Master’s Thesis HRM
How to decrease defensive behavior when discouraging people to smoke with moral judgments? An experimental study to the moderating effects of identification with society on motivation to quit smoking.
Date: 08-07-2016
Student: Simone Willigers (s1957996) Supervisor: S. Täuber
Abstract
This study focuses on smoking and ways to discourage people to smoke. While people assume moral framing is effective when discouraging smoking, earlier research has shown moral framing might evoke defensiveness instead of motivation to change. With this study we answer the research question: How can we decrease defensive behavior when discouraging people to smoke with moral judgments? We argue the way to do this is to take identification with smokers and moral judgments into account. People want to belong to a group, and when they are morally judged they might experience fear of rejection and exclusion from their group. This fear of rejection and exclusion is the underlying mechanism of peoples motivation to change their behavior: They change their behavior to conform to the norms of their group, to prevent rejection and exclusion from the group. A boundary condition we take into account is identification with society. Every smoker belongs to society as well. In this study we study the effect of the identifications to these two groups, and try to find favorable conditions in which people will behave less defensive and more motivated to quit smoking.
2
Introduction
Problem
Smoking causes harm to people who smoke and to those around them. Second-hand-smoking, smoking because you are nearby someone who smokes, can cause serious health issues, such as increases of the risk of lung cancer (20%), heart diseases (20-30%), lower weight and height of children (20-40%), and infections and airway symptoms by children (20-50%) (Gezondheidsraad, 2016). Because of these severe consequences of smoking, the government tries to find ways to encourage people to quit smoking.
The Dutch government for instance undertakes action to discourage people to smoke.
The Dutch government stimulates programs which help people to stop smoking, and introduces new regulations, such as smoking restrictions in pubs (Rijksoverheid, 2016).
Besides these regulation, the government is putting its hopes on awareness in order to discourage smoking. This can be done in two ways: appealing to people’s morality or using competence-related appeals. The government tries to discourage smoking by making smokers aware of the negative consequences of smoking, by obliging warnings on the packages of ciagerettes (Wetten overhead, 2016), hereby choosing not for competence-related appeals but for morality-related appeals. However, the actions of the government are not successful: In 2014, still 24,9% of the Dutch population smoked, and 19,1% of the Dutch population smoked every day (CBS statline, 2016).
When discouraging people to do certain behavior there are two possible outcomes:
people are either motivated to change their behavior, or they are defensive, not willing, to change their behavior. The government’s actions are not successful, meaning people react defensively on the government’s actions. To discourage people to quit smoking, we need to find ways for people to react motivated, not defensive. We need to find a way to make people behave less defensively. When do people react motivated, and when do they react defensive?
Earlier research has shown the way people react depends on the way the behavior is framed. For example, the behavior can be framed in an immoral way, as the government is doing with smoking. This does not mean smoking is in fact immoral behavior: When one smokes alone, with no other people nearby who could experience damage, many would agree that this form of smoking is not immoral behavior. The government is using an immoral priming to make people think they behave immoral and make them aware of this immoral
3 behavior, in order to discourage them to smoke. However, earlier research has shown framing behavior as immoral is not an effective way to discourage the behavior (Täuber, van Zomeren, & Kutlaca, 2015). It appears to have the opposite effect: Moral judgments lead to defensiveness to change their behavior. Non-moral judgments, on the other hand, seems to make people more motivated to change their behavior. When we apply these results to smoking, this would mean moral judgments would make smokers unwilling to change their behavior, while non-moral judgments from would make them more willing to change their behavior.
In this study we try to find a way in which moral judgments can be useful in discouraging smoking. We try to answer the research question: How can we decrease defensive behavior when discouraging people to smoke with moral judgments? The answer on this research question can help the government. The government will be able to adjust their regulations in order to make them more effective, and thereby do a better job at discouraging people to smoke.
Current study
In this study we will expose smokers to moral and non-moral judgments about smoking.
Based on the earlier studies of Täuber and van Zomeren (2012; 2013), we expect people will fear to be rejected by other people, will want to maintain their current identity of smoker, and will not be willing to change their behavior. This study will extend the earlier studies in several ways. In earlier studies on moral judgments, the groups compared to each other were two distinct groups: an in-group and an out-group (Täuber & van Zomeren, 2012; 2013).
Instead of comparing two distinct groups to each other, we are interested in the effect of judgments while dealing with two in-groups. This means we will not compare smokers to non-smokers as two separate groups. Instead, in this study we will compare smokers to the rest of society. As smokers are a subgroup of the larger group society, they will be compared to their own other in-group, society. Besides the focus on two in-groups instead of on an in- group versus an out-group, we extend the earlier studies in another way. In earlier research, the moral and non-moral judgments came from a third out-group, who had nothing to do with the comparison between in-group versus out-group (Täuber & van Zomeren, 2012; 2013). In our study, the judgments smokers are exposed to will come from society. This means the judgments will come from an in-group, moreover from the same in-group they are being compared with.
4 Including the in-group society in the study as comparison group and the group the moral judgments come from, might be a way to motivate smokers to quit smoking. Appealing on their sense of belonging to a larger group than solely the group smokers might get them to care more about their membership of society than about their membership of the group smokers. The focus of the current study on two in-groups is an innovative aspect of this study, but therefore it is hard to predict the behavior of smokers when confronted with judgments from their own in-group. We take their identifications with both groups into account, for these might be predictors of their reactions (see Appendix 1 for the conceptual model). When smokers identify with the group smokers, they would want to conform to the norms of the group smokers, but when smokers identify with the group society, they would want to conform to the norms of society. To what norm will smokers want to conform when they identify with smokers as well as with society?
5
Theory
Social Identity Theory
We will start with describing the Social Identity Theory. This theory helps us explain why people choose to behave defensive and keep smoking, or why they would be willing to commit to the norms of society. A defensive reaction would mean they keep their social identity of smoker, while a motivated reaction would mean they give up their social identity of smoker.
The social identity theory derives from peoples fundamental needs in life. It is a fundamental need for people to belong to groups and they define themselves often in terms of the groups they belong to (Deci & Ryan, 1985). One of the first things people see about themselves when introducing themselves is about their social identity, for example “I am a student”, or “I live in the city Groningen”. According to the Social Identity Theory, social groups define the identities of the group members. The group members define themselves as a group member, as a part of that certain group (Hogg, 1992; Hogg, Terry, & White, 1995).
This does not just mean how people describe themselves, but much more: Social groups prescribe the attitudes, perceptions and behaviors of the group members. The sense of social belongingness deriving from a social identity motivates people to make favorable comparisons between their own group and out-groups (Tajfel & Turner, 2004). Groups strive for positive distinction from other groups. When comparing themselves with relevant outgroups, belonging to such a positive group enhances the self-concept and self-esteem of the group members, and reduces uncertainty about people’s self-concept (Turner & Tajfel, 1979; Hogg, 2006). Social identity has a positive impact on well-being and health, on work, and on life satisfaction: Social identity is associated with support and appreciation from the group, which has healing effects and protects individuals from sickness and stress. Because of this, social identity has long-term effects on health, well-being and morale (Haslam, Jetten, Postmes, & Haslam, 2009). The group members want to keep being part of their group and not be rejected by them, for this would mean an identity loss. Losing a social identity will have consequences for their self-esteem and self-concept (Hogg, 2006).
6 Defining concepts
Social identity
According to the Social Identity Theory discussed above, social groups determine how people define themselves. This part of their identity, defining oneself in terms of affiliation with a group, is called their social identity (Hogg, Terry, & White, 1995). It is the part of their identity of which people define themselves as a group member, according to the perceptions, characteristics, and attitudes. Because they share all these aspects with the group, people identifying with a group tend to follow the group norms. This indicates that people’s social identity determines their behavior (Haslam, Jetten, Postmes, & Haslam, 2009).
People can belong to more groups at the same time. Their identification with multiple groups can differ from each other: some social identities are stronger than others. However, the salience of social identities can differ over time as well (Hogg, Terry, & White, 1995). In the current study we will focus on two different social identities. We study the group smokers as part of the in-group smokers and as part of the in-group society. Therefore, we will focus both on identification with smokers and on identification with society. The extent of identification with smokers and with society might be determining the reaction of people: it might be explaining why they would choose to keep smoking or to change their behavior.
This will depend on the strength of the two identities: when smokers identify with smokers, they will probably keep smoking, but if they identify with society, they would be more willing to change their behavior. In this study, identification with smokers will be manipulated as a dichotomous variable: identification is either high or low. Identification with society will be measured as a continuous variable on a scale from 1 to 7.
Moral judgments
We have discussed the importance for people to belong to groups to have a social identity.
Social groups prescribe the behavior of group members. Now we will discuss what happens if people behave differently than their social group prescribes. They will risk being judged by their group for their behavior. Social judgments can come in various form, for instance moral judgments. Moral judgments are social judgments concerning the immoral behavior of others.
Social judgments are only moral judgments when certain circumstances are in order, and they are not based on the behavior itself, but on the effect the behavior might have (Van Bavel,
7 Packer, Haas, and Cunningham, 2012). Behavior is morally judged by group members in case of two conditions: when the behavior is causing harm to others, and when the individual is responsible for his own behavior (Mulder, 2014). According to Mulder (2014), both factors need to be present in order to define behavior as immoral. When one of those factors is missing, no moral judgments of the behavior will be made. When a certain behavior causes no harm to others, the behavior might be accepted by the group. However, when this same behavior causes harm to other people, the behavior might be judged. The other aspect, own responsibility, is just as important. When the individual who performs the behavior is not responsible for the behavior, this behavior will not be judged as immoral.
We will demonstrate this with two examples. When the behavior causes harm to others but is not the own responsibility of the individual performing the behavior, the behavior and the caused harm were not his fault, thus the behavior will not be morally judged. An example of this is when someone causes a car accident, because he had a head injury. He could cause harm with this accident, although the accident is not his own responsibility. This behavior will not be judged as immoral. When a behavior is the own responsibility of the individual performing the behavior, but causes no harm to others, other people could disagree with the behavior but not morally judge it. For example, when someone only smokes when no one is around, in an environment where it will not bother anyone (for example in his own home), this behavior will not cause any harm to other people, and thus not be judged as immoral behavior. In conclusion, group members will make moral judgments when the behavior both causes harm to others and is the own responsibility of the individual performing the behavior.
Moral judgments are often unpleasant: They suggest to the target person he is not competent enough to behave in a moral way (Monin, 2007). According to Monin (2007), moral judgments can have three negative effects: people can think they are not capable of behaving morally, people can get confused about their own behavior and whether they behave moral or not, and people can experience resentment because of the moral judgments. People resent being morally judged by others, even if the judgments are incorrect (Monin, 2007).
In this study, we study the effect of moral judgments. To be able to distinct moral judgments from other sorts of judgments, we include non-moral judgments in the study as well. Non-moral judgments are judgments based on competence. It suggests the behavior is caused by incompetence of the target person, instead of immorality. The judgments participants in this study will be exposed to will be manipulated as either moral judgments or non-moral judgments.
8 Fear of rejection and exclusion
First, we have discussed peoples fundamental need to belong to a group. People are attached to their group, and they define themselves as part of the group. Then we discussed what happens to people who deviate from group norms: They risk being judged by their group.
Now we will turn to what happens if people risk being judged by their group: They would experience fear to be rejected and excluded. When they would be rejected and excluded from the group, they would lose their social identity. Because people want to belong to their social groups and keep their social identities, social judgments can evoke fear of being rejected and excluded from the group. This is not an unrealistic fear: People who deviate from the norms of the group they belong to are treated worse than people who behave in the same way but do not belong to the group (Marques & Páez, 1994). Especially immoral behavior can be a threat for exclusion from the group (Opotow, 1990). According to Gausel and Leach (2011), moral judgments, rather than non-moral judgments, have a negative effect on the individual’s self- image.
Fear of rejection and exclusion by their group might be an important predictor in why people choose to defend their behavior or to change it. People who fear being rejected by society might be more willing to quit smoking, than people who do not experience fear. In this paper we deal with different groups, namely the larger group society and the subgroup smokers. When further mentioned in this paper, fear of rejection and exclusion will cover fear of rejection and exclusion from the larger group society.
Motivation and defensiveness
Above we have discussed the importance of belonging to a group, the judgments people face when they behave differently than their group, the fear of rejection and exclusion this will cause them. Last, we will discuss the two possible reactions when people are confronted with this situation: motivation or defensiveness.
People do not want to be excluded from their group. When they risk being rejected, they will make sure they behave according to the group norms to prevent this. This study includes two groups: smokers and society. The group norm of smokers would be to keep smoking, while the group norm of society would be to quit smoking. When a smoker keeps smoking, he would conform to the norms of his in-group smokers, but risk being rejected by his in-group society. When a smoker quits smoking, he would lose his smoker identity, but
9 conform to the norms of his in-group society. To which group norms people will conform depends on the judgments they are exposed to and the group they identify with more, for the different reactions have different consequences for their group memberships.
Group members are more motivated to restore group status when the threat concerns a moral domain, compared with a threat concerning incompetence, a non-moral domain (Täuber
& van Zomeren, 2012; Täuber & van Zomeren, 2013). Moral judgments will make a smoker more defensive and more unwilling to change his behavior. Non-moral judgments, however, will decrease the defensiveness and make a smoker more motivated to change his behavior.
Besides the judgments it matters which social identity is more important. When group members experience a moral threat for their group status, they focus on defending their social identity and keeping the status quo, rather than improving their social identity (Täuber & van Zomeren, 2013). This means that a smoker who identifies with smokers would want to defend his smoker identity and keep smoking. He would not want to improve his behavior and be motivated to quit smoking, because this would be losing his group status. However, when a smoker identifies more with society, he would not care for his smoker identity. He would not want to defend his smoker identity and probably be more motivated to quit smoking to conform to the norms of society.
Hypotheses
As we have discussed all constructs of this study above, we will turn to our expectations for this study. First, we will discuss the effects of identification with smokers and moral judgments on the reaction of the smokers: motivation to quit or defensiveness. Then, we will explain the expected underlying process of these effects and the boundary conditions.
The effect of identification with smokers and moral judgments on motivation to quit smoking
The importance of identification with smokers and moral judgments on the reaction of the smoker is already explained above. Moral judgments lead to defensiveness, while non-moral judgments lead to motivation. When confronted with moral judgments, high identification with the target group of the judgments leads to defensiveness, while low identification does not (Täuber & van Zomeren, 2012; 2013). However, the groups studied in these earlier studies are positive groups, while smoking is a negative group. This difference changes some of our expectations.
10 For high identifiers with the group smokers nothing changes. People who identify high with the group smokers might see their own group more positive and not care about a negative group image. Therefore, we expect high identifiers to have the same results as in earlier studies, they will be more motivated to quit smoking by non-moral judgments. We expect different results for people who identify low with the group smokers. According to Becker and Tausch (2014), people do not want to belong to groups of which memberships are negative, problematic, or painful. They might want to distance themselves from these unwanted group identities by physical escape, if possible, or disidentification. The smoker identity is easy to escape physically, because people can change their behavior and stop smoking. Therefore we expect low identifiers to be as motivated to quit smoking for both moral and non-moral judgments.
Hypothesis 1: High identifiers with smokers will be more motivated to quit smoking when confronted with non-moral judgements rather than moral judgments.
The mediating effect of fear of rejection and exclusion on motivation to quit smoking
Now we will try to explain and predict the underlying process of these effects of identification with smokers and moral judgments. Moral judgments have a negative impact on an individual’s self-image and on the social image (Gausel & Leach, 2011). People who are judged for not behaving according to the group norms risk their social identity. Because they behave different, not according to the norms, they risk being rejected (Marques & Páez, 1994). Thus people who are judged for smoking do not behave according to the societal norm and risk being rejected from society. As we discussed above, rejection from the group means loss of social identity, which has consequences for peoples self-concept. Because of these consequences people do not want to lose their social identity: the risk of rejection may provoke fear. We argue fear of rejection and exclusion may be the explanation of the link between identification with smokers, moral judgments and motivation.
Especially immoral behavior is a threat for exclusion (Opotow, 1990). According to Alicke (2000), when an individual behaves less morally than other people, the individual may experience fear of exclusion and rejection. Based on this we expect moral judgments provokes more fear of rejection and exclusion than non-moral judgments. Because people fear being rejected by society, they would want to prevent this by conforming to the norms of society. Out of fear of being rejected by society, smokers would be motivated to quit smoking. This way, they would be able to keep being a group member of society.
11 Hypothesis 2: For moral judgments, rather than for non-moral judgments, fear of rejection and exclusion mediates the relationship between identification with smokers and motivation to quit smoking.
The moderating effect of identification with society on motivation to quit smoking
Now we will turn to the boundary condition of the effect of identification with smokers and moral judgments on motivation to quit smoking. For the moral judgments come from society, we expect identification with society matters in how much people care for these judgments and how they will react on them. When they identify highly with society, they probably would care more for the moral judgments and be more motivated to change their behavior, than when they do not identify with society.
In earlier research, distinction is made based on whether individuals are “insiders” or
“outsiders” in their different effects on moral judgments. Insiders are the people who belong and identify with the group. Moral judgments may cause them to experience guilt, remorse and self-blame for their immoral behavior. Outsiders are the people who do not belong and not identify with the group. They may not experience any response (Opotow, 1990). This would suggest smokers who identify with society experience guilt, remorse and self-blame and be motivated to change their behavior, while smokers who do not identify with society may not experience anything and will not be motivated to change their behavior. In this study, however, we do not deal with an in-group and an out-group, but with two in-groups: a larger group society and a subgroup smokers. Although people might not identify with society, they still do belong to society. According to Alicke (2000), the relationship between individual and the group is a possible moderator of the effect of moral judgments and the consequences of this behavior. This would suggest identification with a group has the same boundary effect as whether someone belongs to a group. Therefore we expect identification with society functions as a moderator of the effect of identification with smokers, moral judgments and motivation to quit smoking.
The extent to which people identify with society and with smokers can differ, and dominate over each other. How motivated people are to change their behavior may differ across the different extents of identifications. People who identify highly with society want to conform to the norms of society: They would probably quit smoking. However, they belong to the group smokers as well and will identify to some extent with the group smokers as well.
When they quit smoking, they would lose their smoker identity. When smokers identify more
12 with society than smokers, it is likely to assume they are motivated to quit smoking in order to behave to the norms of society. When smokers identify more with smokers than with society, they would probably want to conform to the norms of the group smokers. This would mean they will be defensive to quit smoking. However, do the extents of identifications to both groups vary in the same amount as the final motivation of the smoker? Where is the line between defensiveness and motivation to change?
We expect the following effects of identification with society as boundary condition.
For people who identify low with society, we expect identification with society will have no effect. For these people we expect the results described earlier: High identifiers with smokers will be more motivated when confronted with non-moral judgements rather than moral judgments. Low identifiers with smokers will be motivated to quit smoking when confronted both with moral and non-moral judgments. For people who identify highly with society, we expect some differences in the effects of identification with smokers and moral judgments.
First, high identifiers with society who identify low with smokers will be motivated to change their behavior for both moral and non-moral judgments, because they would want to get rid of the membership of a negative group, as described above. However, high identifiers with society who identify high with smokers as well will have different results. When people identify with both groups, with two different group norms, they will never lose their social identity completely: They always have one group left which they identify highly with.
According to the Social Identity Theory, people behave defensively because they would lose their social identity if they are rejected by the group (Turner & Tajfel, 1979). But in this case, they would never lose their social identity completely, so it would not be necessary to behave defensively. In this situation, people would behave less defensive. This means that if their identity of smokers is threatened by moral judgments, they do not have to fear exclusion, because they still belong to the group society. We expect high identifiers will be more willing to change their behavior when they identify with society as well.
Hypothesis 3: Identification with society moderates the effect of identification with smokers on motivation to quit smoking for moral judgments. For moral judgments rather than non- moral judgments, we expect high identifiers with both smokers and society will be more motivated to quit smoking.
13
Method
Participants
This study was set up as an online study and distributed online in America via AmTurk.
Participants of the study could earn £1.05 dollars for their participation. In total, 130 people participated, all of them were smokers. Participants who only filled in the survey partly and participants who did not pass the attention check were deleted from the study. After deletion of these participants, 122 participants were left. 72 participants were male, 50 participants were female. The participants had a mean age of 37.55 (standard deviation of 10.67). The highest education level of the participants was a high school diploma (n=57), bachelor’s degree (n=45), master’s degree (9), associate’s degree (6), PhD (3) and technical education (2). The current occupation of the participants was employee (81), self-employed (21), unemployed (10), retiree (6), and student (4). The nationality of the participants was mostly American (120), one of the participants was Canadian and American and one of the participants was Indian. The participants were randomly assigned to the four conditions. In two conditions were 30 participants, in the other two conditions 31.
Procedure and manipulations
When participants started the survey, they were shown an informed consent. In the informed consent, information about the topic of the study, voluntary participation, confidentiality and anonymity, time and rewards were explained (see Appendix 2).
Next, the participants filled in a questionnaire for identification with society (see Appendix 3). To measure the participants identification with society, we used a questionnaire for measuring in-group identification by Leach et al. (2008). This questionnaire measures group-level self-investment as well as group-level self-definition. The questionnaire has 20 items. Participants had to indicate the extent of agreement with statements such as: “I feel a bond with Americans” and “I am similar to the average American person”. The questionnaire has a Cronbach’s alpha of 0.97.
Then, the participants were manipulated for the two independent variables. The participants were first manipulated for their identification with smokers, second for moral or non-moral judgments. In this way, the judgments could not influence their identification with smokers. To manipulate their identification with smokers, participants were told they would
14 do a test to measure their identification with smokers. The test was a questionnaire of six items to measure nicotine addiction (Fagerström, Heatherton & Kozlowski, 1990). Two items were added to lay more focus on identification instead of addiction, for example: “Do you feel connected to other smokers?” (see Appendix 4). After the test, participants had to wait for a few seconds, to make it seem their test scores were being computed. Then their test results were shown, with an identification either below or above average. As a manipulation check, the participants had to answer the question: “According to the test, my identification with other smokers is” on a scale of 1 to 7, 1 being below and 7 being above average. The manipulation of identification with smokers did not work for 5 participants.
For manipulation of judgments, participants were shown the following text about tobacco smoking: “Tobacco smoke is enormously harmful to your health. Due to its severe consequences, the American society wants to discourage smoking. According to representative polls, public opinion is that 80% of the population sees smoking as …”.
Depending on the condition, the dots were filled with immoral or incompetent. When they had read the text, the participants had to answer the following question: “According to representative polls, 80% of the population sees smoking as” on two items on a scale of 1 to 7. For the first item, 1 was immoral, 7 moral. The second item, 1 was incompetent, 7 competent. The manipulation of judgments did not work for 7 participants.
Questionnaires
After the manipulations the participants filled in a questionnaire for fear of rejection and exclusion (see Appendix 5). Fear of rejection and exclusion is measured with two scales of the questionnaire used by Gausel et al. (2012; 2015). The scales we used to measure fear of rejection and exclusion are the scales concern for condemnation by others, rejection regarding health and lifestyle and moral failure. In total, the participants had to indicate for 12 statements the extent to which they agreed on a scale of 1 to 7. Examples of questions are
‘I feel rejected when I think about my smoking” and “I feel condemned by other people because I smoke”. We took the three scales together as a measurement for fear of rejection and exclusion. This questionnaire for fear of rejection and exclusion has a Cronbach’s alpha of 0.94.
The last questionnaire the participants had to fill in was a questionnaire to measure their motivation to quit smoking (see Appendix 6). It was a questionnaire of 6 items.
Participants had to indicate the extent of agreement with the items on a scale of 1 to 7.
15 Examples of items are “I am motivated to stop smoking” and “I am going to change my smoking habits”. The questionnaire for motivation to quit smoking has a Cronbach’s alpha of 0.94.
16
Results
We will start this result section with the outcomes of the manipulation check. For identification with smokers, participants had to indicate whether their identification with smokers was below or above average on a scale from 1 to 7. We performed an ANOVA in which we added identification with smokers as factor and the answer on the scale from 1 to 7 as dependent variable. The ANOVA shows significant results (F(1,120)=2456.73, p<.001), indicating a successful manipulation of identification with smokers. The same is done for the manipulation of judgments. Participants have indicated whether people thought of smokers as immoral or incompetent, both on a scale from 1 to 7. We compared both answers on the immoral and incompetent question, the immoral question used to check for the manipulation of moral judgments and the incompetent question used to check the manipulation of non- moral judgments. The ANOVA shows significant results for moral judgments (F(1,120)=84.44, p<.001) and for non-moral judgments (F(1,120)=37.00, p<.001), indicating a successful manipulation of judgments.
Next, we will discuss the results we have found in our study. At first we conducted a multivariate ANOVA to see if the covariates age and gender have any effect on the other variables. In this way it would be possible to decide whether we would analyse the data according to our expectations, or to analyse the data separately for different demographic groups. Identification with smokers, judgments and gender are added as factors, age as covariate and fear of rejection and exclusion and motivation as dependent variables. The interaction effect of identification with smokers, judgments and gender is significant:
F(2,111)=3.12, p=.048. This suggests a difference in effects for males and females. We will analyze the data separately for males and females.
The mean scores and standard deviations on fear of rejection and exclusion and motivations are computed again for the different conditions for males and females separately (see Appendix 7). Some differences are already visible: female participants are more motivated to quit smoking when confronted with non-moral judgments, while male participants are more motivated to quit smoking when confronted with moral judgments.
To analyze the direct effects of identity with smokers and judgments on motivation, a two-by-two ANOVA is conducted. Identity with smokers and judgments are added as factors in the ANOVA, motivation as dependent variable. Gender is added as moderator, to take differences in the effects for males and females into account. The ANOVA shows no
17 significant results (see Appendix 8), meaning we have found no support for our first hypothesis, that high identifiers with smokers are more motivated to quit smoking when confronted with non-moral judgments rather than moral judgments.
Mediating effect of fear of rejection and exclusion
Now we turn to the mediation effect of fear of rejection and exclusion. Based on the descriptive statistics of males and females on fear (see Appendix 9) we can see different patterns for the different judgments. For both male and female participants, the different judgments seem to have an opposite effect for different identifications with smokers. We will therefore analyze the data separately for high and low identification with smokers. We start with low identification with smokers. We perform the mediation analysis PROCESS in which we add judgments as independent variable, motivation to quit as dependent variable, fear of rejection and exclusion as mediator and gender as moderator. Fear has a significant positive effect on motivation to quit: B=.35, t=2.85, p=.006. When examining the indirect effect of judgments through fear on motivation, the bootstrap interval gives indication for a mediation effect only present for males. For males, the bootstrap confidence interval is completely above zero, while for females this interval includes zero. This suggests an effect of judgments through fear on motivation for males. For females probably no such effect exists. We interpret this mediating effect of fear for male participants with a Sobel test. We perform two regression analyses for males and low identification with smokers in which we include judgments as independent variable and motivation as dependent variable. In the second one we include fear as independent variable in the regression model. The Sobel test comparing the coefficients is not significant (test statistic=1.49, p=.136). This means that although the data indicates a mediating effect of fear for males and low identification with smokers, no significant mediating effect is found in our data.
The same analyses are done for high identification with smokers. The interaction effect of judgment and gender on fear is significant (B=1.21, t=2.51, p=.015), suggesting a different effect of judgments on fear for males and females. Fear has again a positive effect on motivation to quit: B=.39, t=3.13, p=.003. The bootstrap intervals for males and females indicates a mediation effect for females only: the interval is completely above zero, while for males the bootstrap confidence interval includes zero. This suggests an effect of judgments through fear on motivation, only for females. For interpretation of this effect we perform regression analyses and a Sobel test, which gives a significant result (test statistic=1.73,
18 p=.08). This means that for females, fear mediates the relationship of judgments on motivation to quit. The regression analyses suggest that non-moral judgments lead to more motivation to quit, through more fear of rejection and exclusion.
Based on these results we have found some partial support for a mediating effect of fear, only present for female participants. However, we did expect to find this mediating effect for moral judgments, rather than for non-moral judgments. Instead, we found this to be true for non-moral judgments. Based on these results, we reject our hypothesis that fear mediates the relation between identification with smokers and motivation to quit for moral judgments. Based on the results we can say there is a mediating effect of fear for non-moral judgments, rather than for moral judgments.
Moderating effect of identification with society
At last we will examine the boundary condition of our model, identification with society. We will analyze the effects separately for male and female participants. We conduct a two-by-two ANOVA in which we add identification with smokers and judgments as factors, identification with society as covariate and motivation as dependent variable. The interaction effect of identification with smokers, judgments and identification with society is significant (F(3, 64)=6.35, p=.001). Next we do a regression analysis in which we include identification with smokers, judgments and identification with society, and their interactions in all possible combinations. The regression analysis is significant (F(7,64)=3.12, p=.007, R2adj=.17). The three-way interaction is again significant (β=-5.15, t=-4.08, p=<.001). The interaction of identification with smokers and identification with society is significant (β=4.14, t=3.45, p=.001). The interaction of judgments and identification with society is significant (β=4.93, t=4.18, p<.001). To be able to interpret these interaction effects, we look at the data separately for high and low identification with smokers.
First, we will look at low identification with smokers. We do a regression analysis with identification with society, judgments, and their interaction. The regression model is significant: F(3,31)=4.56, p=.009, R2adj=.24. The interaction effect of judgments and identification with society is significant (β=1.478, t=3.242, p=.003), indicating a different effect of moral and non-moral judgments on motivation to quit for different levels of identification with society. To interpret this interaction effect we conduct a simple slope analysis, in which we break up the interaction into simple slopes at one standard deviation below the mean of identification with society and one standard deviation above the mean of
19 identification with society (see Appendix 10). It seems that people who identify high with society have higher motivation to quit when confronted with non-moral judgments rather than moral judgments (β=.50, t=2.22, p=.034). In contrast, people who have low identification with society are more motivated to quit when confronted with moral judgments, rather than non- moral judgments (β=-.59, t=-2.62, p=.013). This suggests that moral judgments lead to more motivation to quit for people who identify low with both smokers and society.
Now we will do the same moderation analyses for males and high identification with smokers. We do a regression analysis with identification with society, judgments and their interaction, F(3,33)=2.19, p=.107, R2adj=.09. The interaction effect of judgments and identification with society is significant (β=-1.50, t=-2,47, p=.019), suggesting a different effect for moral and non-moral judgments on motivation to quit smoking for different levels of identification with society. Again, a simple slope analysis is done for one standard deviation below the mean of identification with society and one standard deviation above the mean of identification with society (see Appendix 10). People who identify highly with society are more motivated to quit when they are confronted with moral judgments rather than non-moral judgments (β=-.66, t=-2.37, p=.024). People who have low identification with society are less motivated to quit smoking when confronted with moral judgments rather than non-moral judgments (β=.27, t=1.22, p=.230). These results support our third hypothesis, that people who identify both with society and smokers are more motivated to quit when confronted to moral judgments rather than non-moral judgments.
For females, the same analyses are done as the analyses for males described above.
The regression model with identification with smokers, judgments and identification with society (F(7,42)=.37, p=.913, R2adj=-.10) has no significant results.
Based on the moderation analyses, we can support our third hypothesis, that identification with society moderates the effect of identification with smokers on motivation to quit smoking for moral judgments. For moral judgments rather than non-moral judgments, high identifiers with both smokers and society are more motivated to quit smoking.
Moderated mediation analysis
At last we will do a moderated mediation analysis to test the whole conceptual model. We analyze the data again separately for male and female participants, and high and low identification with smokers. We start with the moderation effect of identification with society, to check whether this effect still exists when controlling for fear of rejection and exclusion as
20 mediating variable. We do a regression analysis in which we add judgments, identification with society, their interaction and fear of rejection and exclusion as independent variables and motivation to quit as dependent variable. For male participants and both moral and non-moral judgments, the interaction of judgments and identification with society is significant when controlling for fear of rejection and exclusion as mediating variable (identification with smokers low: β=1.24, t=2.76, p=.010; identification with smokers high: β=-1.60, t=-2.46, p=.019), meaning there still is a moderating effect of identification with society when controlling for fear of rejection and exclusion.
Furthermore, the mediation effect of fear of rejection and exclusion is computed when controlling for identification with society as moderator. Separately for male and female participants and high and low identification with smokers we conduct a Sobel-test with the interaction of judgments and identification with society as independent variable. None of the Sobel-tests are significant. This means there is no indication for a mediation effect of fear of rejection and exclusion, when controlling for identification with society. This means we have no support for our second hypothesis, that fear of rejection and exclusion mediates the relationship of identification with smokers and motivation to quit for moral judgments, when controlling for identification with society.
21
Discussion
Conclusion
The aim of this study was to answer the research question: How can we decrease defensive behavior when discouraging people to smoke with moral judgments? Hereby we focused on moral judgments, and tried to find ways in which moral judgments would be able to discourage smoking. To answer the research question we measured the motivation of our participants to quit smoking under certain conditions. We manipulated the participants in either high or low identification with smokers, and exposed them to either moral or non-moral judgments. We expected people who identified highly with smokers would be more motivated to quit smoking when confronted with non-moral judgments rather than moral judgments.
Then we tried to explain underlying processes of peoples reaction in either motivated or defensive behavior by fear of rejection and exclusion. We expected that fear of rejection and exclusion mediated the relationship between identification with smokers and motivation to quit, for moral rather than non-moral judgments. Last, we tried to test boundary conditions at which people would be more motivated to quit smoking. We expected identification with society would moderate the effect of identification with smokers and motivation to quit smoking. We expected for moral judgments rather than non-moral judgments, that high identification with both smokers and society would motivate people to quit smoking.
Findings
Now the findings of this study will be discussed and interpreted. First, we found an interaction effect of gender with identification with smokers and moral judgments. Based on this effect we decided to analyze the results for males and females separately. We did not expect to find different effects for males and females. An explanation of this effect might be that females are more influenceable and more inclined to conform to group norms, while males attach less value to the opinions of others (Eagly, 1978).
We expected to find a mediating effect of fear of rejection and exclusion for moral judgments, rather than non-moral judgments. Instead, we found for female participants a significant mediating effect of fear of rejection and exclusion for non-moral judgments.
Females might experience more fear they appear incompetent, than immoral. Earlier research
22 has already found that females are more afraid of appearing incompetent than males (Brenner
& Tomkiewicz, 1982). This might be an explanation the mediating effect of fear for females is stronger for non-moral judgments.
We found a significant result for our third hypothesis: for moral judgments rather than non-moral judgments, high identification with both society and smokers lead to more motivation to quit smoking. An explanation of this result is the following. When confronted with moral judgments, high identifiers with the group smokers would behave defensive, because they did not want to lose their social identity. When they identify high with society as well, they have another social identity to fall back on. They will never lose their social identity completely, because they still have their social identity of society member. Based on this moderating effect of identification with society we can answer our research question:
How can we decrease defensive behavior when discouraging people to smoke with moral judgments? According to Täuber and van Zomeren (2012; 2013), high identifiers with the target group react defensively at moral judgments. In our study we have extended this result by specifying boundary conditions. People who identify with society as well will not behave defensively, but more motivated to change.
Limitations
Besides the result of gender and the moderating effect of identification with society, we did not find support for our first and second hypothesis: We did not find that high identifiers with smokers were more motivated to quit smoking when confronted with non-moral judgements rather than moral judgments, and we did not find a mediating effect of fear of rejection and exclusion for moral judgments, rather than non-moral judgments. In this section we will discuss limitations of our study that might have caused the lack of support for these hypotheses.
A first limitation of this study that could be reason for the lack of support for the first two hypotheses is that the type of group in this study is not comparable with the type of groups in earlier studies. Therefore the effects of the groups in earlier studies might not be transferable. This study focused on the group smokers. Smokers have a negative image in society, therefore smokers can be seen as a negative group. Previous research was based on positive groups. We used the results of these previous studies to argue for our hypotheses, while the groups are of a different type and as such the effects might not be transferable to one another.
23 Another reason for the lack of result for the first and second hypotheses could be the importance of identification with society in our study. In earlier studies (e.g. Täuber & van Zomeren, 2012; 2013), participants were exposed to judgments from a third, unrelated group.
In this study participants were exposed to judgments from society, their own in-group.
Besides their in-group smokers, participants had their other in-group society to worry about.
These two in-groups might be the reason we did not find any significant results at first, when we did not control for the participants’ identification with society. For the third hypothesis, when we included identification with society as a moderator, we did find significant results.
Practical and theoretical implications
We have found two results, the mediating effect of fear for females and non-moral judgments, and the moderating effect of identification with society. These results have several practical and theoretical implications.
First, we will discuss implications for the moderating effect of identification with society. As we expected, moral judgments can be used to discourage people to smoke, when smokers identify highly with both smokers and with society. This result has some practical implications. The social identities of smokers and society should be emphasized when using moral judgments. We advise the government to make both social identities salient when using moral judgments to discourage people to smoke. This can be done by, for example, emphasizing both identities in advertisements on television or on the warnings at cigarette packages. National symbols on the advertisement could make the social identity of society salient (Carvalho & Luna, 2014). In this way, the government can make more effective use of moral judgments when discouraging people to smoke.
Now we will discuss implications for the effect of gender. We have made some speculations about the reason for the different effects we found for males and females: it could be females are more willing to conform to group norms and attach more value to social judgments. Not much earlier research has focused on the effects of gender when using social judgments and identification to motivate people. We suggest more research will be done focused on the specific gender effects.
24
References
Alicke, M. D. (2000). Evaluating social comparison targets. In Suls, J. & Wheeler, I. (Eds) Handbook of Social Comparison: Theory and Research.
Baumeister, R. F. (2012). Need-to-belong theory. In P. M. Van Lange, A. W. Kruglanski, E.
T. Higgins, P. M. Van Lange, A. W. Kruglanski, E. T. Higgins (Eds.) , Handbook of theories of social psychology (Vol 2) (pp. 121-140). Thousand Oaks, CA: Sage Publications Ltd.
Brenner, O. C., & Tomkiewicz, J. (1982). Sex differences among business graduates in fear of success and fear of appearing incompetent as measured by objective instruments.
Psychological Reports, 51(1), 179-182.
Carvalho, S. W., & Luna, D. (2014). Effects of national identity salience on responses to ads.
Journal Of Business Research, 67(5), 1026-1034. doi:10.1016/j.jbusres.2013.08.009
Deci, E. L., & Ryan, R. M. (1985). The general causality orientations scale: Self- determination in personality. Journal of research in personality, 19(2), 109-134.
Eagly, A. H. (1978). Sex differences in influenceability. Psychological Bulletin, 85(1), 86.
Fagerstrom, K. O., Heatherton, T. F., & Kozlowski, L. T. (1990). Nicotine addiction and its assessment. Ear Nose Throat J, 69(11), 763-5.
Gausel, N. & Leach, C. W. (2011). Concern for self-image and social image in the management of moral failure: Rethinking shame. European Journal of Social Psychology, 41, 646-478.
Gausel, N., Leach, C. W., Vignoles, V. L., & Brown, R. (2012). Defend or repair? Explaining responses to in-group moral failure by disentangling feelings of shame, rejection, and inferiority. Journal of personality and social psychology, 102(5), 941.
Gausel, N., Vignoles, V. L., & Leach, C. W. (2015). Resolving the paradox of shame:
Differentiating among specific appraisal-feeling combinations explains pro-social and self- defensive motivation. Motivation and Emotion, 1-22.
Haslam, S. A., Jetten, J., Postmes, T., & Haslam, C. (2009). Social Identity, Health and Well Being: An Emerging Agenda for Applied Psychology. Applied Psychology: An International Review, 58(1), 1-23.
Hogg, M. A. (2006). Social identity theory. Contemporary social psychological theories, 13, 111-1369.
25 Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: A critical comparison of identity theory with social identity theory. Social psychology quarterly, 255-269.
Leach, C. W., van Zomeren, M., Zebel, S., Vliek, M. L., Pennekamp, S. F., Doosje, B., ... &
Spears, R. (2008). Group-level self-definition and self-investment: a hierarchical (multicomponent) model of in-group identification. Journal of personality and social psychology,95(1),144.
Marques, J. M., & Paez, D. (1994). The ‘black sheep effect’: Social categorization, rejection of ingroup deviates, and perception of group variability. European review of social psychology, 5(1), 37-68.
Monin, B. (2007). Holier than me? Threatening Social Comparison in the Moral Domain.
Revue Internationale de Psychologie Sociale, 20(1), 53-68.
Mulder, L.B., Rupp, D. E., & Dijkstra, A. (2014). Making snacking less sinful: (Counter-) moralising obesity in the public discourse diferentially affects food choices of individuals with high and low perceived boy mass. Psychology & health, 30(2), 233-251.
Opotow, S. (1990). Moral Exclusion and Injustice: An Introduction. Journal of Social Issues, 46(1), 1-20.
Tajfel, H., & Turner, J. C. (2004). The Social Identity Theory of Intergroup Behavior. In J. T.
Jost, J. Sidanius, J. T. Jost, J. Sidanius (Eds.) , Political psychology: Key readings (pp.
276-293). New York, NY, US: Psychology Press.
Turner, J. C., Brown, R. J., & Tajfel, H. (1979). Social comparison and group interest in ingroup favouritism. European Journal Of Social Psychology, 9(2), 187-204.
Täuber, S., van Zomeren, M., & Kutlaca, M. (2015). Should the moral core of climate issues be emphasized or downplayed in public discourse? Three ways to successfully manage the double-edged sword of moral communication. Climate Change, 130, 453-464.
Täuber, S., & van Zomeren, M. (2013). Outrage towards whom? Threats to moral group status impede striving to improve via out-group-directed outrage. European Journal of Social Psychology, 43, 149-159.
Täuber, S., & van Zomeren, M. (2012). Refusing intergroup help from the morally superior:
How one group’s moral superiority leads to anouther group’s reluctance to seek their help.
Journal of Experimental Social Psychology, 48, 420-423.
Van Bavel, J. J., Packer, D. J., Haas, I. J., & Cunningham, W. A. (2012). The Importance of Moral Construal: Moral versus Non-Moral Construal Elicits Faster, More Extreme, Universal Evaluations of the Same Actions. PLoS ONE, 7(11), 1-14.
26 https://www.rijksoverheid.nl/onderwerpen/roken/inhoud/roken-ontmoedigen, acquired on 10 01-2016.
http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=83021ned&D1=0-10,19
26&D2=0-13,30-42&D3=0&D4=l&HDR=T&STB=G1,G2,G3&VW=T, acquired on 10-01 2016.
http://www.gezondheidsraad.nl/sites/default/files/0321nher.pdf, acquired on 10-01-2016.
http://wetten.overheid.nl/BWBR0004302/geldigheidsdatum_24-07-2015, acquired on 10-01 2016.
27
Appendix 1: Conceptual model
Figure 1: Conceptual model
28
Appendix 2: Informed consent
Welcome to this survey! The purpose of this investigation is to examine new perspectives on smoking behaviors. You are hereby invited to participate in this investigation. Participation is voluntary. You will be asked questions about your identification with society and smokers and about your smoking behaviors. All the information that you provide will be treated confidentially and kept anonymous. The study will take approximately 15 minutes, and you will be provided with $1.05 for your efforts. If you agree that you have read this text, that you do not have any questions regarding study participation, that you are at least 18 years old, and that you wish to participate in this study, please check the ‘I AGREE’ box and press next at the bottom of the page.
I AGREE
I DO NOT AGREE
29
Appendix 3: Questionnaire for identification with society
Scale for group-level self-investment
I feel a bond with Americans
I feel solidarity with Americans
I feel committed to Americans
I am glad to be an American
I think that Americans have a lot to be proud of
It is pleasant to be an American
Being an American gives me a good feeling
I often think about the fact that I am an American
The fact that I am an American is an important part of my identity
Scale for group-level self-definition
Being an American is an important part of how I see myself
I have a lot in common with the average American person
I am similar to the average American person
American people have a lot in common with each other
American people are very similar to each other
I see myself as an American
I identify with other American people
Being an American just feels natural to me
I feel (personally) implicated when American people are criticized
Americans are an important group to me
I have a lot of respect for American people
30
Appendix 4: Manipulation identity smokers
How soon after you wake up do you smoke your first cigarette?
Do you feel connected with other smokers?
Do you find it difficult to refrain from smoking in places where it is forbidden, e.g. in church, at the library, in cinema etc.?
Which cigarette would you hate most to give up?
How many cigarettes per day do you smoke?
Do you smoke more frequently during the first hours after waking than during the rest of the day?
Do you smoke if you are so ill that you are in bed most of the day?
Do you more often smoke together with other people or alone?
31
Appendix 5: Questionnaire for fear of exclusion and rejection
I feel condemned by other people because I smoke
I feel that other people have a strong opinion about me being a smoker
Other people give me the feeling of being antisocial because I smoke
Other people make me feel like a bad person because I smoke
When I think about me being a smoker, I think that something is wrong with me
When I think about me being a smoker, I think that I am a moral failure
When I think about my smoking, I think I could be isolated from the company of others
When I think about my smoking, I think others might not have respect for me
I feel rebuffed when I think about my smoking
I feel rejected when I think about my smoking
I feel withdrawn when I think about my smoking
I feel alone when I think about my smoking
32
Appendix 6: Questionnaire for motivation to quit smoking
I am motivated to stop smoking
I would like to work on smoking less during the next month
I am going to try to smoke less
I think I smoke too much
I think I should cut down on my smoking
I am going to change my smoking habits
At present, how much do you want to cut down the number of cigarettes you smoke?
Are you seriously thinking of cutting down the number of cigarettes you smoke?
How much do you want to quit smoking?
Do you plan to quit smoking?
If you plan to quit smoking, how determined are you to quit?
How determined are you to cut down?
If you decided to quit smoking completely, how sure are you that you would be able to do it?
If you wanted to cut down now, how sure are you that you would be able to do it?
How important for your friends is to be a non-smoker?
How important in our culture is - in your opinion - to be a non-smoker?
How important is to be a smoker to you personally?
To what extent is being a smoker in your opinion unattractive?
To what extent is being a smoker in your opinion unhealthy?
In our culture, being a smoker is considered a bad thing.
In our culture, not smoking is an important part of being attractive.
33
Appendix 7: Descriptive statistics on motivation
Table 1: Mean and standard deviations of male participants on motivation
Moral judgments Non-moral judgments Total Identification smokers high 5.01 (1.14) 4.76 (1.30) 4.89 (1.21) Identification smokers low 5.06 (1.53) 4.88 (1.75) 4.98 (1.61)
Total 5.04 (1.34) 4.81 (1.50) 4.93 (1.41)
Figure 2: Bar diagram of the scores on fear of rejection and exclusion for male participants, split by identification with smokers and clustered by judgments.
Table 2: Mean and standard deviations of female participants on motivation Moral judgments Non-moral judgments Total Identification smokers high 4.61 (1.73) 4.78 (1.49) 4.64 (1.77) Identification smokers low 4.47 (2.12) 5.11 (1.09) 4.88 (1.41)
Total 4.54 (1.88) 4.95 (1.28) 4.76 (1.58)
3 3,5 4 4,5 5 5,5 6
Moral judgments Non-moral judgments
Motivation to quit smoking
Identification smokers high Identification smokers low
34 Figure 3: Bar diagram of the scores on fear of rejection and exclusion for female participants, split by identification with smokers and clustered by judgments.
3 3,5 4 4,5 5 5,5 6
Moral judgments Non-moral judgments
Motivation to quit smoking
Identification smokers high Identification smokers low
35
Appendix 8: ANOVA of the effects of identification smokers and judgments on motivation
Table 3. Results of a two-by-two ANOVA for the main effects and interaction effect of identification with smokers and judgments on motivation to quit smoking, with gender added as moderator.
Source Sum of squares df Mean square F p
Identification smokers .34 1 .34 .32 .571
Judgments .12 1 .12 .11 .739
Gender .45 1 .45 .43 .512
Identification smokers*Judgments .05 1 .05 .05 .825
Identification smokers*Gender .07 1 .07 .07 .790
Judgments*Gender 1.30 1 1.30 1.24 .267
Identification smokers*Judgments*Gender .01 1 .01 .01 .914
Error 118.84 114 1.04
Total 121.00 121
