Do illustrations influence our
attitudes towards Muslims?
Do stereotyping illustrations of Muslims effect our attitude
toward Muslims?
Stefan J. F. Hendriks Communicatiewetenschap College of Communication Dr. A. Van Oosten
[1] ABSTRACT
In The Netherlands, Dutch citizens do not have a very positive attitude toward Muslims. Media play a role in the shaping of these negative attitudes. Former research has shown that stereotyping result in a more negative attitude. The Integrated Threat Theory states that negative stereotypes create negative attitudes toward out-group members. This study examines the effect of stereotyping illustrations of Muslims on the attitude toward Muslims. This is the first study that investigates the effect of stereotyping illustration on the attitude toward Muslims. An online experiment has been conducted among Dutch young adults (N = 62). Results show that stereotyping illustrations of Muslims indeed create a more negative attitude toward Muslims. This effect is not moderated by in-group identification or intergroup contact. These results are useful for the Dutch government and Dutch media operators,
because they have a social responsibility to create a media landscape which is unbiased and which is a reflection of our multicultural society.
INTRODUCTION
After the terrorist attacks on September 11, 2001, Muslim communities became victims of increased hostility across multiple European countries (Allen & Nielsen, 2002). The Netherlands is no exception. Within a month after September 11, 2001, the European Centre on Racism and Xenophobia reported 42 incidents of violence and hostile treatment against Muslims in The Netherlands (EUMC, 2001). The Dutch majority considered the Western and Muslim way of life as opposites that do not coincide (Gijsberts, 2005).
Furthermore, 54% of the Dutch adolescents were prejudiced toward Muslims, whereas only 21% had a positive attitude toward Muslims (González, Verkuyten, Weesie, & Poppe, 2008). Prejudice is an incorrect or unjustified attitude toward an individual based solely on their membership of a social group (Devine, 1989). In the last two years, there have been multiple terrorist attacks throughout the entire world (e. g. Brussels, Belgium; Paris, France; Kukawa, Nigeria; Baghdad, Iraq; Sinai, Egypt), claimed by different terrorist organizations. The majority of these terrorist organizations are working from their religion, which often is the Islam. These recent terrorist attacks may cause the same effect on the population’s attitude toward Muslims as it did 15 years ago, after September 11, 2001.
The media have had a big influence on the shape and reshaping of attitudes toward women, gays, and also Muslims (Adamczyk & Pitt, 2009; Brewer, Graf, & Willnat, 2003; Tower, 2000). However, never has there been any research conducted on the effect of the use
[2]
of illustrations that are used as supporting material. Previous research has shown that the use of stereotyping video material leads to prejudice (Burgess, Dill, Stermer, Burgess, & Brown, 2011). Previous research has also shown that prejudice leads to a negative attitude (Devine, 1989; Kraus, 1962). Prejudice can be explained by the Integrated Threat Theory (ITT;
Stephan & Stephan, 1993, 1996). The ITT is a basic threat model which includes four types of threats that are believed to play a role in causing prejudice. Negative stereotyping had the strongest correlation with prejudice toward Muslims (González et al., 2008), and therefore this study will further investigate this relationship.
Much research has been conducted on stereotyping, especially among gender and bodyweight stereotyping (Spitzer, Henderson, & Zivian, 1999; Stice, Schupak-neuberg, Shaw, & Stein, 1994; Tartaglia & Rollero, 2015), but research on stereotyping toward Muslims is very limited. Research has shown that media portrayals of social groups can contribute to the development and perpetuation of stereotypes, thereby influencing interpretations of and behavioral tendencies toward stereotyped targets (Oliver, Hoewe, Ash, Kim, Chung, & Shade, 2014). Stereotyping of Muslims is widely seen in news coverage (Suleiman, 1988), and therefore it is plausible that the way Muslims are presented in the news has an impact on the way people develop attitudes toward Muslims. Gonzales et al. (2008) found a correlation between the use of negative stereotyping frames and prejudice toward Muslims. However, they did not include supporting illustrations. Studies explored the possibility that the content of stereotypes consists not only of abstract trait concepts, but includes visual images (Brewer, 1988; Brewer, Dull, & Lui, 1981; Lynn, Shavitt, & Ostrom, 1985). Research has shown that stereotyping illustrations have an impact on the way people shape their attitudes (Ferree & Hall, 1990; Sever & Grillo, 2015).
The objective of the present study will be to investigate if the uses of stereotyping illustrations have an effect on the attitude toward Muslims. Besides trying to investigate the effects of stereotyping illustrations of Muslims, this study will also investigate two plausible interaction effects. One will be focusing on the effect of in-group identification based on the Social Identity Theory (SIT) (Tajfel & Turner, 1979) and the second interaction effect will be focusing on the effect of intergroup contact based on the contact hypothesis (Allport, 1954). Furthermore, this study will try to expand the current knowledge of the ITT, SIT, and the contact hypothesis. The main research question for this study is: “To what extent do stereotyping illustrations of Muslims influence the attitude toward Muslims among Dutch adolescents?”
[3]
The following paper will provide an overview of the theories concerning stereotyping and prejudice, followed by a section which will describe how the research has been
conducted. The results of this research will be described in detail, followed by a conclusion. Finally, we will discuss the research, and provide suggestions for future research.
THEORETICAL FRAMEWORK Integrated Threat Theory
The basic threat model includes four types of threat that are believed to play a role in causing prejudice (Stephan & Stephan, 2000). These four threats are: realistic threat, symbolic threat, intergroup anxiety, and negative stereotypes. Only two threats (symbolic threat and negative stereotypes) correlate with prejudice toward Muslims (Gonzalez et al., 2008). Symbolic threats are threats to the morals, values, standards, beliefs, and attitudes of the in-group (Stephan & Stephan, 2000). These threats arise because the in-in-group believes in the moral rights of its system of values. Most of prominent theories of prejudice in social psychology deal with symbolic threats. Other theories, however, argue that feeling that your values are threatened by an out-group is a form of prejudice (e.g., Kinder & Sears, 1981; McConahay, 1986; Sears, 1988), whereas Stephan and Stephan (1993) believe that feeling that your values are threatened by an out-group is a cause of prejudice. The more the group’s values, customs, or traditions are blocked by an out-group, the more negative the in-group’s attitude toward the out-group will be (Haddock, Zanna, & Esses 1993).
One of the functions of stereotypes is to serve as a basis for expectations concerning the behavior of members of the stereotyped group (Hamilton, Sherman, & Ruvolo, 1990). Because most expectations are negative, conflictual or unpleasant interactions are likely to be anticipated. Although throughout history stereotypes have been associated with prejudice (Allport, 1954; Brigham, 1971), recent studies of stereotyping and prejudice have failed to show a substantial relationship. The major reason was that most studies have ignored the valence of the traits of stereotypes. However, negative stereotypes create a social identity threat (Tajfel & Turner, 1979; Steele, Spencer, & Aronson, 2002). A social identity threat is the fear of losing one’s identity when interacting with non in-group members. Previous research has shown that by seeing Muslims, people automatically feel a social identity threat (González et al., 2008), which lead to shaping a negative attitude (Ojala & Nesdale, 2004). Therefore, it is plausible that illustrations of Muslims create the same reaction. This research will try to expand the current knowledge on the ITT by studying it from a new perspective.
[4]
Currently the ITT was only related to direct contact between individuals, this study will try to see if the same rules apply when people are exposed to stereotyping illustrations of out-group members. Based on these findings the following hypothesis is formulated:
H1: Negative stereotyping illustrations will have a negative influence on the attitude toward Muslims.
In-group identification
The Social Identity Theory (IST) (Tajfel & Turner, 1979) states that people are motivated to develop and maintain a positive sense of their social self. Establishing a
favorable evaluative distinctiveness of one’s group compared to other groups helps to achieve a positive group identity. However, negative out-group evaluation is not an automatic product of group distinctions. Territory boundaries and group differentiation can be used as a tool to prevent conflict among individuals instead of supporting it. Prejudice between in-groups and out-groups are caused by relative favoritism toward the in-group. Out-groups can be evaluated with indifference and sympathy, as long as the intergroup distinctiveness is maintained. The more people identify with their own group, the more likely they are concerned about their group interests and to consider it important to preserve their groups’ culture. Individuals who identified strongly with their in-group were more likely to perceive the presence of ethnic minorities as a threat to their culture and society; these results also apply to the Netherlands (Van Oudenhoven, Prins, & Buunk, 1998). When an in-group member sees an out-group member, he or she tries to avoid contact (Brewer, 1996). However, when avoiding contact is not possible, they will feel that their social identity is threatened (Brewer, 1996). Therefore, it is plausible that, by seeing a simple illustration of an out-group member, the same reaction might appear. This study will try to expand the current knowledge of the social identity theory. So far the social identity theory only states that individuals establish a favorable evaluative distinctiveness of one’s group compared to other groups to achieve a positive group identity. This study will investigate if they simultaneously create a more negative attitude toward these other groups. Based on these findings the following hypotheses are formulated:
H2: The stronger Dutch adolescents identify themselves with their in-group, the more negative their attitude toward Muslims will be.
[5]
H3: The stronger Dutch adolescents identify themselves with their in-group, the more powerful the effect of a stereotyping illustration will be on their attitude toward Muslims.
Intergroup contact
A meta-analysis showed that the quantity of intergroup contact has a positive effect on prejudice (Pettigrew & Tropp, 2006). The contact hypothesis states that direct interpersonal contact between two members of different social groups could lead to a less negative attitude toward the other group (Allport, 1954). However, research shows that the intergroup contact might only be successful when contact between intergroup members is as inconspicuous as possible and when the interaction is intimate (Brewer & Miller, 1984). The mere exposure theory could also explain how frequent contact with out-group members may reduce their negative attitude toward the group (Bornstein, 1989). Frequent contact with out-group members might also lead to a decategorization of group members, and thereby reduce the stereotypical thinking. People can acquire out-group knowledge through frequent contact, and thereby learn to correct negative stereotypes. Many studies have found that intergroup contact does indeed reduce stereotyping (Messick & Mackie, 1989). This study will try to support the contact hypothesis by investigating if more frequent contact with Muslims indeed decreases the negative attitude toward Muslims. Also will this study try to expend the current
explanatory power of the contact hypothesis by investigating the exposure to stereotyping illustrations and if this will influence the existing attitude. Based on these findings the following hypotheses are formulated:
H4: The more frequent Dutch adolescents are in contact with Muslims, the less negative their attitude toward Muslims will be.
H5: The more frequent Dutch adolescents are in contact with Muslims, the less powerful the effect of a stereotyping illustration will be on their attitude toward Muslims.
METHOD Design
An experiment has been conducted to measure to what extent stereotyping illustrations of Muslims effect the attitude toward Muslims among Dutch adolescents. The experiment is
[6]
the only research method that is able to show a causal relation between the content and its effect on attitudes. A between subject-design was used for this experiment. During the experiment there were two different conditions. Participants were placed randomly in one of the two conditions. After the participants were exposed to the stimulus, a post-test was conducted. The post-test consisting of a questionnaire, is intended to measure the attitude toward Muslims, in-group identification, and intergroup contact.
Sample
An experiment has been conducted among 71 Dutch young adults aged 18 through 24 (M = 20.66, SD = 2.01). In total 10 participants removed from further analysis, because they did not finish the experiment. From the 61 participants 33 were female (54.1%), and 28 were male (45.9%). The participants were college students, of which 21 were MBO educated (34.4%), 16 were HBO educated (26.2%), and 24 were WO educated (39.3%). Several trade schools, colleges, and universities located in Arnhem and Nijmegen, were randomly
approached with the request to let their students be part of a scientific research. From all students that were approached, 20% (n = 71) was willing to participate.
Material
All participants in both the experimental and the control condition were exposed to a news article concerning Muslims. The control group read a news article with a neutral frame supported with a neutral illustration. The experimental group read a news article with a neutral frame supported with a stereotyping illustration of Muslims. Participants within both groups were given five minutes to read the news article. To ensure participants read the news article, the following multiple choice question was asked at the end of the survey: ‘What was the news article about?’ None of the participants answered the question incorrect and
therefore no participants were deemed from further analysis. To ensure that the participants noticed the illustration, the following multiple choice question was asked: ‘What did you see on the illustration next to the news article?’ None of the participants answered the question incorrect and therefore no participants were deemed from further analysis. These questions were asked at the end of the survey to prevent any suspicion.
Pilot test
Before the actual experiment was executed, a pilot test was executed among a
[7]
members of the researcher. The participants of the pilot test had similar characterizations as the target group, and had no knowledge of participating in an experiment, which made them an excellent test group. The purpose of this pilot test was to confirm that questions were clearly formulated, respondents could follow the given directions, manipulation material was effective, and most importantly, all questions measure as expected. Every participant’s feedback has been reviewed seriously, resulting in some small changes to avoid uncertainties. For instance, it seemed that the questionnaire had some grammatical errors. Also an error was found in one of the scales, where participants could only score between 1 and 4 instead of 5.
Measures
Attitude toward Muslims. Participants’ completed a questionnaire with several
questions concerning Muslims. Examples of the items that were used are: “How would you feel about Muslims living in your city?”, “How would you feel about Muslims living in your street?”, and “How would you feel about a close relative marrying a Muslim person?” The items were measured on scales ranging from 1 (very pleased) to 5 (very unpleased). These items were not recoded. The scale also included one item measured on a 1 to 3 point Likert scale. Because it is unknown how to imply an item with a different measuring scale into the analysis, this item is taken out of the questionnaire. The items used were taken from the Attitude toward Muslim Proximity Index (AMPI). A factor analysis using varimax rotation showed one factor with an Eigenvalue of 4.27. The factor accounted for 85.47% of the variances. Factor loading ranged between 0.87 and 0.97. The compute option was used to create a new scale of all the variables that measured the attitude toward Muslims (MEAN (Var1, Var2, Var3, etc)). The scale is reliable with a Cronbach’s alpha of 0.95 (M = 3.02, SD = 1.00). The attitude toward Muslims Former research indicated that the six items from AMPI formed a good scale with good internal reliability with a Cronbach’s alpha of 0.81 (Brockett, Village, & Francis, 2009).
In-group identification. To measure in-group identification, participants were asked
to answer eight items that were taken from previous Dutch research (Verkuyten, 2005). These items measure the importance attached to one’s ethnic background. The items are similar to the Identity and Membership subscales of the Collective Self-Esteem Scale (Luthanen& Crocker, 1992). Examples of the items which were used are: “I am a worthy member of the social group I belong to”, “I feel I do not have much time to offer to the social group I belong to” (reversed coded), and “The social group I belong to are an important reflection of who I
[8]
am” The items were measured on scales ranging from 1 (strongly disagree) to 5 (strongly agree).
A factor analysis using varimax rotation showed two factors. When both factors are combined, in a reliability test, the scale is not reliable (The first factor has an Eigenvalue of 3.79 and measures the value of the participation one has in their social groups. The first factor accounted for 47.43% of the variances. Factor loading for the first factor ranged between 0.60 and 0.86. The compute option was used to create a new scale of all the variables that
measured the participation one has in their social groups (MEAN (Var1, Var2, Var3, etc)). The scale is reliable with a Cronbach’s alpha of 0.79 (M = 3.81, SD = 66). The second factor has an Eigenvalue of 1.29 and measures to what extend the social group reflects the
participant’s personality. The second factor accounted for 16.15% of the variances. Factor loading ranged between 0.53 and 0.85. The compute option was used to create a new scale of all the variables that measured to what extend the social group reflects the participant’s personality (MEAN (Var1, Var2, Var3, etc)). The scale is reliable with a Cronbach’s alpha of 0.77 (M = 3.33, SD = 0.74). Together both factors account for 63.58% of the variance. Former research shows that the eight items formed a good scale with good internal reliability with a Cronbach’s alpha of 0.73 (Verkuyten, 2005). A higher score indicates stronger ethnic group identification. Interesting is that they only found one factor. The reason that there is a difference in the findings might be explained by the translation from English to Dutch.
Intergroup contact. To measure intergroup contact, participants were asked to answer
four items that were taken from previous research (González et al., 2008). Examples of the items which were used are: ‘How many Muslims friends do you have?’; and ‘Do you have contact with Muslims at school?’. The first item was measured on a 4-point scale, ranging from (1) ‘none’ to (4) ‘only Muslim friends’. The other items were also rated on 4-point scales, ranging from (1) ‘never’ to (4) ‘always’. The original scale from González et al. (2008) for the last three items, used a 4-point scale ranging from (1) ‘never’ to (4) ‘often’. This has been changed because ‘often’ is not an exhausting answer. There is a possibility that participants are in contact with Muslims all the time (at school, neighborhoods, and other places), but simply are unable to notify so in the questionnaire. A factor analysis using varimax rotation showed one factor with an Eigenvalue of 2.37. The factor accounted for 59.14% of the variance. Factor loading ranged between 0.69 and 0.8. Cronbach’s alpha for the four item scale has been found to be 0.70 in former research (González et al., 2008). The compute option was used to create a new scale of all the variables that measured the
[9]
intergroup contact (MEAN (Var1, Var2, Var3, etc)). With the enhanced scale, the reliability of the scale has been improved, with a Cronbach’s alpha of 0.76 (M = 2.29, SD = 0.87).
Creating new variables
Before analyses could be executed, scales had to be created as described above. However, before executing factor analyses and reliability tests, some variables were to be recoded due to their opposite Likert scale. In total 4 variables from the in-group identification scale had to be recoded. This was done in SPSS by using the ‘recode into same variables..’ option. This was done before scales were conducted.
To execute any analyses based on the experiment and its conditions, both condition had to be merged to one variable. This was done by multiplying both variables. If a
participant took part in the controlled condition, he or she scored 1 on that variable versus 0 for the participants who did not participate in the controlled condition. If a participant took part in the experimental condition, he or she scored 2 on that variable versus 0 for the participants who did not participate in the experimental condition. Adding the variables together via ‘Compute’ making a new variable, participants would score 1 if they participated in the controlled condition or 2 if they participated in the experimental condition.
Last, to be able to execute an ANOVA, independent variables need a nominal
measuring level. Because the moderators ‘in-group identification’ and ‘intergroup contact’ are measured interval, they are equally divided in three groups: low (1 – 2.33), medium (2.34 – 3.67), and high (3.68 – 5). This is done with the ‘Recode into different variables..’ option. By doing this, the variables can be used in an ANOVA.
Variables for explorative analyses
Questions concerning demographic variables were asked solely for the purpose of explorative analyses. These questions include: age, gender, education, and religion.
Analysis
A correlation analysis was conducted between all control variables and the dependent variable. If there was a relationship between any of the control variables and the dependent variable, they were added as a covariance in the analysis in question. Levene’s test was executed before each ANOVA to ensure the usage of an ANOVA was justified. To measure the effect of the stereotyping illustration on the attitude toward Muslims a one-way Analysis
[10]
of Variances (ANOVA) was executed. Because the independent variable ‘stereotyping illustration’ is dichotomous, and the dependent variable ‘attitude toward Muslims’ is interval an ANOVA is the best way to analyze the results. When results are significant the Eta square will be mentioned.
RESULTS Sample distribution
In the controlled condition (n = 32) and the experimental condition (n = 29), the variables gender, education, and age were mostly equally divided. Also within each condition, variables were equally divided. 53.1% (n = 17) of the participants in the control group were man, and 46.9% (n = 15) were women with an average age of 20.9 years old. In the
experimental group, 37.9% (n = 11) of the participants were men, and 62.1% (n = 18) of the participants were women with an average age of 20.38. Table 1 shows an overview of the deviations of education and religion. All demographic variables were statistically tested; these results are displayed at the end of the results paragraph.
Table 1
Sample deviation variables education and religion
Education Controlled condition Experimental condition
MBO 28.1% (n = 9) 41.4% (n = 12) HBO 28.1% (n = 9) 24.1% (n = 7) WO 43.8% (n = 14) 34.5% (n = 10) Religion Non-religious 81.3% (n = 26) 79.3% (n = 23) Muslim 6.3% (n = 2) 6.9% (n = 2) Catholic 6.3% (n = 2) 6.9% (n = 2) Protestant 3.1% (n = 1) 6.9% (n = 2) Other 3.1% (n = 1) 0% (n = 0)
[11]
The sample deviation of all control variables are equally divided over both conditions. This means that if any significant differences are found these cannot result from the sample deviation, but are caused by the manipulation material.
Stereotyping illustration
Before an ANOVA can be executed, the dependent variable has to be normally distributed in every population. When creating a frequency table in SPSS, a histogram is selected with the option to show a normal curve on the histogram. However, when executed, it is notable that the dependent variable ‘Attitude toward Muslims’ is not equally divided. This might be due to the low number (n = 61) of participants. Therefore, a one-way ANOVA is still executed. A Levene’s test for Homogeneity of Variances show that the variances are not statistically different from each other (p = .628). The one-way ANOVA shows that the mean of attitude toward Muslims differ significantly between the controlled (M = 2.80, SD = 0.93) and experimental (M = 3.27, SD = 1.03) conditions, F (1, 59) = 3.49, p = 0.067, η2 = 0.06. Therefore, the first hypothesis “Negative stereotyping illustrations will have a negative influence on the attitude toward Muslims” is accepted.
In-group identification
To test the second hypothesis, the stronger Dutch adolescents identify themselves with their in-group, the more negative their attitude toward Muslims will be, 2 one-way ANOVAs were executed. The in-group identification items created 2 components, meaning that for every scale (one for each component) a one-way ANOVA was executed. A Levene’s test for Homogeneity of Variances show that the variances of the first scale are not statistically different from each other (p = 0.406). Because the independent variable has 3 groups the Bonferroni post-hoc test is selected, and will show if there are any significant differences between groups. The first scale, which measures the value of the participation one has in their social groups, shows to have no impact on the attitude toward Muslims, F (2, 58) = 0.99, p = 0.378. The Bonferroni post-hoc test shows that there are no significant differences between participants that have a low (M = 4.00, SD = 0.28), medium (M = 2.98, SD = 1.00), or high (M = 3.00, SD = 1.01) participation in their social group.
A second Levene’s test for Homogeneity of Variances show that the variances of the second scale are not statistically different from each other (p = 0.074). Because the
independent variable has 3 groups the Bonferroni post-hoc test is selected, and will show if there are any significant differences between groups. The second scale, which measures the
[12]
influence a social group has on the participant’s self image, shows to have no impact on the attitude towards Muslims, F (2,58) = 1.24, p = 0.298. The Bonferroni post-hoc test shows that there are no significant differences between participants whose social group have a low (M = 3.50, SD = 0.56), medium (M = 2.86, SD = 0.91), or high (M = 3.03, SD = 1.15). influence on their self image. Therefore, the second hypothesis “The stronger Dutch adolescents identify themselves with their in-group, the more negative their attitude toward Muslims will be” is rejected.
To test the third hypothesis, the stronger Dutch adolescents identify themselves with their in-group, the more powerful the effect of a stereotyping illustration will be on their attitude toward Muslims, 2 two-way ANOVAs were executed. One two-way ANOVA was for the first scale of in-group identification and a second two-way ANOVA for the other scale. A Levene’s test for Homogeneity of Variances show that the variances of the first scale are not statistically different from each other (p = 0.685). The first scale, which measures the value of the participation one has in their social groups, does not interact with any of the conditions of the experiment, F (1, 56) = 1.81, p = 0.184. Results of the two-way ANOVA show no main effect of the experimental condition on the attitude toward Muslims, F (1, 56) = 0.96, p = 0.332. The participation in a social group has no main effect on the attitude toward Muslims,
F (2, 56) = 0.74, p = 0.481.
A Levene’s test for Homogeneity of Variances show that the variances of the second scale are not statistically different from each other (p = 0.409). The second scale, which measures the influence a social group has on the participant’s self image, does not interact with any of the conditions of the experiment, F (2, 55) = 2.03, p = 0.141. The results show no main effect of the experimental condition on the attitude toward Muslims, F (1, 55) = 2.37, p = 0.129. The influence of a social group on the participant’s self image has no main effect on the attitude toward Muslims, F (2, 55) = 1.49, p = 0.234. The results of the two-way
ANOVAs show that the third hypothesis “The stronger Dutch adolescents identify themselves with their in-group, the more powerful the effect of a stereotyping illustration will be on their attitude toward Muslims” is rejected. However,
Intergroup contact
To test the fourth hypothesis, the more frequent Dutch adolescents are in contact with Muslims, the less negative their attitude toward Muslims will be, a one-way ANOVA was executed. A Levene’s test for Homogeneity of Variances show that the variances are not
[13]
statistically different from each other (p = 0.282). Because the independent variable has 3 groups the Bonferroni post-hoc test is selected, and will show if there are any significant differences between groups. The results of the ANOVA show a significant effect of
intergroup contact on the attitude toward Muslims, F (2,58) = 5.45, p = 0.007, η2 = 0.16. The Bonferroni post-hoc test shows that participants that have low (M = 3.22, SD = 1.04)
intergroup contact have a more negative attitude toward Muslims than participants that have high (M = 1.60, SD = 0.94) intergroup contact, p = 0.005. Furthermore, participants that have medium (M = 2.97, SD = 0.72) intergroup contact have a more negative attitude toward Muslims than participants that have high (M = 1.60, SD = 0.94) intergroup contact, p = 0.027. Therefore, the fourth hypothesis “The more frequent Dutch adolescents are in contact with Muslims, the less negative their attitude toward Muslims will be” is accepted.
To test the fifth hypothesis, the more frequent Dutch adolescents are in contact with Muslims, the less powerful the effect of a stereotyping illustration will be on their attitude toward Muslims, a two-way ANOVA was executed. A Levene’s test for Homogeneity of Variances show that the variances of the first scale are not statistically different from each other (p = 0.057). Intergroup contact does not interact with any of the conditions of the
experiment, F (2, 55) = 0.65, p = 0.526. Intergroup contact does have a significant main effect on the attitude toward Muslims, F (2, 55) = 5.40, p = 0.007. The results show no main effect of the experimental condition on the attitude toward Muslims, F (1, 55) = 2.07, p = 0.156.
Explorative results
Out of all participant’s, men (M = 2.53, SD = 0.89) show to have significant more contact with Muslims than women (M = 2.08, SD = 0.81), t (59) = 2.03, p = 0.047, d = 0.53. This means that, following the contact hypothesis, men should have a less negative attitude toward Muslims than women. However, there has been found no significant difference in the attitude toward Muslims between men (M = 3.06, SD = 1.07) and women (M = 3.00, SD = 0.95), t (59) = 0.24, p = 0.808. There is no significant difference between men (M = 3.96, SD = 0.56) and women (M = 367, SD = 0.71) on the first scale of group identification, t (59) = 1.74, p = 0.087. Neither was there a difference between men (M = 3.38, SD = 0.85) and women (M = 3.29, SD = 0.65) on the second scale of group identification, t (50) = 0.44, p = 0.659. The level of education seems to have no significant influence on the frequency of contact with Muslims, the in-group identification, or the attitude toward Muslims. In total of
[14]
all 61 participants, only 4 were Muslims, 4 were Catholic, 3 were Protestant, and 1 participant had a different religion. The other 49 participants were not religious.
CONCLUSION & DISCUSSION
This research tried to show the effect that stereotyping illustrations of Muslims might have on the attitude toward Muslims. Furthermore, the interference of in-group identification and intergroup contact were taken into account as moderators. Due to the lack of research to the influence of stereotyping illustrations, this online experiment tries to make people aware of this problem. The research question to this research was: “To what extent do stereotyping frames and illustrations of Muslims influence the attitude toward Muslims among Dutch adolescents?”
The exposure to negative stereotyping illustrations of Muslims results in a more
negative attitude toward Muslims. Therefore, the first hypothesis is accepted. This means that these results support the Integrated Threat Theory (ITT) from a new perspective. As the ITT suggests, stereotypes increase prejudice toward out-group members, and create negative expectations such as conflictual or unpleasant interactions. However, until this day there has not been any research that could show that stereotyping illustrations could create the same reaction. This is the first study that shows that stereotyping illustrations indeed create a more negative attitude toward Muslims.
Based on the Social Identity Theory (SIT), in-group identification was expected to interact with the effect that stereotyping illustration of Muslims have on the attitude toward Muslims. SIT states that people create favorable evaluative distinctiveness of one’s group compared to other groups to achieve a positive group identity. However, results show no interaction effect nor a main effect on the attitude toward Muslims. Therefore, hypothesis 2 and 3 are rejected. These results add new information to the SIT. Regardless of how strong participants identify themselves with their social group, it does not impact their existing attitude toward Muslims when exposed to stereotyping illustrations of Muslims. Also, the in-group identification has no direct effect on the attitude toward Muslims. Therefore, SIT is supported in the fact that people only create favorable evaluative distinctiveness of one’s group to increase there positive feeling of the social group they belong to.
[15]
The contact hypothesis states that direct interpersonal contact between two members of different social groups could lead to a less negative attitude toward the other group. Also, the mere exposure effect states that, the more people are exposed to something, the more positive their attitude becomes. Therefore, it was expected that participants who spend more time with Muslims would have a less negative attitude toward Muslims. Results show that people who spend much time with Muslims indeed have a less negative attitude toward Muslims than people who spend almost no time with Muslims. Therefore, hypothesis 4 is supported. These results support both the contact hypothesis and the mere exposure effect. It was also expected that people who spend more time with Muslims would be less effected by stereotyping illustrations of Muslims than people who spend almost no time with Muslims. However, results show that the frequency of intergroup contact does not interact with the effect of stereotyping illustrations of Muslims on the attitude toward Muslims. This means that hypothesis 5 is rejected. These results do not support the contact hypothesis nor the mere exposure effect. It might be that, when someone has a positive attitude toward an out-group member, their positive attitude is more easily disrupted when seeing a negative stereotyping illustration than when it would enhance an in-group member. However, only further research could explain the exact causes.
This study shows that stereotyping illustration about Muslims increase the negative attitude toward Muslims. However, the study shows a number of flaws. First, the number of participants was low (n = 62). This has caused a power problem with several analyses, leading to an increase in the chance of an error of the second kind. Second, the items of the concept in-group identifications were misinterpreted by many MBO students. It was pointed out that some choice of words, like social groups, was not clear defined causing multiple
interpretations. This might be the reason why there has been found no effect of in-group identification on the attitude toward Muslims. This might also be the reason why two components were found in the factor analyses, leading to two scales. Third, just before the data collection, a terrorist attack took place in Brussels, which could have had a temporary effect on the attitude toward Muslims. Fourth, the manipulation material was shown as supporting material to a news article. Therefore, participants might have been more
concentrated on the news article itself than the supporting illustration. Fifth, due to a mistake in the questionnaire, the manipulation test was not successful. Therefore, there is no guarantee that participants did in fact notice the manipulation material.
[16]
In case of repeating this research there are a number of suggestions one should take into account. First, follow-up studies should rewrite the items for the Dutch scale in-group identification and make sure that the concept has a clear definition which is understandable to all participants. Second, one should consider using only the illustrations as manipulation material and therefore do not use any supporting news article or any other form of supporting content. This ensures not only that the manipulation is indeed noticed, but also avoids the risk that the results could be assigned to the supporting content. Third, a higher number of
participants is recommended to overcome any power problems. Fourth, one might consider to address an older target group who are more engaged with the topic. Also, the younger
generation is used to growing up with Muslims in their direct surroundings. Older people had to adjust to the integrating Muslims, creating mixed feelings toward these immigrants (due to unemployment of Dutch citizens, unfamiliar religion, etcetera).
For further research addressing the same topic a number of suggestions are given. First, for now the effect of stereotyping illustrations only are applicable to Muslims. This means that future research should address different minorities or disadvantaged groups like homosexuals or women. Second, the previous study only studied the effect of stereotyping illustrations. It is however plausible that stereotyping video, game or music material will have the same effects as stereotyping illustrations. Therefore, it should be considered to study a different form of media content. Third, people who spend much time with Muslims have a less negative attitude toward Muslims than people who spend almost no time with Muslims. However, because hypothesis 5 was rejected it might be interesting to find out what caused this. Therefore, a research should find out why people who spend much time with Muslims are just as effected by stereotyping illustrations as people who spend almost no time with Muslims.
With these results the Dutch government and Dutch media operators are made aware of the power that stereotyping illustrations have on people’s attitudes. They have a social responsibility to create a media landscape which is unbiased and which is a reflection of our multicultural society. This was the first ever conducted experiment that tested the use of stereotyping illustrations. This is an old existing phenomenon, but sadly it has hardly ever been investigated. Especially in the world we live in today, with all our multicultural backgrounds and globalization we should pay more attention to issues which involves stereotyping and or the disadvantage of minorities.
[17]
LITERATURE
Adamczyk, A., & Pitt, C. (2009). Shaping attitudes about homosexuality: The role of religion and cultural context. Social Science Research, 38(2), 338-351.
Allen, C., & Nielsen, J. (2002). Islamophobia in the EU after 11 September 2001 [online]. Vienna: European Monitoring Centre on Racism and Xenophobia. Available from:
http://www.raxen.eumc.eu.int/1/webmill.php
Allport, F. H. (1954). The structuring of events: outline of a general theory with applications to psychology. Psychological Review, 61(5), 281.
Bornstein, R. F. (1989). Exposure and affect: Overview and meta-analysis of research, 1968– 1987. Psychological bulletin, 106(2), 265.
Brewer, M. B. (1988). A dual process model of impression formation. Advances in Social
Cognition, 1, 1-36
Brewer, M. B., Dull, V., & Lui, L. (1981). Perceptions of the elderly: Stereotypes as prototypes. Journal of Personality and Social Psychology,41(4), 656.
Brewer, M. B., Miller, N. (1984). Beyond the contact hypothesis: theoretical perspectives on desegregation. See Miller & Brewer 1984, pp. 281-302.
Brewer, M. B. (1996). When contact is not enough: Social identity and intergroup cooperation. International Journal of Intercultural Relations, 20(3), 291-303.
Brewer, P. R., Graf, J., & Willnat, L. (2003). Priming or framing media influence on attitudes toward foreign countries. Gazette, 65(6), 493-508.
[18]
Burgess, M. C., Dill, K. E., Stermer, S. P., Burgess, S. R., & Brown, B. P. (2011). Playing with prejudice: The prevalence and consequences of racial stereotypes in video games. Media
Psychology, 14(3), 289-311.
Devine, P. G. (1989). Stereotypes and prejudice: their automatic and controlled components.
Journal of Personality and Social Psychology, 56(1), 5.
EUMC (2001). Report to EUMC on ‘post September 11th developments’. Dutch Monitoring Centreon Racism and Xenophobia. Available from:
http://www.lbr.nl/internationaal/DUMC/publicatie/index.html
Ferree, M. M., & Hall, E. J. (1990). Visual images of American society: Gender and race in introductory sociology textbooks. Gender & Society,4(4), 500-533.
Gijsberts, M. (2005). Opvattingen van autochtonen en allochtonen over de multi-etnischesamenleving. In Jaarrapport integratie 2005. The Hague: SCP/WODC/CBS.
González, K. V., Verkuyten, M., Weesie, J., & Poppe, E. (2008). Prejudice towards Muslims in The Netherlands: Testing integrated threat theory. British Journal of Social Psychology, 47, 667-685.
Haddock, G., Zanna, M. P., & Esses, V. M. (1993). Assessing the structure of prejudicial attitudes: The case of attitudes toward homosexuals. Journal of Personality and Social
Psychology, 65(6), 1105.
Hamilton, D. L., Sherman, S. J., & Ruvolo, C. M. (1990). Stereotype‐based expectancies: Effects on information processing and social behavior. Journal of Social Issues, 46(2), 35-60.
Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics: Symbolic racism versus racial threats to the good life. Journal of personality and social psychology, 40(3), 414.
Kraus, S. (1962). Modifying prejudice: Attitude change as a function of the race of the communicator. Educational Technology Research and Development, 10(1), 14-22.
[19]
Lynn, M., Shavitt, S., & Ostrom, T. (1985). Effects of pictures on the organization and recall of social information. Journal of personality and social psychology, 49(5), 1160.
McConahay, J. B. (1986). Modern racism, ambivalence, and the modern racism scale.
Messick, D. M., & Mackie, D. M. (1989). Intergroup relations. In C. Sedikides, J. Schopler, and C. A. Insko (Eds.), Intergroup relations, Mahwah, New Jersey: Lawrence Erlbaum associates.
Ojala, K., & Nesdale, D. (2004). Bullying and social identity: The effects of group norms and distinctiveness threat on attitudes towards bullying. British Journal of Development
Psychology, 22(1), 19-35.
Oliver, M. B., Hoewe, J., Ash, E., Kim, K., Chung, M., & Shade, D. (2014).Media and social groups. In M. B. Oliver and A. A. Raney (Eds.), Media and social life. New York, NY: Routledge.
Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of personality and social psychology, 90(5), 751.
Spitzer, B. L., Henderson, K. A., & Zivian, M. T. (1999). Gender differences in population versus media body sizes
:
A comparison over four decades. Sex Roles, 40, 545-565.Stephan, W. G., & Stephan, C. W. (1993). Cognition and affect in stereotyping: Parallel interactivenetworks. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping:Interactive processes in group perception (pp. 111-136). Orlando: Academic Press.
Stephan, W. G., & Stephan, C. W. (1996). Predicting prejudice. International Journal of
[20]
Stice, E., Schupak-Neuberg, E., Shaw, H. E., & Stein, R. I. (1994). Relation of media exposure to eating disorder symptomatology: An examination of mediating mechanisms.
Journal of Abnormal Psychology, 103(4), 836-840.
Sears, D. O. (1988). Symbolic racism. In Eliminating racism (pp. 53-84). Springer US.
Sever, B., & Grillo, M. (2015). Do Our Books Reinforce Criminal Justice Stereotypes? An Analysis of the Images in Introductory Criminal Justice Textbooks. Journal of Criminal
Justice Education, 1-23.
Steele, C. M., Spenser, S. J., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. Advances in experimental Social
Psychology, 34, 379-440.
Suleiman, M. W. (1988). The Arabs in the mind of America. Brattleboro, VT: Amana Books.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict.The social
psychology of intergroup relations, 33(47), 74–78.
Tartaglia, S., &Rollero, C. (2015). Gender stereotyping in newspaper advertisements: A cross-cultural study. Journal of Cross-Cultural Psychology, 46(8), 1103-1109.
Tower, K. (2000). In our own image: Shaping attitudes about social work through television production. Journal of Social Work Education, 36(3), 575-585.
Van Oudenhoven, J. P., Prins, K. S., & Buunk, B. P. (1998). Attitudes of minority and majority members towards adaptation of immigrants. European Journal of Social
