1
Master’s thesis
Take care and carry one: the effects of self-persuasion campaigns on young women carrying condoms with them
______________________________________________________________________
Graduate School of Communication; Master’s programme Communication Science; University of Amsterdam Student: Esther van der Werf (10852417)
Supervisor: Marieke Fransen Date of completion: 29/01/2016
2 Abstract
This study examined the effects of a self-persuasion strategy compared to a direct persuasion strategy for safe sex campaigns motivating young women to carry a condom when they go out. Furthermore, it investigated reactance as a mediator and self-esteem as a moderator for the effects of self-persuasion. The results showed that self-persuasion leads to more positive attitude regarding carrying condoms. However, direct persuasion reduced image concerns among the target group. Intentions to carry a condom depended on both the persuasion strategy used and the levels of self-esteem of women. Women with low self-esteem had higher intentions when exposed to a persuasion poster, whereas women with high self-esteem had higher intentions when exposed to a direct persuasion poster. At last, reactance was not found to mediate the effects of self-persuasion.
Introduction
In a study regarding condom possession and carriage, Jellema et al. (2013) made a couple of comparisons between young women and men. They reported that 80% of men compared to 66% of the women indicated to have used a condom during their last intercourse.
Furthermore, they found that 41% of men and 62% of women reported not to carry condoms in general, 47 % of men and 73% of the women reported not to carry condoms when they go out, and 26% of men and 59% of women reported not to carry condoms when expecting to have sex. Apparently, women both use and carry condoms less than men. How can we explain this gender difference? Several studies suggest that women hold more positive attitudes regarding condoms than men. However, they are more inhibited than men about keeping condoms. Women are mostly concerned that others will perceive them as sexually promiscuous. Will.i.am, one of the world famous band members of the Black Eyed Peas illustrated these concerns in an interview with Elle Magazine in April 2011:
Elle: If you walked into a woman's house, what one item would convince you that you weren't compatible?
Will.i.am: If she had condoms in her house, that would just f[---]in' throw me off. That's just tacky.
Elle: Well, okay, I could see if she had a candy bowl full of them on the coffee table. But if she's got a few in a drawer, wouldn't that simply suggest she's health-conscious?
Will.i.am: I just think, like, if you're into someone and you guys get to that level, then that's something you should converse about together and say, "Hey, maybe we should get some."
3
them available. Therefore, women still rely on their male partners to buy and keep condoms (Campbell, Peplau, & DeBro, 1992; Helweg-Larsen & Collins, 1994; Sacco, Levine, Reed, & Thompson, 1991). If not only men, but also women start carrying condoms, chances on safe sex will increase. More young adolescents will be carrying a condom with them, and will be prepared in the case they get into a situation of sexual encounter. Doyle et al. (2008) also found that young women/adolescents who carry condoms with them also have greater intentions to use them.
The previously mentioned studies and the interview with Will.i.am show an unresolved issue in the field of safe sex and the possession of condoms by women in particular. Women mostly think about reasons why not to possess and carry even though they are very aware of the risks of unsafe sex. Therefore, it is unlikely that new interventions telling women to keep condoms with them supported by arguments about the risks of unsafe sex will be effective. Many safe sex campaigns focus on making people aware of the risks of unprotected sex and teaching adolescents how to have safe sex. However, the studies mentioned above showed that most women know what the risks of unsafe sex are and therefore already have positive attitudes towards condoms. For women especially, not a lack of knowledge or positive
attitudes, but rather embarrassment seems to be the actual matter that keep them from carrying them.
Another type of campaign strategy may be more effective. First of all, carrying a condom is found to be a behavior that young women are resistant to because of image concerns. Furthermore, it is a kind of behavior they find hard to change. Therefore, the effectiveness of a direct persuasion strategy which commands them to carry a condom in a forceful way may not be great. Young women may not like to be forced to carry a condom since they consider this type of behavior embarrassing and difficult to perform. Second, since women are found to focus on reasons why they should not carry a condom, it may be interesting to look at the effects of a campaign that motivates them to think about the benefits instead. This way they will focus on the health advantages of carrying a condom rather than on their image concerns. Dutch young women are found to have sufficient knowledge regarding the topic of safe sex and condom use. Therefore, they are expected to be able to come up with reasons why to carry a condom. The campaign strategy that tends to change people’s behavior by motivating them to think about the benefits of the behavior, and tends to let them persuade themselves, is called self-persuasion (also known as ‘indirect persuasion’). This type of strategy may be more effective than the presently used direct persuasion campaigns in which people are directed to carry a condom. A reason for this could be that self-persuasion is an indirect type
4
of persuasion and could therefore lead to less reactance compared to more direct messages. When people recognize the persuasive attempt of a message, they get more critical and
resistant towards it (Friestad & Wright, 1994). Young women may not like the forceful way in which present campaigns tell them to keep or carry condoms. This study’s goal is to answer the following research question:
RQ: To what extent does a safe sex campaign using self-persuasion have more effects on young women’s intentions and behavior to carry a condom with them than a safe sex campaign using direct persuasion?
Theoretical background
Self-persuasion vs. direct persuasion
According to Aronson (1999), self-persuasion is an indirect technique of persuasion contrary to the traditional direct persuasion. It tends to motivate people to persuade
themselves to change their own attitudes and behavior. Thus, the persuasion takes place in a more indirect way. Self-persuasion occurs for example when a parent is trying to change his child’s behavior by asking him why it was wrong what he did, rather than just telling him it was wrong. Another example is telling ourselves we do not like the bad food (like fries) that much when we try to eat healthier. An example of a campaign that uses self-persuasion is a leaflet of the NHS (National Health Service) England (see appendix figure 1). This leaflet motivates women to get a breast cancer screening. The leaflet shows a woman with a
‘dilemma’, balancing a benefit and risk of breast cancer screening. It does not tell them to get a screening directly but it lets them decide for themselves. It does even address a negative element of a screening (the possibility of finding and being treated for a breast cancer that would have never been life-threatening). However, receivers of this leaflet will come to the conclusion that this ‘risk’ is not really a risk compared to what could happen to them if they would not get the breast cancer screening (death). This way the leaflet leads women to persuade themselves that getting a breast cancer screening is the better option in a
non-forceful way. An example of a breast cancer screening campaign that uses direct persuasion is a campaign of Baycare Health Systems (see appendix figure 2). This campaign does not motivate women to make the decision themselves, but rather tells them that getting a screening in the right thing to do in a more direct and forceful way.
Aronson argues that for commercial advertising direct advertising can be highly successful because the change from, for example, one toothpaste brand to another does not involve big changes in lifestyle (you will only have to move your hand a little bit further to
5
the left or right of the shelf in the supermarket). However, when it bigger behavioral changes like stopping smoking or condom possession, direct persuasion is likely to be less effective. According to Aronson, the problem with direct attempts at persuasion when it comes to more important decisions stems from the very fact that they are direct. Aronson explains that when a person sees a poster or TV commercial persuading him to make an important change in behavior, he will weigh the pros and cons and ask himself ‘Why am I (not) doing this?’. He will suppose he would do it because the poster or TV commercial tells him to. However, right after that he will asks himself: ‘Do I really want this myself?’. Aronson states that with self-persuasion something different happens. With self-self-persuasion, people motivate themselves and come to believe that they really want to change their behavior (Aronson, 1999). Aronson and Carlsmith (1963) found that self-persuasion has longer lasting effects compared to direct persuasion. In their study, they let children play with some toys and after a while they forbid them to play with their favorite one. The experimenter gave half of the group a mild threat, telling them they were allowed to play with any of the toys except from a robot toy (Aronson & Carlsmith, 1963). The other half of the group of children received a severe threat in which he added that he would be very angry if they would play with the robot toy and that he would have to do something about it. Both, the children in the severe as well as the mild threat condition stopped playing with the toy. However, this effect lasted longer for the children with the mild threat. Even 2,5 months after the threat they refused to play with the toy. The researchers found that whereas the children in the severe threat condition did not play with the toy because of the severity of the threat, the children in the mild threat condition came up with arguments why not to play with the toy themselves, because the threat itself was not that severe.
The present study investigates whether the same thing applies for safe sex campaigns focusing on young women. These long term effects are very relevant for condom possession too, since it is important that men and women keep remembering to possess and carry condoms in the case a sexual encounter could be a possibility; whether this is a day or three months after exposure to a campaign. In order to reach these effects, a message that motivates women to persuade themselves to start carrying condoms (self-persuasion) may be more effective than the traditional forceful message in which they are persuaded by someone else (direct persuasion).
Self-persuasion and cognitive dissonance
6
In short, the theory of cognitive dissonance suggests that ‘dissonance’ (an unpleasant feeling) arises when a person does or says something that goes against his or her own beliefs
(Aronson, 1999). In order to reduce this dissonance, people will try to get those contradicting actions and cognitions in greater harmony. According to Aronson, one of the underlying processes of dissonance is counterattitudinal advocacy. In one of their experiments, Nel, Aronson and Helmreich (1969) asked people who were against the use of marijuana to give a speech to others, advocating the use of marijuana. They received a minimal reward in return. These people were asked to convince others of a position that was opposed to their own privately held beliefs. In order to do so, they had to find arguments for the justification of marijuana use. The result was a less negative attitude towards marijuana use. In a later study regarding condom use, Aronson, Fried and Stone (1991) encountered the problem that
counterattidudinal advocacy does not apply to condom use among adolescents because people already believe that AIDS is a problem and condom use is a good thing. Thus, regarding this topic, there is no counterattitude to advocate for the target group. However, they also found that the target group did not always practice what they preached: when you ask people about condom use, they will mostly affirm that everyone should use them. However, they often fail to use them themselves (Jellema et al, 2013; Aronson, Fried and Stone, 1991). Therefore, Aronson, Fried and Stone conducted a study among college students in which they confronted them with their hypocrisy regarding condom use. When confronted with this, their
self-concept of being a person with integrity got threatened. In order to recover this they needed to start practicing the way they preached. Aronson, Fried and Stone found that hypocrisy was indeed created by first asking their participants to persuade others to use condoms, before making them aware of their own past failure to use condoms. They found that this induction of hypocrisy lead to decreased levels of denial and this improved intentions to use condoms in the future.
However, the effectiveness of a self-persuasion strategy has never been examined in the context of mass safe sex campaigns. These kind of campaigns could be used by health organizations like for example UNAIDS, WHO and SOA AIDS Fonds Netherlands. In previous self-persuasion studies, self-persuasion was mostly forced by asking people to think about a certain behavior and let them advocate this behavior to other people in a speech for example. This study will test two kinds of self-persuasion campaign effects. First, of all, it examines the effects of a campaign when people are asked to write in advocacy of carrying condoms (active processing), Second, it investigates the campaign effects when a poster simply asks receivers to think about the benefits of carrying a condom, but does not force
7
them to write something about it (passive processing). Previous campaigns using direct persuasion have not yet lead to the desired effects: women still consider carrying condoms as embarrassing and do not take responsibility for them. However, what if they are exposed to a campaign poster using a self-persuasion strategy that motivates them to focus on the benefits of carrying a condom rather than the concerns regarding their image? Furthermore, this type of persuasion is expected to lead to more positive attitudes towards the campaign and its message (‘carry a condom’) because it does not force women to carry a condom. It rather motivates them to persuade themselves that carrying a condom is the right thing to do. H1: Safe sex campaigns using self-persuasion lead to more positive attitudes of young women to carry a condom with them, increased intentions to carry a condom, increased interest in condoms and actual carrying a condom with them, compared to direct safe sex campaigns.
Self-persuasion and reactance
The studies discussed above suggested an interesting mediator of the self-persuasion strategy, namely cognitive dissonance. This study will examine another mediator: a decline in reactance towards a campaign. This effect is expected because the persuasive intent is less obvious in a self-persuasion campaign compared to a direct persuasion campaign. For young women, lowering reactance is especially important. They are already relatively resistant towards carrying a condom, since they consider carrying a condom embarrassing and therefore hard to perform. According to Aronson (1999) the effectiveness of self-persuasion over direct persuasion is due to the following: when people are persuaded directly, they are aware of the fact that someone is trying to convince them. Friestad and Wright’s Persuasion Knowledge Model (1994) suggests that when people recognize persuasive attempts, they start processing information more critically. This leads to more counterarguing and more
reactance. However, in the case of self-persuasion, people believe that the motivation to change comes from themselves. Williams, Fitzsimons and Block (2004) found that simply asking people an intention question about a desired behavior leads to an overproduction of someone’s likelihood to engage in this behavior and ultimately results in a greater likelihood to actually engage in the behavior. They state that persuasion knowledge mediates the mere-measure effects: the effects occur because people do not recognize the persuasive intent of the intention question. However, they found these effects in the field of commercial purchases, and thus not for health communication.
8
threaten people’s negative faces, which leads them to resist the message (Jenkins &
Dragojevic, 2011). However, when a persuasive campaign is triggering receivers to come up with arguments themselves, the negative face of receivers will remain unthreatened, and maybe even reinforced since it redefines the relation between source and receiver, and puts the receiver in an expert role. Since this is a very indirect and less forceful way of persuading people, it is expected that reactance will be lower. This will lead to more persuasion and to more behavioral change. This study examines whether a self-persuasion campaign is more effective because it is indirect and motivates women to come up with reasons to keep a condom with themselves. Expected is that women will be less likely to counter such a message since they motivate and come up with arguments themselves. Furthermore, because of this same reason, they are expected to have more positive than negative feelings and thoughts regarding the poster. They will come to belief that they actually want to change their behavior (rather than feeling forced by the organization behind the poster). This leads to the following hypothesis:
H2: The effects of safe sex campaigns using self-persuasion are mediated by a decline in reactance towards the message.
Self-persuasion and self-esteem
Self-persuasion is a persuasive strategy that motivates people to come up with reasons why they should perform a certain behavior. This leads them to persuade themselves that performing the communicated behavior is the right thing to do. Thus, in the case of this study: young women will be motivated to create the belief that carrying condoms is right. Since people with high self-esteem are more likely to act upon their beliefs and assumptions, this kind of strategy is likely to be more effective for people with high self-esteem compared to people with low self-esteem (Baumeister, 1998). People with low self-esteem are thus expected to be less persuaded by their own arguments in a process of self-persuasion.
Therefore, hypothesized is that people with low self-esteem are more persuaded by arguments from (knowledgeable) others than from their own arguments (direct persuasion rather than persuasion). Petty et al. (2002) also provided evidence for the moderating effects of self-esteem. They found that when one is confident that his or her attitude-relevant thoughts are correct, those thoughts have a strong impact on their attitude. However, when someone is not confident that his or her attitude-relevant thoughts are correct, those thoughts do not have much effect on attitude. Therefore, when women with high self-esteem create positive thoughts regarding carrying a condom (by motivating them to think of arguments why
9
carrying a condom is favorable), this should enhance their attitudes towards carrying a condom. This leads to the following hypothesis:
H3: The effects of safe sex campaigns using persuasion are moderated by women’s self-esteem levels.
Methods
Sample and design
For the purpose of this study (motivating women to start carrying condoms with them when they go out) a sample of heterosexual female respondents without a relationship or a
relationship of less than six months was used. A total of 505 female respondents started this study. Out of these respondents, 253 respondents were dropped from the study because they had been in a relationship for 6 months or longer; 123 respondents were dropped because they did not complete the survey; 1 respondent was dropped because the stimulus (poster) did not download on her computer and 1 respondent was dropped because she was homosexual and would therefore not be interested in condoms because of her sexual orientation. In the final sample of 127 respondents, the mean age of participants was 25.5 (SD = 7.5). Furthermore, 60 % of the respondents indicated to have a University degree, 24% a Dutch HBO diploma, 11% a Dutch MBO diploma, and 5% a high school diploma. To test the hypotheses, an experiment was conducted with a two factor (type of persuasion and self-esteem) between-subjects design. Four levels of persuasion were studied: persuasion with active processing, self-persuasion with passive processing, direct self-persuasion, and a no campaign/control group. Self-esteem was measured and divided with a median split in two categories: low and high self-esteem.
Procedure
Participants were recruited through Facebook. A message was posted in which females were asked to participate. In this message a link to the survey was included. Participants were told that the survey consisted of two parts. First, they had to indicate which items they would bring on a weekend trip, a day at school and a night out in order to measure their current intentions to carry condoms in a subtle way. In the second part of the survey participants were randomly assigned to one of the four conditions. They were either exposed to a poster using
self-persuasion (active and passive processing self-self-persuasion group), a poster using direct persuasion (direct persuasion group) or no poster (control group). After that, respondents answered questions regarding these posters and the act of carrying condoms. Finally, all participants were asked to fill in their e-mail addresses and to indicate whether they were
10
willing to participate in a future study and/or whether they wanted to get informed about the purpose of the study.
Independent variables
Poster/stimuli. Participants were exposed to either a campaign poster using self-persuasion as a strategy (appendix figure 3), a campaign poster using direct persuasion (appendix figure 4) or no poster. The self-persuasion poster and direct persuasion poster both consisted of a picture of a woman talking to a man in a bar. They are both raising a glass of wine. Furthermore, both posters consist of a logo of a fictitious health organization called SOA Stichting, and a #TakeCareAndCarryOne referring to a non-existing Twitter page. The self-persuasion poster poses a question at the top of the poster: ‘Why should women carry a condom?’ and additional motivation to think about the message and respond to it: ‘Let us know what YOU think!’ in the right corner of the poster. Two groups of participants (active and passive processing self-persuasion) were assigned to this poster. However, the difference between two groups was that the active processing group received the assignment to respond to the self-persuasion poster and the passive processing self-persuasion group did not. Participants in the active processing self-persuasion condition were asked to reflect on the poster. They were told that the poster of the SOA Stichting invited them to think about why women should carry a condom with them when they go out. They were asked to imagine that they wanted to respond to this poster and send a message to the SOA Stichting, answering this question. Two options were offered: posting their response on social media with the
#TakeCareAndCarryOne, or send a private message to the SOA Stichting. After they chose between these two options, they were asked to write down their message. Participants in the passive self-persuasion condition did not have to write down their response. The reason why both an active and passive condition were used, is that in the case of a normal mass media poster campaign there is no active processing involved (receivers are not asked to write something down, but just motivated to think about the message). The purpose of measuring both conditions was to investigate whether just motivating women to think about carrying a condom (passive self-persuasion) without asking them to write something down would have enough effect.
Different than the self-persuasion poster, the direct persuasion poster consisted of the direct message that women should carry a condom with them and provided a reason why women should do this (thus women were not motivated to think about this themselves). In order to make the difference between the self-persuasion and direct persuasion as small as
11
possible (and the quality of the argumentation in particular) a pre-test was conducted in which women were asked to list reasons why they thought it would be good if women would carry condoms with them when they go out. Out of the 53 participants, 15 women suggested that you do not know what will happen on a night out and it would thus be good to carry a condom in order to be prepared. Since this was the reason that was listed most often, the direct
persuasion poster used in this study states: ‘You never know how your night will end’ at the top and the direct message ‘Always carry a condom with you’ in the right corner. The difference between the two posters is the following: the self-persuasion poster tends to motivate women to think about arguments why to take a condom with them and does not force women to perform the desired behavior (carrying a condom). The direct persuasion poster makes a clear statement ‘Always carry a condom with you’ and gives a reason as well ‘You never know how your night will end’. As already described above, this reason was formulated with the results of the pre-test. This poster does not motivate women to think about carrying a condom themselves. Furthermore, this poster does force women to carry one. In short, in the case of the self-persuasion poster, receivers become the persuaders, whereas in the case of the direct persuasion poster the SOA Stichting (the organization behind the poster) is the persuader.
Self-esteem (moderator). To measure this factor, Rosenberg’s RSES scale (1965) was used. Examples of items of this scale are ‘I feel I have a number of good qualities’, ‘I am able to do things as well as most other people’ and ‘I feel I do not have much to be proud of’.
Participants were asked to indicate their answers on a scale ranging from 1-5.
Measures
3.4.1 Descriptive variables
Demographics. Age was measured with the open-ended question ‘What is your age?’. Educational level was measured with the following open-ended item: ‘What is the highest degree of school you have completed? If currently enrolled, please indicate the level or degree you are on now’. Relationship status was measured with the following closed-ended item: ‘Are you currently in a relationship? If yes, for how long have you been in this relationship?’. Participants could choose between three answer options: ‘No’, ‘Yes, I have been in a
relationship for less than six months’ or ‘Yes, I have been in a relationship for more than six months’.
3.4.2 Dependent variables
12
Participants read a cover story telling that the study they were participating in was about checklists of items to bring on a night out, on a weekend trip, and to school. For each situation a list of items were given, and respondents were asked to indicate which of those items they would bring. The cover story was used in order to measure condom possession in a subtle way. This way the participants would not already understand that the study was about condom carriage. Participants were asked to complete these checklist items both before exposure to the poster and two weeks after exposure to the poster in a post-test.
1. ‘Imagine you are going on a weekend trip to a European city. You will stay at a hotel. Which of the items below will you bring?’
2. ‘Imagine you and your friends are going to some bars or a club tonight. What will you put in your purse?’
3. ‘Imagine you are going to the university for a full day (two lectures, a lunch and some group studying). Which of the items below will be in your handbag?’
Participants were able to choose a lot of items from the lists. One of these items was ‘condoms’, which was the only item that was actually interesting for this study.
Actual behavior – condom possession. Participants were asked whether they kept condoms at home at the moment. They could answer with either ‘yes’ or ‘no’. This question was asked two weeks after exposure to the poster. This question was asked at the end of the post-test on a separate page. This was done in order to influence the items they would pick from the checklists as less as possible.
Attitude towards the ad/poster. Participants were asked to rate the poster on a seven-point semantic differential scale composed of the following bipolar adjectives.
1. ‘Bad/good’ 2. ‘Uninteresting/interesting’3. Not-enjoyable/enjoyable’ (e.g. Moorman, Neijens, & Smit, 2002) (Cronbach’s alpha = .84).
Attitude towards carrying a condom. Participants indicated their attitude towards carrying a condom with them on a series of 7-point semantic differentials: bad/good,
unfavorable/favorable, negative/positive, unwise/wise, unnecessary/necessary, difficult/easy. A factor analysis was performed to establish the construct validity of this measure. The
outcome showed a clean factor structure for the items of the first scale for attitude towards carrying a condom. All items loaded on one factor which was labeled ‘’Attitude towards carrying a condom – personal’’ (6 items, alpha = .85) and thus, all items were kept. Attitude towards the SOA Stichting. Participants indicated their attitude towards the SOA Stichting (presented source of campaign poster) on a series of 7-point semantic differentials reflecting expertise (incompetent, not credible, knowledgeable and uninteresting) and
13
likeability (good and sincere). This measure was inspired by see Zuwerink and Devine’s Source Evaluation Scale (1996) (cronbach’s alpha = .84). A factor analysis was performed for this adjusted ‘attitude towards the SOA Stichting’ scale. The items of this scale all loaded on one factor which was labeled ‘Attitude towards SOA Stichting’ (6 items, alpha = .78). Intentions. To measure this variable, participants were asked to indicate to what extent they agreed to the following statements on a scale from 1 (Completely disagree) to 5 (Completely agree).1. Next time I go out, I will take a condom with me.2. I am going to purchase condoms to keep at home. A correlation test was performed for this ‘Intentions to carry a condom’ scale. The results showed a significant correlation. (r = 0.418 , N = 127 , p = .000) Interest in condoms. Interest in getting condoms was measured with the following item. Participants were told a price would be raffled among them and participants got to choose what price they would be most interested in. ‘To thank my participants, I will raffle a price among them. Please pick the price you would prefer among the options below.’ The options were a Bol.com voucher (value €20, -), a box of condoms (value €20,-) or a theater voucher (value €20,-). Answers were coded with either ‘0’ (no interest in condoms, when participants chose either the Bol.com voucher or a theater voucher) or ’1’ (interest in condoms, when participants picked the box of condoms).
Concerns regarding image. Participants were asked to indicate to what extent they agreed to the following statements on a scale from 1-7: ‘Carrying a condom with me when I go out is embarrassing’ and ‘Carrying a condom with me when I go out gives people a wrong
impression of me’. A correlation test for the ‘Concerns regarding image’ scale showed a significant strong correlation between the two items of this scale (r = .725, N = 127, p = .000).
3.4.3 Mediating variable - Reactance Thought-listing
Participants were asked to take about 3 minutes to write down any thoughts they had while they saw and read the poster (Petty & Cacioppo, 1986). This reactance measure was coded as follows: thoughts were categorized as ‘negative thoughts’, ‘positive thoughts’ and ‘neutral thoughts’. All types of negative and positive thoughts were counted (both thoughts related to the message of the poster and the design of the poster). The reasoning behind this is that all negative thoughts can be interpreted as reactance. When someone does not like the message of the poster, he or she can criticize all aspects of the poster, even though some of those thoughts may not be related to the message particularly. Negative thoughts were subtracted
14
from positive thoughts and this total was used. The lower the score, the higher the reactance. Scores range from -5 (big reactance) to 3 (small reactance).
Affect. Participants were asked to indicate how they were feeling when they were looking at the poster by rating a series of 10 positive and negative affect items from 1 (Very slightly to not at all) to 7 (Extremely). This measure is inspired from Jacks and Devine’s Affect scale (2000) (cronbach’s alpha = .93). The feelings they had to rate were: determined, attentive, alert, inspired, active, irritated, aggravated, upset, angry, annoyed. The positive affect elements were reversely recoded. A factor analysis was conducted for this adjusted Affect scale. The items of this scale all loaded on one factor. There were no loadings lower than .75 (10 items, alpha = .83).
3.5 Post-test
At the end of the survey participants were asked to write down their e-mail address and they were asked whether they would be willing to participate in a next study (post-test) and/or whether they wanted to be informed about the purpose of the study. The post-test’s purpose was to find out whether the behavior of the respondents had changed after three weeks. Specifically, the post-test examined whether more respondents possessed condoms at home after three weeks, and whether more of them would bring condoms on a night out. Thus, the post-test consisted of two items:
Actual behavior – condom carriage. Participants completed the checklists of items they would bring on a night out, a day at school and a holiday trip again. This item was almost the same as the item in the experiment. However, in the experiment women were able to add some items that were not on the list. These items were added to the lists now, in order to make them think the lists were improved for this next study.
Actual behavior – condom possession. Participants were asked whether they currently kept condoms at home (Yes/no).
Unfortunately only 23 respondents participated in the post-test. Therefore, the results of this test do not have enough power to be able to make conclusions.
Results
Hypothesis 1 and 3
Hypotheses 1 and 3 were investigated together. First of all, the following tests investigate whether a safe sex campaign using a self-persuasion strategy leads to 1) a more positive attitude towards carrying a condom, 2) less concerns regarding image, 3) a higher interest in condoms and 4) more intentions to carry a condom. Second, they investigates whether
self-15
esteem of women moderate these effects. For self-esteem a median split was performed (low vs. high self-esteem). Two one-way ANOVA’s, a factorial logistic regression and a
MANOVA were performed in order to test the two hypotheses.
A first ANOVA showed that intentions to carry a condom did not differ across the 4 conditions, F(3, 126) = 1.19, p = .318. Furthermore, intentions did not differ across the different self-esteem levels (low or high), F(1, 126) = .41, p = .532. However, a significant interaction effect between persuasion type and self-esteem, F(3, 126) = 2.85, p = .040. Thus, there is a difference for the effects of the persuasion type depending on the level of self-esteem. A plot showed that women with low self-esteem had the highest intentions in the passive self-persuasion condition and the lowest intentions in the direct persuasion group. Women with high self-esteem had the highest intentions in the direct persuasion condition and the lowest in the control group (see appendix figure 5). First of all, against expectations, women with lower self-esteem in the control group were found to have more intentions to carry a condom (M= 3.08, SD = .230) than women with higher self-esteem (M = 2.29, SD = .268) , p = .026 in the control group. Second, women with high self-esteem in the direct persuasion group had marginally significantly more intentions (M= 3.31, SD = .278) than with low self-esteem in the direct persuasion group (M= 2.68, SD = .230), p= .086. Third, women with high self-esteem in the direct persuasion group had significantly higher
intentions (M= 3.31, SD = .278) than women with high self-esteem in the control group (M= 2.29, SD = .268), p= .009. Fourth, women with lower esteem in the passive
self-persuasion group had marginally significantly more intentions to carry a condom (M= 2.32, SD = .243) than women in the direct persuasion group (M= 2.68, SD = .230), p= .058. Fifth, women with high self-esteem in the passive self-persuasion group had marginally
significantly more intentions (M= 2.96, SD = .289) than women with high self-esteem in the control group (M= 2.29, SD = .268) p = .090. Sixth, women with high self-esteem in the active self-persuasion group had marginally significantly higher intentions (M= 2.86, SD = .213) than women with high self-esteem in the control group. (M= 2.29, SD = .268), p = .094. A factorial logistic regression was performed to find out whether persuasion type and self-esteem levels have effects on interest in condoms. The analysis showed that the overall model was not significant t(X²= 3.090, p = 0.543). Furthermore, none of the coefficients (neither of the individual persuasion types or self-esteem) were significant either.
A MANOVA with attitude and image concerns as dependent variables was performed. The results showed that there was no significant interaction effect for self-esteem and condition, F(6, 234) = .852, p = .531; Wilk’s Λ = .958. Furthermore, two separate ANOVA’s were
16
performed to examine attitude and image concerns separately. A first ANOVA showed that only a marginal significant effect was found for persuasion type on attitude F(3, 126) = 2.19, p = .093. Attitude did not differ across the different self-esteem levels F(1, 126) = .46, p = .497. Furthermore, no significant interaction effect was found between persuasion type and self-esteem, F(3, 126) = 1.07, p = .367. A post hoc test for the effects of persuasion type on attitude showed that attitude was significantly more positive among women in the passive self-persuasion group (M= 5.52, SD= .276) compared to the direct persuasion group (M= 4.87, SD = .263), p = .050. Furthermore, attitude was significantly more positive among women in the passive self-persuasion group (M= 5.52, SD= .276) compared to control group (M= 4.68, SD = .258), p = .030. The second ANOVA showed that image concerns
significantly differs between the 4 conditions, F(3, 126) = 3.17, p = .027. However, no
significant effects were found for the effect of self-esteem on image concerns F(1, 126) = .02, p = .885. A significant interaction was not found for image concerns either F(3, 126) = .945, p = .421. A post hoc test for the effect of the persuasion type on image concerns showed the following: image concerns were significantly less among women in the direct persuasion group (M= 2.57 , SD= .317) compared to women in the active self-persuasion group
(M=3.66, SD= .336), p= .025. Image concerns were also significantly less among the direct persuasion group (M= 2.57 , SD= .317) compared to the control group (M= 3.85, SD= .310), p= .009.
To conclude, H1 was not supported. Self-persuasion only had more positive effects on attitude towards carrying a condom. H3 was not supported either. Self-esteem only affected intentions depending on the persuasion type. However, these effects worked in the opposite direction of what was expected. Women with low self-esteem had more intentions to carry a condom when exposed to a (passive) self-persuasion poster instead of a direct persuasion poster. Women with high self-esteem had higher intentions when exposed to the direct persuasion poster compared to the self-persuasion poster.
Hypothesis 2
To investigate whether a decline in reactance mediates the effects of a safe sex campaign using self-persuasion a MANOVA was performed in order to examine whether there are differences between the conditions in levels of reactance (measured with thoughts and affect). The results show that there are no significant differences across the conditions in levels of (positive or negative) thoughts F(3, 126) = .572, p = .634, and affect F(3, 126) = 1.44, p =
17
.234. Therefore, a mediation effect also does not have to be examined further, and H2 was rejected.
Conclusion and discussion
This study had three purposes. The first purpose was to gain knowledge of effects of safe sex campaigns for young women using a self-persuasion strategy compared to campaigns using a direct persuasion strategy. The second purpose was to examine a potential reduction in reactance towards the campaign as an underlying process for the effects of a campaign using self-persuasion. The third purpose was to investigate whether self-esteem of women
moderates the effects of the different strategy types.
First of all, self-persuasion was found to have more positive effects on attitudes towards carrying a condom than a campaign using direct persuasion. This is consistent with previous studies on the positive effects of self-persuasion on attitude. The self-persuasion campaign poster motivated women to think about the benefits of carrying a condom. This lead them to persuade themselves that carrying a condom is the right thing to do. However, women were found to suffer less from image concerns after exposure to the direct persuasion campaign compared to the self-persuasion group. A possible explanation for this result is that the direct persuasion poster gives women a reason to carry a condom when they go out (‘Because you never know what happens’). This argument forces women to focus on a benefit, and it may distract them from thinking about image concerns. This is important, since it was found that women worry carrying condom will make them look sexually promiscuous. These image concerns keeps them from carrying them, even though their attitudes towards condoms are very positive. Considering the fact that women already hold positive attitudes regarding condoms, it may be more efficient to focus on these image concerns rather than their attitude. In that case, the results of this study suggest to use direct persuasion in a campaign. However, self-persuasion may be more effective when the answer to a question asked by the poster is more clear. Some participants commented on the self-persuasion poster that they did not understand its message and purpose: ‘The poster poses a question, but does not answer it’. Women may have found the self-persuasion posted used for this study unclear. Indeed, a study conducted by Ball and Goodboy (2014) on student evaluations of assignments showed that students were more willing to complete an assignment if the teacher was forceful but and clear about it rather than indirect and vague. It would maybe be better if women would be pushed more towards the right answer, especially since carrying a condom is considered an embarrassing behavior. Clarity of the self-persuasion message could be taken as a moderator
18
in a future study. Furthermore, future research could investigate what happens when self-persuasion and direct self-persuasion are combined. What could for example be examined is giving receivers a direct persuasion message, but also ask them ‘What do YOU think?’. Persuasion type did not have any effect on interest in condoms. It did have some effects on intentions to carry condoms. However, this depended on the level of self-esteem of women. However, the fact that not all expected positive effects of self-persuasion were found in this study may have been caused by other factors. First of all, some respondents found the poster look slightly unprofessional. Second, some women interpreted the poster in an unexpected way. They stated that they believed that men and women both have responsibility for condoms instead of just women. They thought the poster was communicating that women should take the responsibility for condoms instead of men. Third, some women found the message odd. They believed that the poster motivated women to behave in a sexual
promiscuous way. Fourth, some respondents evaluated the message very positively. However, they believed that the message was meant for others, not for them. They believed the message did not apply to them.
Second, women with low esteem had the highest intentions in the (passive) self-persuasion condition and the lowest intentions in the direct self-persuasion group. It makes sense that intentions did differ across the different self-esteem levels and attitudes (attitude and image concerns) did not, because the first two concern behavior (actually getting and carrying condoms) which is something you have to feel confident about, and attitude not. It is possible to have a positive attitude regarding condoms. However, you may not feel confident to carry them. Women with high self-esteem had the highest intentions in the direct persuasion condition and the lowest in the control group. This was against expectations: the
self-persuasion poster motivated women to persuade themselves with their own arguments. It was expected that women with a high self-esteem would have more confidence in their own arguments and would thus be more persuaded by such a self-persuasion poster than women with low self-esteem. An explanation could be that women with low self-esteem do not like to be forced into a type of behavior they consider embarrassing and difficult to perform whereas women with high self-esteem worry less about that. Women with higher self-esteem may find it less embarrassing and easier to take a condom. Therefore, they would maybe care less when they are forced by a knowledgeable organization. The self-persuasion poster leaves the
decision to the receiver, and does not force them to carry a condom. This could explain why a direct persuasion poster (that is more forceful and clear) is more effective for women with high self-esteem and a self-persuasion poster (that is less forceful and lets women decide for
19
themselves) for women with low self-esteem. In this case attitude and efficacy possibly mediate these effects on intentions to carry a condom.
Third, reactance was examined as an underlying process of the effects of self-persuasion. It was expected that – since a self-persuasion poster is less forceful and directive – people will show less reactance towards it. The results show that this is not the case. However, in this study reactance was measured with affect (negative feelings versus positive feelings) and thoughts (negative thoughts versus positive thoughts). It did not measure exactly whether they felt forced by the poster to perform the behavior. Future studies could look into this.
Furthermore, condoms are a type of behavior most women already do consider important and positive. Therefore, women probably already had less negative feelings regarding the poster because they already consider the message as something positive. A future study could test if women feel forced into the behavior, and whether that has certain negative effects.
To conclude, if we want to stimulate women to carry a condom too, campaigns should first start focusing on this fear of women to get a bad image when they carry condoms. This study confirms that women indeed consider carrying a condom as embarrassing. Future studies could focus more on what type of strategies can tackle this. New York based condom-start up brand Lovability is already working on this issue. They came up with a condom that is packaged in a more discreet way and could therefore be more comfortable for women to carry around. Future safe sex campaigns should focus on normalizing condom possession for young women, and making these women more confident about carrying one.
References
Aronson, E. (1999). The power of self-persuasion. American Psychologist, 54: 875-884. Baumeister, R. E (1998). The self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., pp. 680-740). New York: McGraw-Hill.
Bryan, A., Fisher, J. D. & Fisher, W. A. (2002) Tests of the mediational role of preparatory safer sexual behavior in the context of the theory of planned behavior, Health Psychology, 21(1), pp. 71–80.
Eagly, A., & Chaiken, S. (1993). The psychology of attitude. Attitude structure and function. Handbook of Social Psychology, 2 (4th ed.), edited by D. Gilbert, S. Fiske, and G. Lindzey, 269–322.
Eagly, A., & Kulesa, P. (1997). Attitudes, attitude structure, and resistance to change: Implications for persuasion on environmental issues. Environment, ethics, and behavior: The psychology of environmental valuation and degradation, edited by M. Bazerman.
20
Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with persuasion attempts. Journal of Consumer Research, 21(1), 1-31.
Jellema, I.J., Abraham, C., Schaalma, H.P., Gebhardt, W.A and van Empelen, P. (2013) 'Predicting having condoms available among adolescents: The role of personal norm and enjoyement'. British Journal of Health Psychology 18 (2), 453-468
Jenkins, M., & Dragojevic, M. (2011). Explaining the process of resistance to persuasion: A politeness theory-based approach. Communication Research, 40(4). 559-590 .
Petty, R. E., Brinol, P., & Tormala, Z. L. (2002). Thought confidence as a determinant of persuasion: The self-validation hypothesis. Journal of Personality and Social Psychology, 82, 722–741.
Williams, P., Fitzsimons, G.J., & Block, L. G. (2004). When consumers don’t recognize “benign” intentions questions as persuasion attempts. Journal of Consumer Research, 21, 540–550.
Zuwerink, J.R., Devine, P.G. (1996). Attitude importance and resistance to persuasion: It’s not just the thought that counts. J. Pers. Soc. Psychology. 70:931–44
Jacks, Z.J., Devine, P. (2000). Attitude importance, forewarning of message content, and resistance to persuasion. Basic Appl. Soc. Psychology. 22:19–29
Appendix
21
Figure 2: breast cancer screening campaign Baycare Health Systems
22 Figure 4: direct persuasion campaign poster
Figure 5: interaction effects persuasion type and condition on intentions 0 0,5 1 1,5 2 2,5 3 3,5
Low self-esteem High self-esteem
