ‘If a returnee would be moving in on my block, I would feel less safe than I feel now’. These questions were answered with a Likert scale of 5 points, with 1 (totally disagree) to 5 (totally agree). Moreover, a short introduction on the term ‘returnees’ is given and two flows in political opinions are given. Option 1: The Netherlands should help foreign fighters return, after which they are tried and reintegrated. Option 2: The Netherlands should leave foreign fighters where they are and they should be tried there. The
respondents could choose with which option they agreed most and could also give their own opinion if it differed from the two options in an open question.
In the next part of the survey, the cognitive component against fear was tested by six questions about the perceived risk of becoming and/or about a relative becoming a victim of terrorism. For instance: “How big do you think the chance is that you become a victim of terrorism?” with a Likert scale from 1 (very small) to 5 (very big). To measure the affective component, twelve questions were asked about certain feelings the respondents experience most in specific situations. An example is: ‘What feeling do you experience most when you see terrorism-related images or texts?’. The respondents could choose between stress, fear, concern, anger and sadness. Followed by two questions: ‘How often do you experience this feeling?’ and ‘To what extent do you experience this feeling?’. Both follow-up questions had to be answered on a 5-point Likert scale (1, almost never – 5, very often; 1, barely – 5, strongly). Another question to measure the affective components was: ‘To which extent do you have faith in the fellow man?’ with a Likert scale from 1 (none) to 5 (a lot).
To measure the third and last component, the conative component, five questions were asked to see if the fear of terrorism of respondents causes certain actions (Rosenberg & Hovland, 1960). An example is: ‘How often do you avoid certain places because of your concern about terrorism? (i.e. city squares, festivals, stations, airports)’. The answer is again on a 5-point Likert scale with 1 (never) to 5 (very often).
Lastly, the respondents were thanked for their time and offered the e-mail address of the researcher if they had any questions or wanted to know the outcome of the research.
3.3 Respondents and ethical aspects 3.3.1 Respondent recruitment
To answer the question about the Dutch society, the survey was conducted among random Dutch citizens. Therefore, the survey was in Dutch. Because of the topic, terrorism and returning foreign fighters, the minimum age of the survey was 18 years as this topic is heavy loaded. The respondents were recruited in my own social network. I posted several messages on my LinkedIn, Facebook and Instagram account. I also asked family members, coworkers and friends to share the message in their own social networks.
The respondents could reach my survey by clicking on a link. That link sent them to the Google Forms page containing the survey. My supervisor and I set the bar of minimum respondents on 200 to write a good and reliable research. In total 212 people responded to my call for help. The respondent minimum of 200 was confirmed by the sample calculator of
SurveyMonkey (https://nl.surveymonkey.com/mp/sample-size-calculator/). With a margin of error of 7% and a population of approximately 13,500,000 people. The population was based on the numbers of the Central Bureau of Statistics (CBR, 2019) about registered adults in the Netherlands. The sample size calculator indicated a minimum of 196 respondents to write a research with a reliability of 95% which is good for a quantitative research.
3.3.2 Ethical aspects
In every research, it is important to keep ethical aspects in mind. Baarda (2009) wrote a guide about how to write a research and enlightens the aspects of ethics when performing a
research. He states that the ethical requirements in every research come down to: the respondents participate voluntarily, the respondents are clearly informed about the goal and method of the research, the handling of the information of the respondents is confidential and preferably anonymous, the outcomes of the research are in no way negative for the
respondents, and the research is honest and objectively performed (Baarda, 2009).
First of all, all respondents reacted voluntarily to the survey due to the messages that were posted on social networks. Moreover, they were informed about their voluntary service in the introduction of the survey and were told that they could stop the survey anytime they wanted. In the introduction, the respondents are also informed about the goal and method of the research without spoiling too much in order to prevent biased outcomes. With this, the second ethical aspect is covered.
The information of the respondents is completely anonymous and therefore, the third aspect is maintained. The information about the respondents is handled confidential as well. The outcome of the research will not be negative for the respondents as they are anonymous and will not be used in a possible negative way.
The last ethical aspect is about the researcher. I am confident I performed this research in an honest way and will be as objective as I can possible be.
3.4 Variables
The survey questions measure all the variables needed to answer the main question of this research. In this chapter, all the variables are explained and it is shown how they are
constructed for further research. First, the demographic variables are explained. Second, the independent variables of this research are enlightened and explained. Lastly, the dependent variables are covered.
3.4.1 Demographic variables
With performing a survey, a lot of researches use the same demographic variables, such as: gender, age, education, religion, and job situation. Additionally, in this research, questions about political preference, children, living situation, and origin are asked in the survey. These variables are important to measure, in order to get to know your respondent group and to be able to see differences between different groups. Different groups could be male and female, lower educated and higher educated people, or younger and older people. It is interesting to see if there are any differences between these groups because the results are more applicable on society as everyone is different and could not be seen as one and the same. On this note, the demographic variables will be used as independent variables in this research.
3.4.2 Independent variables
In addition to the demographic variables as independent variables, the main independent variables are the components of attitude against the fear of terrorism. These components were the cognitive component, the affective component and the conative component. As explained in chapter 3.2, in order to measure the cognitive component, a 6 items-subscale was used. Participants were asked to rate the items on a scale from 1 (very small) to 5 (very large). In the statistics program, a scale was designed from the eight items. The scale reliability was α = .832, M = 3.03; SD = .72. The higher you get on the scale, the higher your perception of risk is when it comes to terrorism.
To measure the affective component, a total of 22 questions were asked. The first 15 were the questions on which feeling the respondents experienced most when in certain situations, with the questions about frequency and intensity of that feeling. As these are all kinds of different questions, it is not possible to construct one scale. The questions need to be tested separately. The two questions about trust in the government to protect the citizens, were combined into a scale. This scale had a reliability of α = .807, M = 3.33; SD = .97, the higher you score, the more trust you have in the government.
The conative component was measured by using a 5 items-subscale. Participants were asked to rate the items on a scale from 1 (never) to 5 (very often). In the statistics program, a scale was designed from the five items. The scale reliability was α = .710, M = 2.84; SD = .77. The higher you get on the scale, the higher the tendency to act based on the cognitive and affective components.
These components of fear are used as dependent variables as well in the hypotheses of this research. This happens when the demographics are being used as independent variables. For instance, to see if the perception of risk differs between males and females, or in age.
3.4.3 Dependent variables
In addition to the components as dependent variables, the dependent variable used in the main hypothesis (H1) of this research is the opinion about the return of foreign fighters (against or in favor of the return). This variable was measured with a 5 items-subscale. Participants were asked to rate the items on a scale from 1 (totally disagree) to 5 (totally agree). In the statistics program, a scale was designed from the five items. First, two of the five items had to be recoded because the direction of the question was different from the others. This was done in order to develop a reliable scale. The scale reliability was α = .786, M = 3.49; SD = .84. The higher you get on the scale, the more against you are about the return of foreign fighters.
Moreover, a question was asked to clearly ask the opinion of the respondents. It started with an explanation of two flows of beliefs within the political environment around the return of foreign fighters. One was in favour of helping them return and prosecuting the returnees in the Netherlands and option 2 was against helping returnees return to the Netherlands and in favour of prosecution over there.
3.5 Data analysis
After collecting the data of the respondents, the data analysis had to be performed. For the data analysis, the results of the survey are being tested in the program IBM SPSS Statistics 24 with the help of the book of Andy Field (2013): Discovering statistics using IBM SPSS Statistics. Field (2013, p. 4) explains that statistics are a suitable method to test certain theories. The theories in this research are the hypotheses described earlier.
3.5.1 Descriptive statistics
In order to generally oversee the differences between groups (age, gender, education, etc.), the descriptive statistics and the frequencies were tested. When you run a descriptive statistics test on data you get a certain table with all kinds of information. In table 1, an example of a descriptive statistics table is given. As shown, a clear overview is given about the data. The upcoming tables do not contain numbers, this will be done in the next chapter Results. First, you see what question the information is about. N is the indicator for the number of respondents, so in this research N should be 212. Next, the mean of the answers is
given. This gives you an insight in the most popular opinion. On a Likert scale, a mean higher than 3 is closer to the 5 (i.e., totally agree) and 2 or lower is closer to the 1 (i.e., totally
disagree). However, further analysis should be performed to determinate whether or not these differences in mean are significant. Minimum and maximum shows the lowest and highest score for that question. The std. dev. stands for standard deviation which is, according to Field (2013, p. 27) a measure of the dispersion or spread of data around the mean. A large standard deviation assumes that the data points were far from the mean and a standard deviation of 0 would indicate that all scores were the same.
Question N Mean Minimum Maximum Std. Dev.
Question 1 Question 2
Table 1. Descriptive statistics
To oversee the data, a frequencies table is made in SPSS. An example of this table is shown in table 2. These tables show the information per question and for every possible answer.
Moreover, if this example had been a question with a Likert scale, there would have been 5 groups. First, the table shows the frequency of the answer given for that question. Next, the percent and valid percent is given. Last, the cumulative percent is given so the last should make 100 percent.
Question 1
Group Frequency Percent Valid Percent Cumulative Percent
Group 1 Group 2
Table 2. Frequencies
3.5.2 One-way ANOVA
One way ANOVA is a statistical test to check if there are significant differences between three or more groups. When a one-way ANOVA is performed in SPSS, three tables are presented. The first one is the descriptive statistics table as in table 1. The next table is the ‘test of homogeneity of variances’ table. In which the Levene’s test will be shown. The
Levene’s test tests the hypothesis that the variances of each group are equal (Field, 2013, p. 464). Such a table is shown in table 3. In this table, it is important to look at the significance of the Levene Statistic. When it is significant (< 0,05) the variances are not equal, so it is important that the Levene Statistic is not significant (> 0,05).
Question 1
Levene Statistic df 1 df 2 Sig.
Table 3. Test of homogeneity of variances
The last table is the ANOVA table (see table 4). Here is per question tested what the
differences are between the groups and within the groups. The most important information is the F-score and the significance. When the significance is less than 0,05, there are significant differences between or within the groups. This could only be done if the Levene’s test is not significant and so when all variances are equal.
Question 1 Sum of Squares df Mean Square F Sig. Between groups Within groups Total
Table 4. One-way ANOVA
3.5.3 Chi-square test
To find differences between categorical groups like gender, a Chi-square test was performed, hopefully in order to find a significant difference. When a Chi-square test is performed in SPSS, you get three different tables. The first one is the crosstab table, see table 5. In this table, the answers per group are presented. Next is the Chi-square test table, see table 6. Whenever the Pearson Chi-square is significant (<0,05), the differences between the groups are significant. However, sometimes the Chi-square is not significant because of the group is too small. The Likelihood ratio is the second measure you can look at. When this one is significant, the differences are still significant. However, with a group of 212 respondents the
Chi-square test should be significant. When the differences are significant, you could presume the hypotheses. Question 1 1 2 3 4 5 Total Gender Male Female Rather not say Total Table 5. Crosstab Value df Sig. Pearson Chi-Square Likelihood Ratio N of Valid Cases
Table 6. Chi-square test
3.5.4 Correlation
To see difference in, for example, age, a correlation analysis is performed to see if this variable is coherent to the level of fear of terrorism. Correlation measures the coherence between two variables. It tests the relationship between two variables. Field (2013, p. 263) explains that we can express this relationship between variables statistically by looking at two measures: covariance and correlation coefficient. For this research, when relationships are examined, the Pearson’s correlation efficient is tested. Correlation tests if there is a relationship, it does not test causality. In this research, a two-tailed correlation will be performed in order to see the positive and negative relationships.
Correlation (r) is also presented in a table, see table 7. In this table, the direction, strength and significance of a correlation is shown. Direction and strength is noted in the Pearson Correlation column. With a two-tailed correlation, the score could go from -1 to 1. The closer to -1 or 1, the stronger the correlation is. Around 0 is not a strong correlation or even barely
related to each other. From an r of -0.08 and lower or of 0,08 and higher, you could say the variables are strongly related to each other (Field, 2013).
Between all variables, correlation tests are being performed to see where the relations are between the variables.
Variable Variable 1 Variable 2 Variable 1 Pearson Correlation -
Sig.
Variable 2 Pearson Correlation -
Sig.
Table 7. Pearson’s Correlation
3.6 Reliability, validity and limitations
In this part of the chapter, the reliability, validity and limitations of this research are explained. Every research deals with issues on these aspects. The possible issues for this research are described below.
3.6.1 Reliability
The reliability of this research is due to the number of respondents (212) and the calculated minimum of respondents (196) reliable with a percentage of 95%. The reliability of the constructed scales for the components were all larger than 0,7 which means that the scales are reliable enough. When a reliability analysis on all the questions of the survey was performed, the α = .640 which is too low. When the variable age was deleted the α became .906. This is a very reliable scale, but it should not be possible to calculate this with all the different answer categories.
The research is not excluded from incidental errors. An incidental error could be that a respondent did not understand a question and just filled something in or the respondent accidentally clicks on an answer but meant to click on another. Incidental errors could influence the reliability of a research. Moreover, the respondents could fill in the survey wherever they wanted as it was possible to fill in the survey at a smartphone. So, the researcher has no control over the environment during the data collection.
Moreover, the researcher asked various acquaintances to check the survey for errors or uncertainties in order to be sure everyone would understand the questions correctly. This process contributes to the reliability of a research. Also, the researcher obligated each
question, unless you stopped the participation, in order to prevent respondents accidentally forgetting one question.
3.6.2 Validity
When discussing the validity of a research, there are to components to think about: the
internal validity and the external validity. Below, both components are explained in relation to this research.
3.6.2.1 Internal validity
Internal validity is the level in which a causal conclusion is justified based on a certain research. It could be called the methodological validity. An issue in this research for the internal validity is the social network of the researcher that is used to collect respondents. However, the researcher tried to ask family members, coworkers and friends from other environments to share the survey in order to get a group of respondents that reflects the Dutch society. However, it is still possible that the survey was filled in by more young people than older people, or more people who live in urban areas instead of rural areas.
The anonymous survey is designed in order to avoid socially desirable answers, which could affect the internal validity. However, it is not certain that the respondents did not give socially desirable answers.
3.6.2.2 External validity
The external validity is about the generalizability of the outcomes. Due to the choice for quantitative research, this becomes an important aspect. More often than qualitative research, quantitative research is expected to be generalizable to the entire population (in this case Dutch adult society). As explained with internal validity, due to recruiting respondents in social network, the generalizability could be tainted. However, the researcher tried to prevent this by inviting more people to share. The demands for filling in the survey were: resident of the Netherlands, age 18 or older and speaking the Dutch language.
3.6.3 Limitations
Arising from the internal and external validity threats explained above, is one limitation of this research. Very often, survey research is an oversimplification of social reality. As is in this research. Every person is different and with 212 respondents, it is hard to say that the outcomes are the same as the social reality. Respondents do not have the chance to argue their
decisions made in the survey. For a more detailed sketch of the Dutch opinion, interviews would be better. However, this was not possible in the period of time or the researcher only had a few respondents.
4 Results
In this chapter, the results of the data analysis (chapter 3.5) are presented. Starting with the demographics of the respondents in order to get to know the population that is examined. Next, the dependent variables and the independent variables are presented. Lastly, the hypotheses are tested and the results are presented before chapter 5 Conclusion starts.
4.1 Population
A total of 212 persons filled in the survey. 104 respondents were male (49,1%) and 106 were female (50%), two respondents filled in ‘rather not say’ (0,9%). The respondents were between 18 and 82 years old with an average age of 35.
Almost half of the respondents’ highest education is HBO, 99 respondents (46,8%). Furthermore, 14 respondents were master educated (6,6%), 38 were MBO schooled (17,9%), 21 respondents only completed high school (9,9%) and 40 respondents were WO schooled (18,9%). Most respondents were from urban places, 170 respondents lived in cities or suburbs (80,2%) and only 42 respondents were from rural environments (19,8%). The respondents were from different religious backgrounds. Moreover, 145 reported non-believers (68,4%). Only 8 respondents reported to be Muslim (3,8%) and the most others were Catholic (18,4%) or Protestant (6,6%).
71 respondents have children (33,5%) and 141 reported that they do not have children (66,5%). Most respondents have fulltime jobs (45,8%), have part-time jobs (22,2%) or are student (21,7%). Three respondents have a part-time job together with being a student (1,4%).
Political preference is almost equally divided between the three directions. Two other options were included: ‘None’ and ‘Rather not say’. Leftwing preference was reported by 59 respondents (27,8%), rightwing by 60 respondents (28,3%) and middle by 60 respondents as well (28,3%). 30 respondents reported that they do not have a political preference (14,2%) and 3 respondents preferred to keep it to themselves (1,4%).
4.2 Variables
4.2.1 Attitude towards returnees
In the survey, an elaboration was given about two different flows of opinions about the return of foreign fighters in the Dutch political environment. Option 1 was that the returnees are going to be helped to return, after which the returnees would be prosecuted here and helped with their reintegration. Option 2 was that returnees should be left there and prosecuted there. There as in Syria or Iraq. Only 59 respondents (27,8%) voted for option 1 and 153
respondents (72,2%) voted for option 2. Moreover, a 5 items-attitude subscale was covered in the survey, α = .786, M = 3.49; SD = .84, to measure the attitude against returnees. The higher you score on the scale, the more against returnees you are.
4.2.2 Fear of terrorism
To measure the level of fear of terrorism, three components were tested: the cognitive, the affective and the conative component. Two of the three components were constructed into a scale. The cognitive component, or risk perception, had a reliability of α = .832, M = 3.03;
SD = .72. The conative component, or action-related items, had a reliability of α = .710, M =
2.84; SD = .77. The affective component, which measured on the one hand the frequency and intensity of negative feelings against terrorism-related items. On the other hand, there was risk sensitivity, trusting feelings towards government and fellow man, and protection mechanisms.
4.3 Exploratory analysis
For the exploratory analysis of the data, a correlation analysis was performed with the most important variables and displayed in Table 8. According to the expectations, there was a significant positive correlation between the attitude against returnees and two of the three components of attitude against fear (r = .40, p < .001; r = .27, p < .001). So, the more negative the attitude of the respondent, the higher his or her fear of terrorism is (cognitive and
conative). However, this was a bivariate correlation analysis, so the opposite relationship is possible as well. The attitude against returnees also has a positive significant correlation with age (r = .17, p < .05), which would mean that the more negative your attitude against
returnees, the older you are, or the other way around.
The correlation between the cognitive component and the conative component is pretty strongly positive and significant (r = .52, p < .001), logically. As these components are related to each other to form the attitude against fear. Similar with the correlation between the
attitude against returnees and the respondents’ choice for option 1 or 2 in the survey (r = .57,
p < .001). It goes without saying that these two are related to each other because they both
measure the same opinion.
Interesting to see is that the cognitive component is the only variable who significantly correlates with gender (r = .15, p < .05) and the only component who significantly correlates with age (r = .22, p < .001), both correlations are positive.
Table 8
Means, Standard Deviations, and Correlations Between Opinion about returnees and Additional Variables (N = 47) Variable 1 2 3 4 5 6 1. Opinion return – 0.57** 0.40** 0.27** -0.13 0.17* 2. Option 1 or 2 – 0.26** 0.15* -0.01 0.05 3. Cognitive – 0.52** 0.15* 0.22** 4. Conative – 0.27 0.10 5. Gender – -0.06 6. Age – M 3.49 1.72 3.03 2.84 1.52 34.74 SD 0.84 0.45 0.72 0.77 0.52 14.81 Note. *p < .05. **p < .01 4.4 Hypotheses testing
All hypotheses were tested using IBM SPSS Statistics 23. Various statistical tests were used, such as multiple and singular linear regression, ANOVA and MANOVA, Chi-square tests, and correlation.
4.4.1 Main hypothesis
The main hypothesis of this research is: Someone who does not want the Netherlands to help
the foreign fighters return, show a higher level of fear of terrorism than people who do think the Netherlands should help bring the foreign fighters back. Together with H1a: Someone who shows a higher level of fear of terrorism, has a more negative attitude towards returnees than someone who shows a lower level of fear of terrorism.
To start, only the cognitive component - or the risk perception - was tested as independent variable. A singular linear regression was calculated to predict the attitude towards returnees based on the risk perception. A significant regression equation was found (F(1, 210) = 40.506, p < .001), with an R2 of .162. This indicates on a significant effect of
level of risk perception on the level of negative attitude towards returnees. Moreover, a Chi-square analysis was performed on the effect of risk perception between the respondents who chose for option 1 and the respondents who chose for option 2. There was a significant association between the choice of respondents between the options and the level of risk perception, X2(22) = 36.68, p < .001. This outcome indicates that the level of risk perception
has a significant effect on the respondents’ choice for option 1 or 2.
Second, the conative component was tested as independent variable. A singular linear regression was calculated to predict the attitude towards returnees based on the conative component. A significant regression equation was found (F(1, 210) = 16.567, p < .001), with an R2 of .073. This indicates on a significant effect of level of conative component on the
level of negative attitude towards returnees. For testing the option choice of the respondents, a Chi-square analysis was performed on the effect of the conative component. No significant association was found between the choice of respondents between the options and the level of risk perception, X2(19) = 18.08, p > .05. This outcome indicates that there is no significant
effect between the conative component of attitude against fear and the decision between option 1 or 2.
For the affective component, the 22 questions that measured the affective component were tested in a multiple regression model with the attitude towards returnees as dependent variable. The first block was tested with the intensity and frequency of the negative feelings around terrorism related topics. In the second block, the three items of trust towards fellow man and the government were added. In the third block, risk sensitivity was added and in the fourth block, protection mechanism was added. Looking at the ANOVA table that was drawn out due to this multiple regression analysis, all blocks are significant (p <.001). This indicates that measuring attitude towards returnees with these variables is significantly better than
predicting the outcome using the mean. When running the same analysis for the option choices, not all blocks are significant. The first block, with the questions about negative feelings, their frequency and intensity, is not significant (p = .341). The other blocks are significant, p < .05. This outcome indicates that predicting the option chosen by the respondents is not significantly better when using the 15 questions about negative feelings and their intensity and frequency. However, it is significantly better if you use all the 22 questions of the affective component.
4.4.2 Other hypotheses
The other hypotheses in this research were: H2: Females report more often against the return
of foreign fighters than men. H2a: Females report a higher perception of risk of becoming a victim themselves than men. H2b: Females report a higher level of fear of terrorism than men. To test these hypotheses above, the category ‘Rather not say’ was removed so only two
groups remained: male and female. Leaving the total group at 210 respondents with 104 males and 106 females. This approach makes it easier to test the hypotheses. A singular linear regression was calculated to predict the attitude towards returnees based on gender. A
significant regression equation was found (F(1, 208) = 4.779, p < .05), with an R2 of .023.
This indicates on a significant effect of gender on the level of negative attitude towards returnees. However, the regression equation indicates on a negative effect on the attitude towards returnees which indicates that men are more negative towards returnees than females. Furthermore, a Chi-square analysis was performed on the effect of gender between the
respondents who chose for option 1 and the respondents who chose for option 2. No
significant association was found between the choice of respondents between the options and the gender, X2(1) = 0.01, p > .05. This outcome indicates that there is no significant effect
between the gender and the decision between option 1 or 2.
To test hypothesis H2a, a singular linear regression was calculated to predict the risk perception based on gender. No significant regression equation was found (F(1, 208) = 3.806,
p > .05), with an R2 of .018. This indicates that there is no significant effect of gender on the
level of risk perception. This also indicates that gender does not have an effect on the cognitive component of attitude against fear.
Hypothesis H2b is tested with the use of a Chi-square test. The conative component is used as dependent variable and gender as independent. No significant effect was found,
X2(19) = 16.72, p > .05. A multiple ANOVA (MANOVA) analysis was performed to measure
statistically significant difference in the affective component based on gender, (F(17, 192) = 1.746, p < .05). This indicates that the gender does have an effect on the affective component of attitude against fear.
The following hypotheses are: H3: Lower educated people report more often against
the return of foreign fighters than people who are higher educated. H3a: Lower educated people report a higher risk of becoming a victim of terrorism than people who are higher educated. H3b: Lower educated people report a higher level of fear of terrorism than people who are higher educated. In order to test the first hypothesis (H3), a one-way ANOVA
analysis was conducted to compare the effect of education on attitude towards returnees. The analysis of variance showed that the effect of education on attitude towards returnees was not significant, F (4, 208) = 2.084, p > .05. This indicates that education does not have an effect on the attitude towards returnees. A singular linear regression was performed as well, no significant regression equation was found (F(1, 210) = 1.709, p > .05), with an R2 of .008.
This indicates that there is no significant effect of education on the attitude towards returnees. The variable with the options was tested in a Chi-square test. No significant association was found between the choice of respondents between the options and the education, X2(4) =
6.354, p > .05.
Testing hypothesis H3a was done by running a one-way ANOVA analysis. The
analysis of variance showed that the effect of education on risk perception was not significant, F (4, 207) = 1.507, p > .05. This indicates that education does not have an effect on the
perception of risk. In addition to this outcome, for hypothesis H3b, another one-way ANOVA was performed. The analysis of variance showed that the effect of education on the conative component was not significant, F (4, 207) = 1.078, p > .05. This indicates that education does not have an effect on the conative component of attitude against fear.
For testing the affective component of attitude against fear a MANOVA was conducted. Using Roy’s Largest Root, there was a statistically significant difference in the affective component based on education, (F(4, 207) = 2.072, p < .05). This indicates that the education does have an effect on the affective component of attitude against fear.
The last hypothesis of this research is: H4: Muslims report less often against the
return of foreign fighters than non-Muslims. In order to test this, two groups had to be created
in SPSS ‘Muslims’ and ‘Non-Muslims’. Eight respondents reported to be Muslim. With those groups, a singular linear regression analysis was performed. There was a statistically
significant difference in the affective component based on gender, (F(1, 210) = 3.201, p < .05). This indicates that the religion, or being a Muslim, does have an effect on the affective
