• No results found

Persuasion and framing

N/A
N/A
Protected

Academic year: 2021

Share "Persuasion and framing"

Copied!
36
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Persuasion and framing:

The effect of self-persuasion and framing on reactance towards a message

Sietske Abelmann Student ID: 10440720

Master’s Thesis

Graduate School of Communication Master’s program Communication Science

Supervisor: Dr. M.L. (Marieke) Fransen Date of completion: June 30, 2017

(2)

Abstract

This research tries to give insight in how a message to promote ecstasy testing can best be designed. In the Netherlands ecstasy is more commonly used and the number of people using it is rising. Unfortunately, the use of ecstasy is not always without health risks and can be dangerous since the pills might be poisonous. By testing the drugs this can be detected and negative consequences can be prevented. However, at this moment only a small amount of the ecstasy used is being tested. It has been proposed that the inducing of self-persuasion in a message might lead to less reactance and increase the intention to test ecstasy. The way this message is framed – positive or negative – might influence that effect, though this has not been researched before. This research therefore tries to give answer to the question: ‘What is the effect of self-persuasion on reactance and intention to test ecstasy and what is the role of framing in this?’. To answer this question an experiment with a 2 (direct vs. self-persuasion) x 2 (gain- vs. loss-frame) between-subjects design is conducted. The results showed a marginal significant effect of type of persuasion, in which a direct persuasion message led to a higher intention to test ecstasy. It is also shown that type of persuasion has an influence on reactance, but only on certain aspects. A direct effect of type of persuasion on threat to freedom and negative affect is found, but in the opposite direction than was expected. In these cases self-persuasion led to more reactance. This research is not in line with previous research, but may provide an interesting basis for future research.

Keywords: persuasion, framing, reactance, negative affect, threat of freedom, source derogation.

(3)

Introduction

After the Amsterdam Dance Event in 2014, the media reported three deaths in which drugs were involved. One of these deaths was assured to be caused by the use of ecstasy (Klompenhouwer, 2014). Ever since the 1990s, the drug ecstasy has entered the drugs world (Wouters, Nabben, Benschop, & Korf, 2014). According to the Global Drug Survey, Dutch people use ecstacy more often than other nationalities. Worldwide, 19% of people had used ecstasy in the last year, whereas this was 48.9% among Dutch respondents (Unity, 2017-a). The drug is mostly used among young adults and the number of people using it is rising (Trimbos, 2016-a). Ecstasy has physical and psychological effects. Usage evokes a sense of social and open behavior to others. It creates feelings of euphoria and joy, and makes people feel cuddly. It intensifies the music, lights and surroundings, and people get more energy (Bogt, Engels, Hibbel, van Wel, & Verhagen, 2002). The physical effect, among other things, can increase one’s heart rate, blood pressure, body temperature and muscle strain (Niesink, Brunt, & Croes, 2015).

Unfortunately, the use of ecstasy is not always without health risks. The amount of ecstasy related incidents has raised to 51% of the total number of incidents reported at first aid posts, at festivals in 2015 (Trimbos, 2016-b). It turns out that not the ecstasy usage itself is the problem that can cause death, but death is caused by overdoses, unsafe usage and poisonous pills (Jellinek, 2016). In the Netherlands, it is possible to test drugs. Contributors at testing facilities are available to inform users about the composition of the drugs and provide information on save usage. The tests can show whether a pill is poisonous and when a poisonous pill is discovered, a nation-wide or regional warning campaign (Red Alert) is distributed to warn drug users about this pill (Trimbos, 2017). The news broadcast and festivals distribute information about what the pill looks like and why it is poisonous. At the moment, only a fraction of the used drugs is being tested: only 11.5% of people that use ecstasy had it tested before usage (Unity, 2017a). To increase the safety of using ecstasy and decrease the incident rates, it should become more common to have pills tested before usage. In 2015, the campaign ‘XTCfacts.nl’ was launched, in which people are directed to a site with the most important facts about ecstasy. As with most health campaigns, this campaign tries to show the risks that are related to behavioral choices and should be seen as an intervention to promote healthy behavior (Krischler & Glock, 2015). Health campaigns usually use a message that provides arguments and reasoning to stimulate receivers to behave in a certain way (Hamilton, Biener, & Brennan, 2008). This is a form of direct persuasion.

(4)

With direct persuasion, it is clear to the receiver that the intent of the message is to persuade the receiver into certain behavior (Munger, 2011).

The use of direct persuasion has the disadvantage that it can give the receiver the feeling that his or her freedom is being imposed. According to the reactance theory (Brehm, 1966), this can cause reactance, which can be perceived as resistance against this limitation of freedom and can cause arousal. When people experience reactance, they are motivated to restore their freedom of choice. They can do this by, for example, disregarding the message or behaving in the opposite direction (Miron & Brehm, 2006). Young adults in particular often experience reactance (Burgoon, Alvara, Grandpre, & Voulodakis, 2002), therefore - in the case of drugs testing - it is very important to prevent reactance from occurring.

Something that can counteract the appearance of reactance and therefore might lead to less resistance to the message, is approaching the receiver in an indirect way (Fransen, Verlegh, Kirmani, & Smit, 2015). A way to do this, is stimulating self-persuasion through a message. In this way, the receiver of the message is stimulated to come up with own arguments in accordance with or in contrast to a certain behavior. In this way, the receiver is persuading him- or herself (Baldwin, Rothman, Vander Weg, & Christensen, 2013). According to different studies, this could even be the most effective way to persuade someone, since the motivation to change is not coming from outside, but from the individual itself (Aaronson, 1999; Baldwin et al., 2013). The amount of research about self-persuasion is limited in comparison to the amount of research about direct persuasion. It has been proven that persuasion can be very effective. However, the circumstances in which self-persuasion works best, remains a question (Damen, Müller, Baaren, & Dijksterhuis, 2015). When someone provides their own arguments for a behavior, these arguments can focus on the positive or the negative aspects of the behavior in question. They can emphasize the benefits of taking action - then the argument is in a so-called gain-frame - or they can emphasize the costs of performing a behavior - then the argument is in a so-called loss-frame. In health messages, framing is a commonly used technique with direct persuasion messages, and research shows that a particular frame (positive or negative) works better for specific behaviors (Rothman, Bartels, Wlaschin, & Salovey, 2015). This might be provoked in self-persuasion messages to steer people in a certain direction. The frame that is used in a message can influence the reaction of the receiver (Rothman et al., 2015). For example, a gain-framed argument for testing ecstasy can be that testing ecstasy will make usage more enjoyable and safe, since the composition is known, and thus lead to more positive feelings. In contradiction,

(5)

a loss-framed argument can be that not testing ecstasy may cause danger, since the pill might be poisonous and might lead to feelings of anxiety.

The combination of self-persuasion and framing this in different ways, has not been studied before. Since the effects of framing tend to be different in several situations when direct persuasion is used (Van ‘t Riet et al., 2016), it is very important to examine the effect when self-persuasion is used. This knowledge will contribute to how health messages can best be designed and communicated to people, which is very relevant, since this can help prevent unhealthy behavior and increase the overall health of civilians. The aim of this study will therefore be to give insight in the effects of the type of persuasion used, and the way the message is framed. It will provide an answer to the question:

‘What is the effect of self-persuasion on reactance and intention to test ecstasy and what is the role of framing in this?’

Theoretical Framework

Persuasion and intention

Persuasion is the act of influencing someone into doing something (Cobuild, 2006). This can be done in different ways. Two ways, that are seen as opposites of each other, are direct persuasion and self-persuasion (Baldwin et al., 2013).

Messages using direct persuasion, use arguments that are provided by the source in the message, to provoke certain attitudes or behaviors. In this kind of message, the goal is to directly persuade the receiver (Cialdini, 2014). In advertising this is commonly used; the creator of the advertisement emphasizes specific aspects of a product or service by using arguments about these aspects, aiming to change the convictions of the receiver (Yi, 1990). The stronger and more qualitative these arguments are - and the more reliable the source of the message - the more perceptive the receiver is for persuasion (Pornpitakpan & Francis, 2001). Direct persuasion is mostly used to change the convictions of the receiver. Changing these convictions can positively influence the attitude towards the product and the intention to buy the product (Yi, 1990). This is also commonly used within health messages, since these provide arguments and reasons to stimulate receivers to change their unhealthy behavior (Hamilton, et al., 2008).

Another effective way to persuade receivers is by using self-persuasion (Krischler & Glock, 2015; Müller et al., 2009). By inducing self-persuasion, people are encouraged to generate their own arguments about a specific behavior and consequently influence themselves (Baldwin et al., 2013). This can be stimulated by asking a question, like ‘why

(6)

would you test your ecstasy?’ instead of using a statement or argument, like ‘you should test your ecstasy’. When a person is triggered by reading a question, the receiver is likely to come up with arguments. Providing arguments is a process that is assumed to be automatic (Fitzsimons & Williams, 2000). Recent research has shown that writing down these self-generated arguments is not necessary to induce self-persuasion and that exposure to a question is sufficient (Bernritter, van Ooijen, & Müller, 2017).

There are two components that may explain why these self-generated arguments can influence attitudes and behavior (Baldwin et al., 2013). First, people are their own source of arguments. People do not correct self-generated information, whereas people do with external generated information (Mussweiler & Neumann, 2000). Besides, it is known that the characteristics of the source are of influence on the persuasiveness of arguments. Arguments are more convincing when they come from a reliable source, than from an unreliable source (Baldwin et al., 2013). People tend to assume that their arguments, since they come from themselves, are true, and therefore will be persuaded by it (Kahneman, Knetsch, & Thaler, 1991). Second, when coming up with their own arguments, the ones that fit their convictions are likely to be most prevalent (Baldwin et al., 2013). People also tend to come up with arguments they find most convincing (Briñol, McCaslin, & Petty, 2012). People will therefore come up with different arguments when they are asked, for example, why they would test their ecstasy. These arguments suit the individual and will be more convincing to them (Noar, Benac, & Harris, 2007). This effect is especially of great importance in health messages, since these are personally relevant for the receiver, which enlarges the effect that self-relevant thoughts have a stronger effect on one’s attitude (Kreuter, Bull, Clark, & Oswald, 1999). One’s attitude towards a subject is seen as one of the primary conditions for an intention to change behavior, which most likely influences actual behavioral change according to the theory of planned behavior (Ajzen, 1991). In general, the more favorable the attitude, the stronger one’s intention is to perform a certain behavior. Therefore, while self-generated arguments can influence attitudes, it is likely that they influence their intention as well. People are able and likely to resist others’ persuasive attempts, but they fail to resist persuasive attempts by themselves (Knowles & Linn, 2004). One explanation for this might be that people want to be consistent in their thoughts and behavior, otherwise cognitive dissonance might occur. The cognitive dissonance theory of Festinger (1957) states that someone becomes agitated when two or more cognitions contradict each other. Next, one can respond by rejecting a cognition, to include new information or to reduce the importance (Burns, 2006). This happens mostly with external cognitions and is thus less likely to occur

(7)

with self-persuasion. This might therefore have a greater impact on the intention of behaving in a certain way than direct persuasion. This argumentation leads to the following hypothesis: H1: People who are exposed to a message that induces self-persuasion will have a higher intention to test ecstasy than people who are exposed to a message that uses direct persuasion

Persuasion, reactance and intention

Psychological reactance is more commonly labeled as potential explanation for resistance, because of the failure of different media campaigns to promote healthy behavior (Burgoon et al., 2002). The reactance theory of Brehm (1966) suggests that a message that tries to change the current attitude and behavior of the receiver, might be perceived as a threat to the receivers’ freedom of choice. This is independent of whether the sender has the receivers’ best interest in mind, which is the case within the promotion of healthy behavior - like for instance testing ecstasy. The theory states that when people perceive their freedom being threatened, psychological reactance may occur. This is a feeling of arousal that can be seen as a motivational state stimulating the desire to reestablish the free behavior that has been eliminated, or has been threatened. People do not like to be told what to do and tend to have a high need for autonomy as they like to think that they are in control of their own fate. This motivates them to restore their freedom of choice (Burgoon et al., 2002).

Persuasive communications poses a potential threat to people’s freedom. This is especially the case for direct persuasion, since this explicitly aims to change the convictions of the receiver (Cialdini, 2014). It is than more likely that reactance will occur, which can lead to ignorance of the message, counter-arguing, derogating of the source and even conducting the undesired behavior as a demonstration of choice or restoring attitudinal freedom (Burgoon et al., 2002). For example, when a message tells you to stop drinking alcohol because research shows it is dangerous, you might go and drink even more alcohol, start to think negatively about the company that performed the research, or think of arguments pro drinking alcohol.

Reactance can thus be labeled as a motivational state that mediates the effect of threat to freedom on outcomes such as attitude and behavior (Dillard & Shen, 2005). Research also shows that less counter-arguing - which is one of the commonly used ways to measure reactance - means less reactance is experienced (Jacks & Cameron, 2003). This leads to a more favorable attitude towards a behavior and, in turn, leads to a higher intention to behave that way (Asbeek Brusse, Fransen, & Smit, 2017). Reactance thus negatively influences intention.

(8)

According to Burgoon and colleagues (2002), only certain types of persuasive messages arouse reactance. The more obvious the persuasive intent of a message, the higher the chance that reactance will occur. It can than be stated that when the persuasive intent is less explicit, receivers are more receptive to influence, since reactance is less likely to occur. Messages that use self-persuasion are less explicit in their persuasive intent, since people have to think about it themselves. People are also unable to correct the influence of self-generated information (Mussweiler & Neumann, 2000). People can, however, correct external generated information (Petty & Cacioppo, 1979), therefore direct persuasion can lead to a more defensive reaction to justify one’s current behavior. It could than be that self-persuasion leads to less reactance than direct persuasion, and subsequently result in higher intentions to behave in a specific way. This argumentation leads to the following hypothesis:

H2: The effect of self-persuasion on intention (H1) is mediated by reactance.

Message framing

Besides the use of direct- or self-persuasion, a message can also use different presentations of the same informational content. This manner of presenting information is called message framing (Morton, Rabinovich, Marshall, & Bretschneider, 2011). This might be done in a more positive way, by using a gain-frame, or in a more negative way, by using a loss-frame. A message that uses a gain-frame will refer to the positive consequences or the prevention of bad consequences when a certain, desired behavior is conducted. It emphasizes the benefits of taking action, like testing drugs will make the usage safer and more fun. A message that uses a loss-frame will use an opposite approach: it will refer to negative consequences or the withholding of good consequences when the desired behavior is not conducted. It emphasizes the costs when failing to take action, like not testing drugs will increase the chance of dangerous consequences (Rothman et al., 2013).

Message framing has roots in prospect theory, which proposes that individuals will respond differently when they are presented with different decision scenarios. This depends on whether the consequences of this scenario are framed in potential losses or gains (Tversky & Kahneman, 1981). When people are presented with loss-framed information, people are more willing to take risks, whereas when people are presented with gain-framed information, they tend to become more risk aversive and choose the safe option (van ‘t Riet et al., 2016). The application of risk perception communication is often used in the context of the message framing theory (Meyerowitz & Chaiken, 1987; Rothman & Salovey, 1997). This theory is commonly used with communication about health behavior, since health decisions

(9)

often enhance a choice between a certain immediate outcome and a more uncertain delayed outcome (De Bruijn, Spaans, Jansen, & van’t Riet, 2016). According to this theory, there are two types of behavior - namely detection and prevention behavior – which are also common within health behavior. With detection behavior, an action is taken in order to find out if someone is ill, for example taking a mammography. There is an inherent risk associated with this type of behavior, therefore loss-framed messages should work better. On the other hand, with prevention behavior, an action is taken to prevent someone getting ill, for example using sunscreen. This is considered to be risk aversive, therefore gain-framed messages should work better for the low risk that is associated with this behavior (De Bruijn et al., 2016). Testing ecstasy can be seen as prevention behavior, since this might avoid usage of a dangerous pill and consequently avoid illness.

A direct effect of message framing on reactance can be expected, since messages that use a loss-frame are more likely to evoke a defensive reaction towards the source than messages that use a gain-frame. When a message states the bad consequences of one’s behavior, like with a loss-frame, this might provoke receivers to feel like their judgment is being challenged and their competence about making a judgment is being questioned. This can feel demeaning (Jenkins & Dragojevic, 2011). These effects increase the chance people experience reactance (Burgoon et al., 2002). A loss-framed message might therefore lead to more reactance than a framed message. It has also been shown that the use of a gain-framed message leads to less counter-arguing than a loss-gain-framed message. Counter-arguing is one of the commonly used ways to measure reactance, where more counter-arguing means more experience of reactance (Jacks & Cameron, 2003). It can thus be stated that framing affects the experience of reactance. This argumentation leads to the third hypothesis:

H3: People who are exposed to a gain-framed message will experience less reactance than people who are exposed to a loss-framed message.

Message framing can - like self-persuasion - be seen as an indirect technique for changing behavior (Bernritter et al., 2017). Approaching receivers in a more indirect way, as stated before, reduces the possibility that they will show reactance to the persuasive intent (Fransen et al., 2015). However, whereas the type of persuasion includes how the message is presented to a receiver – for example a question or statement about why people should test drugs - framing is more about how the information inside this message is shaped – for example because it is dangerous versus it is safer (Morten et al., 2011). Framing might moderate the effect of self-persuasion on reactance.

(10)

Self-persuasion provides the receiver the freedom to come up with their own arguments, which makes it an automatic process (Fitzsimons & Williams, 2000). This might also have unwanted results; whereas direct persuasion can steer in a certain direction with the provided argument, the stimulation of one’s own arguments might give more freedom to not forming arguments at all, or forming arguments contradicting the intended message. When a loss-frame is used, a defensive reaction may be provoked (Jenkins & Dragojevic, 2011) and in the case of self-persuasion might thus lead to more reactance. However, people are likely to form the most prevalent arguments, of which can be assumed that this does not require a lot of effortful thinking. Previous research has shown that a gain-frame works better when less thinking is required (Moorman & Van den Putte, 2008) and leads to more positive affect, and thus to less reactance (Van ’t Riet et al., 2016). The effect of self-persuasion on reactance might consequently be even stronger when it is used within a gain-frame. This argumentation leads to the following hypothesis:

H4: The effect of self-persuasion (vs. direct) on reactance will be stronger when a gain-frame (vs. loss-frame) is used.

The fact that people want to be consistent in their thoughts and behavior might lead to a larger effect of a message on intention when self-persuasion is used. However, this might also be affected by the way the message is framed. When a self-persuasion message uses negative framing, this stimulates the receiver to come up with negative consequences of not adopting a specific behavior. This might emphasize the inconsistency between the goal of the message and the current behavior, and lead to more cognitive dissonance than with positive framing (Van ‘t Riet et al., 2016). Positive framing also leads to higher acceptance of the information and a more positive attitude towards the message (Van ’t Riet et al., 2016). Numerous studies have found that message framing can also affect people’s intentions to change their behavior (Rothman et al., 2015), and that the use of a gain-frame, compared to a loss-frame, increases the intention to execute a prevention behavior (Asbeek Brusse et al., 2011). Since the behavior of testing ecstasy can be seen as prevention behavior, a message that induces self-persuasion might lead to an even stronger intention to test ecstasy when a gain-frame is used. This argumentation leads to the following hypothesis:

H5: The effect of self-persuasion on the intention to test ecstasy will be stronger when a gain-frame (vs. loss-gain-frame) is used.

(11)

Moderated mediation

When the use of self-persuasion increases the intention to test ecstasy (H1) and this effect is mediated by reactance (H2), self-persuasion should lead to less reactance, which in its turn leads to a higher intention to test ecstasy. When the use of a gain-frame leads to less reactance (H3), the reactance people experience will be even lower for people who are shown the message that uses self-persuasion combined with a gain-frame (H4). Because this same message leads to a higher intention to test ecstasy (H5), it could be stated that the effect of self-persuasion on intention is mediated by reactance, which is moderated by framing in the case of ecstasy testing. This argumentation leads to the following hypothesis (also see Figure 1 for the conceptual model):

H6: The effect of self-persuasion on intention to test ecstasy is mediated by reactance, which is moderated by framing.

Figure 1: conceptual model

Method

Participants and design

To give answer to the research question an experiment was conducted among participants of the age between 19 and 29. This sample was chosen, since ecstasy is mostly used among young adults in the Netherlands (Trimbos, 2016-a). The amount of people that have used ecstasy increases fast after the age of 18, and decreases from the age of 30 (Centraal Bureau voor de Statistiek, 2017; Jellinek, 2017). A total of 220 respondents started this research. After examining the data for extreme and missing scores, it turned out that 56 respondents (25.5%) did not complete the questionnaire and that 8 respondents were older than the predefined age limit and were therefore deleted from the sample. The final sample exists of

(12)

156 respondents of the age between 19 and 29 (M = 23.17, SD = 1.81), of which 46.2% was male (N = 72) and 53.8% female (N = 84).

The participants were sampled by snowball sampling, and were only included when they had ever used ecstasy. Within the personal network of the researcher, participants were asked to participate and to share the research with people they knew had used ecstasy. Participants were randomly assigned to one of the four conditions of the 2 x 2 factorial design, with type of persuasion as a between-subject variable (self-persuasion vs. direct persuasion) and type of framing as a between-subject variable (gain-frame vs. loss-frame). Both these factors are experimental factors that are manipulated. Also see Table 1 for the distribution among conditions.

Table 1: Distribution of respondents among conditions

Condition Gain-frame Loss-frame Total Self-persuasion N = 41 N = 39 N = 80 Direct persuasion N = 35 N = 41 N = 76

Total N = 76 N = 80 N = 156 (100%)

Procedure

During May 2017, participants were approached via Facebook to participate in this research. They were told that they would help with a graduation research, that it would take approximately ten minutes and that it was only intended for people who had used ecstasy once or more in their lives. The research existed of a questionnaire, in which the stimulus material was included. First, people were stated their rights about participating in the research and needed to give consent to go any further. After that, some background questions were asked about their age, gender, education and whether they had ever used and tested ecstasy. The use of ecstasy was questioned to check whether only targeted people participated in this research. When someone answered not to have used ecstasy before, they were sent to the end of the questionnaire and were thanked for their participation. Participants who had used ecstasy were shown one of the four advertisements, and several questions about the dependent variable (i.e. intention) and mediator (i.e. reactance) were asked. Lastly, participants were asked what, in their opinion, the goal of the research was and were thanked for their participation.

(13)

Independent variables

The independent variable, type of persuasion, is operationalized in two levels, namely direct persuasion and persuasion. This was manipulated by the use of a question for the self-persuasion condition and a statement for the direct self-persuasion condition. The moderator, framing, is also operationalized in two levels, namely gain-frame and loss-frame. Stating the advantage or provoking the formulation of the advantage for the gain-frame condition, and stating the disadvantage or provoking the formulation of the disadvantage for the loss-frame condition, was used to manipulate framing. Therefore, four different messages were used in the stimulus material.

The stimulus material that was used in this experiment was specifically designed for the purpose of this study. Before the stimulus material was made, a pretest was conducted to confirm whether the framing manipulation worked for the self-persuasion condition and which statements should be used for the direct persuasion condition. For this purpose, 28 participants between the ages of 18 and 25 (M = 21.21, SD = 1.91) were asked to answer one of two open questions, namely - with a gain-frame - ‘what are the advantages for you of testing ecstasy before using it?’ and - with a loss-frame - ‘what are the disadvantages for you of not testing ecstasy before using it?’. The results showed that in both conditions people thought about the composition of the pill. The participants that received the question about the advantages answered more in terms of safety, whereas the participants that received the question about the disadvantages answered more in terms of danger. Hence, the framing manipulation was successful. This led to the forming of four messages. First, ‘What are the advantages for you of testing your ecstasy before using it?’ for the self-persuasion and gain-frame condition. Second, ‘What are the disadvantages for you of testing your ecstasy before using it?’ for the self-persuasion and loss-frame condition. Third, ‘Test your drugs! Testing your drugs will make it safer to use since you know what you’ll be taking in’ for the direct persuasion and gain-frame condition. Fourth, ‘Test your drugs! Not testing your drugs can be dangerous since you will not know what you’ll be taking in’ for the direct persuasion and loss-frame condition.

These sentences were used to design the advertisements as stimulus material. The advertisements use the logo of a Dutch non-governmental organization ‘Unity’, which is a volunteering project at which young adults go to parties and festivals to give information about drugs and alcohol and the risks that come with usage (Unity, 2017b). An image was chosen to demonstrate the atmosphere of a party and with a crowd on it, so participants could

(14)

identify themselves with the people in the advertisement and recognize the setting. Except for the text, the advertisements did not differ from each other (see also Appendix 1).

Mediator

The mediator in this research is reactance. Four different operationalizations for reactance are used. Firstly, the cognitive response measurement of Dillard & Shen (2005) is used. This measurement includes the question ‘write down whatever was in your mind when you finished reading the action component of the message’. It is argued that the individual responses will determine whether a message is accepted, resulting in persuasion, or rejected, resulting in resistance (Jacks & Cameron, 2010). Reactance is seen as a perceptual stance toward resistance (Knowles & Linn, 2004). This is also used as a manipulation check to see whether the different manipulations indeed lead to different reactions. The responses on the cognitive response question were coded as supportive, neutral or negative thoughts. Supportive thoughts were coded based on whether the participants had agreement with or a positive thought toward the message or source (reach = 0-4, M = 0.49, SD = 0.75). Negative thoughts were coded based on whether participants disagreed with the message or when a participant was counter-arguing or derogating the source or the message itself (reach = 0-3, M = 0.10, SD = 0.40). Last, neutral thoughts were coded when participants responded non-evaluative of the message, such as ‘nice lights’ (reach = 0-5, M = 2.16, SD = 1.64). Only the negative thoughts are a predictor for reactance (Dillard & Shen, 2005), however, these are not relevant enough to be taken into further account for analysis, since only a small amount of the participants scored on this scale (see the mean score).

Second, as threat to freedom is a cause for reactance (Brehm, 1966), this was measured by the induction check of Dillard & Shen (2005), existing of four items: ‘The message threatened my freedom of choice’, ‘The message tried to make a decision for me’, ‘The message tried to manipulate me’ and ‘The message tried to pressure me’, with a 5-point answering scale between ‘strongly disagree’ and ‘strongly agree’. These items had a Cronbach’s α of .854, therefore the items were averaged into one measure of threat to freedom (reach = 1-4, M = 2.07, SD = .94).

Third, negative affect was measured, since this might be caused by reactance (Brehm, 1966). Reactance involves responding to the persuasive attempt by getting aroused and feeling negative affect (Jacks & Cameron, 2010). This was measured with a scale that has been previously validated (Dillard & Peck, 2000; Dillard, Plotnick, Godbold, Freimuth, & Edgar, 1996). This scale exists of four items; ‘irritated’, ‘angry’, ‘annoyed’ and ‘aggravated’,

(15)

with a 5-point answering scale between ‘none of these feelings’ and ‘a great deal of these feelings’. To prevent participants from getting a negative feeling from the questionnaire, also four positive items were added to this scale. These were only used as a filler questions and to avoid that the scale would negatively affect the rest of the questionnaire. The positive items were ‘happy’, ‘cheerful’, ‘excited’ and ‘joyful’, also with a 5-point answering scale between ‘none of these feelings’ and ‘a great deal of these feelings’. The four items that measure the negative affect had a Cronbach’s α of .836. Therefore, the items were averaged into one negative affect measure (reach = 1-4.75, M = 1.76, SD = 0.76).

Lastly, source derogation was measured. This involves dismissing the expertise or trustworthiness or otherwise rejecting the validity of a source, which is a technique used to resist persuasion (Jacks & Cameron, 2010). This is measured using a previously used scale, existing of seven bipolar items (Jenkins & Dragojevic, 2011): ‘Stupid/smart’, ‘unknowledgeable/knowledgeable’, ‘uninformed/informed’, ‘unintelligent/intelligent’, ‘unqualified/qualified’, ‘unreliable/reliable’ and ‘inexpert/expert’, with answering options between 1 and 7. The items had a Cronbach’s α of 0.864, therefore all items were averaged into one source derogation measure (reach = 1-7, M = 3.33, SD = 1.10).

Dependent variable

The dependent variable in this research is the intention to test ecstasy. To measure this intention, an already existing scale for intention, existing of three items, has been used (Naughton, McCarthy, & McCarthy, 2013) and translated to the behavior of testing drugs. The items of this scale are ‘I want to test my ecstasy the next time I use it’, ‘I plan to test my ecstasy the next time I use it’ and ‘I am going to try to test my ecstasy the next time I use it’. The participants could answer these questions on a 7-point answering scale, with 1 = ‘totally disagree’ and 7 = ‘totally agree’. This scale has a Cronbach’s α of .940, therefore the separate items were averaged into one intention measure (reach = 1-7, M = 3.6, SD = 1.84).

Results

Randomization check

First, to check whether the measured background variables were of influence on the main effect, the random assignment to conditions was tested. Hence, an ANOVA with the type of persuasion as the independent variable and age as the dependent variable was conducted. This test indicated that there were no differences in the distribution of age among the conditions, F(3,152) = 1.95, p = .125, η2 = 0.04. After this, the same test was conducted with education as

(16)

the dependent variable, which showed that there were no differences in the distribution of education among the conditions F(3,152) = 0.33, p = .802, η2 = 0.01. Last, to check the distribution of gender between the conditions, a Chi-square test was conducted. This test indicated that no differences existed between the conditions on gender, Χ2(3, N = 156) = 1.55, p = .672. None of these variables influence the main effect and were consequently not included in the analyses.

Hypothesis testing

The direct and interaction effects of persuasion and framing on intention and reactance were tested first, since literature states that in the case that there is no moderation found on the mediator, or on the dependent variable, the moderated mediation model of Hayes (2013) should not be used. Moderated mediation can only occur when moderation occurs (Muller, Judd & Yzerbyt, 2005). Testing the direct and interaction effects was done by using a two-way analysis of variance and a multivariate analysis of variance.

The effect of persuasion and framing on intention

First, a two-way analysis of variance (ANOVA) with type of persuasion as the independent variable, framing as the moderator and intention as the dependent variable, was used to test whether a message that induces self-persuasion leads to a higher intention to test ecstasy than a message that uses direct persuasion (hypothesis 1) and whether this effect was stronger when a gain-frame instead of a loss-frame was used (hypothesis 5). The results indicate that the type of persuasion (direct vs. self) had a marginal significant effect on intention, F(1, 152) = 3.07, p = .082, η2 = 0.02, which shows that the intention to test ecstasy is a little higher in the direct persuasion condition (M = 3.86, SD = 1.73) than in the self-persuasion condition (M = 3.35, SD = 1.91). This effect is in the opposite direction than was expected, therefore hypothesis 1 is not supported. The group of participants that was exposed to a message that used self-persuasion, did not have a higher intention to test ecstasy. The ANOVA also showed that neither the way the message was framed (gain vs. loss), F(1, 152) = 0.043, p = .837, η2 = 0.00, nor the interaction of the type of persuasion and framing, F(1, 152) = 0.02, p = .897, η2 = 0.00, had a significant effect on the intention to test ecstasy. This means that there is no direct effect of framing or an interaction effect of persuasion and framing on intention; the effect of self-persuasion on intention to test ecstasy is not higher when a gain-frame is used. Hence, hypothesis 5 is also not supported.

(17)

The effect of persuasion and framing on reactance

As mentioned before, reactance was measured by four different scales. Therefore, a multivariate analysis of variance (MANOVA) was conducted with type of persuasion as the independent variable, type of frame as the moderator and source derogation, negative affect and threat to freedom as dependent variables, in order to test hypothesis 3 and 4. It was predicted that a gain-framed message leads to a lower score on reactance than a loss-framed message (hypothesis 3), however, no significant direct effect of type of framing (gain vs. loss) was found, Wilk’s lambda = 0.99, F(3,150) = 0.77, p = .511, η2 = 0.02. The way the message was framed did not influence the amount of reactance people experienced, consequently hypothesis 3 must be rejected. Also, no significant multivariate interaction effect was found, Wilk’s lambda = 0.97, F(3,150) = 1.33, p = .268, η2 = 0.03. This means that the effect of self-persuasion on reactance does not increase when a gain-frame is used, and that there is no interaction effect of persuasion and framing on intention. These findings contradict hypothesis 4, which therefore must also be rejected. This also means that there is no moderation of framing and that the model as a whole cannot be tested. Hypothesis 6, suggesting that the effect of self-persuasion on intention to test ecstasy is mediated by reactance, which is moderated by framing, must therefore be rejected as well.

The results however do show a significant multivariate effect of type of persuasion (self vs. direct) on threat to freedom, source derogation and negative affect, Wilk’s lambda = 0.93, F(3,150) = 3.96, p = .009, η2 = 0.07. When looking at the variables separately, this test indicated that the type of persuasion (direct vs. self) has a direct effect on threat to freedom, but contradicted the expectations, F(1,152) = 8.06 p = .005, η2 = .05. People that were exposed to a direct persuasion condition experienced significantly less threat to freedom (M = 1.86, SD = 0.87), than people that were exposed to a self-persuasion condition (M = 2.27, SD = 0.97). This result was also shown for negative affect. The type of persuasion (direct vs. self) had a direct effect on negative affect, contradicting the expectations, F(1,152) = 6.16, p = .014, η2 = .04. People that were exposed to the direct persuasion condition experienced significantly less negative affect (M = 1.61, SD = 0.62), than people that were exposed to the self-persuasion condition (M = 1.90, SD = 0.85). Type of persuasion, however, did not have a significant effect on source derogation, F(1,152) = 1.05 p = .308, η2 = 0.01.

The mediation effect of reactance on intention

Since the assumed model cannot be used, the second hypothesis - which presumes that the effect of self-persuasion on intention is mediated by reactance - is tested with PROCESS

(18)

model 4 (Hayes, 2013). From the other analyses, it was found that only threat to freedom and negative affect had an effect on reactance, therefore only these were used in further analyses. In the analysis with type of persuasion as the independent variable, intention as the dependent variable and threat to freedom as the mediator, a significant effect of type of persuasion on threat to freedom (b = -.41, SE = .15, p = .006, 95% BCACI [-.70, -.12]) was found and a marginal significant direct effect of type of persuasion on intention (b = .56, SE = .30, p = .065, 95% BCACI [-.03, 1.14]) was found. This shows a trend in which direct persuasion leads to a lower score on threat to freedom and a higher intention to test ecstasy and is in line with the aforementioned results. However, neither an effect of threat to freedom on intention was found (p = .515), nor a mediation effect of threat to freedom, 95% BCACI [-.24, .08].

In the analysis with type of persuasion as the independent variable, intention as the dependent variable and negative affect as the mediator, only a significant effect of type of persuasion on negative affect (b = -.29, SE = .12, p = .018, 95% BCACI [-.52, -.05]) was found. There was no effect found of negative affect on intention (p = .138), nor a direct effect of type of persuasion on intention (p = .149, 95% BCACI [-.16, 1.02]), nor a mediation effect of threat to freedom, 95% BCACI [-.01, .27]. This means that no mediation of reactance was found, so hypothesis 2 must be rejected.

Conclusion & Discussion

Conclusion

This research aimed to answer the question: ‘What is the effect of self-persuasion on reactance and intention to test ecstasy and what is the role of framing in this?’. The results show that the use of direct persuasion leads to a higher intention to test ecstasy than the use of self-persuasion. It was also shown that the type of persuasion that was used influenced reactance, but only on certain aspects. Direct effects of the type of persuasion on threat to freedom and negative affect were found, but contradicting the expected direction: self-persuasion led to more reactance. However, a note should be made about the strength of these effects. In all cases, the effect size was small and should be called very weak. The type of persuasion had no significant direct effect on the last aspect that was used to measure reactance; source derogation. There was also no effect found of framing, neither a direct effect nor an interaction effect. Framing therefore does not play a role in the effect of self-persuasion on reactance, nor on intention.

(19)

The reason why the effect of self-persuasion is in the opposite direction as expected, may be found in the fact that the effectiveness of persuasion is influenced by the experience of control (Damen et al., 2015). Most research on the effectiveness of self-persuasion has been conducted with behaviors like smoking, drinking or eating healthy (Baldwin et al., 2013; Krischler & Glock, 2015; Müller et al., 2009). With these researches, the results show that self-persuasion is more effective. It has been shown that self-persuasion is more effective when people experience high rather than low control (Damen et al., 2015), which is the case with those behaviors, since in order to perform those behaviors, normally no external parties are needed. The behavior of testing drugs could be less, or feel less, in the control of the participants. They cannot test the drugs themselves, but have to bring it to a facility and let it be done for them. This can also be seen in some responses on the cognitive response question; people doubt how the drugs are tested and whether or not this is trustworthy. They know they do not have control over it. Previous research has shown that direct persuasion is more effective when people experience low control (Damen et al., 2015), it could therefore be that in case of testing drugs, as the results show, direct persuasion leads to more effectiveness in the way that it leads to less reactance. The fact that no effect on intention was found, could also be explained by this. Since participants feel less control, they could be more skeptical towards the testing facilities. When the trustworthiness is questioned, this can make the threshold of testing higher and the intention to actually test their drugs lower (Baldwin et al., 2013). Also, the fact that a lot of participants state that they have friends or others who test the drugs for them, or before they buy it, might play a role in this. Participants also stated that they trust the person who they buy ecstasy from. They already think it is safe to use and will as a result have less intention to test it.

The fact that the results show that framing does not play a role, might be explained by the fact that several meta-analyses only found partial proof for the framing theory (O’Keefe & Jensen, 2007, 2009, 2011). Other recent research suggests that this can be explained by the suggestion that the translation of the prospect theory into the health domain might not be completely correct (van ‘t Riet et al., 2016). It seems that perceived risk is the crucial part that has been translated wrongfully (Harrington & Kerr, 2016; Latimer, Salovey, & Rothman, 2007; Van ‘t Riet et al., 2016). The risk-framing theory states, as mentioned before, that the risk is the possibility that something can go wrong or something forms a danger (Van ‘t Riet et al., 2016). As the theory in this research states, risk is about not wanting the consequence of a certain behavior. However, when looking at the prospect theory more closely, risk can be seen as the uncertainty that covers a variety of good and bad results, which are sometimes

(20)

unknown or cannot be estimated which one will occur (Harrington & Kerr, 2016; Van ‘t Riet et al., 2016). In this case, it is not as much about the kind of behavior, but more about how the outcomes or consequences are presented and about the uncertainties that come with this. With the use of a gain-frame, people will perform risk aversive behavior and choose the safe option in order to avoid uncertainty and to ensure a gain, for example when they have to choose how many lives to save. In contrary, with the use of a loss-frame, their choice will favor the uncertainty. They will perform risk-seeking behavior and take a gamble to avoid a loss, for example when they have to choose how many lives are lost (Tversky &Kahneman, 1981). These definitions cannot be seen as completely the same, and the difference between detection and prevention behavior might therefore not be as clear cut as mentioned before. This could explain why framing showed no effect in this case.

These results show that the theory about self-persuasion is not complete yet, and that indeed the conditions under which it works or does not work have to be defined more clearly. This research contributes to reducing the knowledge gap, by showing that self-persuasion might not work as well for any type of behavior, and that in this case, no effect was found when it is combined with different message frames. It also shows that the remarks made about the framing theory should be included in future research, as well as in combination with self-persuasion and different types of behaviors.

Discussion

Although the research did not show the expected results, it is of great importance. Little research has been done on self-persuasion and none in combination with framing. It is therefore still useful to take this research to the next level. However, some constraints of this research should be kept in mind.

This research has a methodological constraint, as people filled in the questionnaire online. This does not give insight in how people were participating - did they focus on the experiment or were they doing other things at the same time? The large amount of people that did not finish the questionnaire, namely 56, may suggest that people were easily distracted or did not feel like filling in the questionnaire after all. This is a difficult decision to take, since it is favorable to test in a natural setting, but it is difficult to say whether or not the results are influenced by other factors rather than the manipulation.

Another methodological constraint is that the used scale for cognitive response measured other factors than it was meant to measure. Participants wrote all kinds of things down, but also a lot of words like ‘party’, ‘lights’ and ‘girl’, which had nothing to with the

(21)

message. The coding did not give a relevant score on negative thoughts - which is a measurement for reactance - to use for analysis. Also, the responses were coded by hand and by one person, it could be that another coder would code differently. Future research should try to avoid this and use multiple coders and formulate the question in a way that the responses are more likely to be an evaluation of the message. Another measurement restraint is that a measurement for reactance, that is used a lot in previous research, was not taken into account for this research, namely counter-arguing (Asbeek Brusse et al., 2016). In this research, it was chosen not to use this measurement, since it might be strange to counter-argue with yourself. This however does have the disadvantage that it is difficult to compare the outcomes about reactance with previous research.

Another note could be made about the stimulus material. The image that was used was chosen because it would stimulate identification of the receiver with the advertisement. An image like the one that was used can on the other hand also influence the results, since it is not a neutral image. It can be very interesting to see whether or not different results can be found for different types of images. The use of an image that already has some sort of message in itself, which is the case with open advertisements, gives more freedom for interpretation, which leads to more appreciation of the advertisement and message (Jans, Ketelaar, & Van Gisbergen, 2008). Further research could look at the effects of neutral images versus no image or a steering image.

Since this research does not find results as expected from the literature, further research about this topic should be done in order to give more relevant recommendations. However, this research does add to the knowledge gap about self-persuasion. It shows that it is important to tailor a health message to the behavior it is trying to change. In this case, direct persuasion leads to less reactance towards messages about testing drugs, whereas previous research shows that for quitting smoking, self-persuasion works better. The amount of control that people experience might explain this and should therefore be taken into account for health messages. Not every marketing-communication technique is effective for any message or behavior. Governments and advertisers should carefully develop the right design for a message to reduce unhealthy behavior among people, and thereby increase overall health among citizens.

References

(22)

Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision processes, 50(2), 179-211. doi: 10.1016/0749-5978(91)90020-T

Asbeek Brusse, E. D., Fransen, M. L., & Smit, E. G. (2016). Framing in Entertainment-Education: Effects on Processes of Narrative Persuasion. Health Communication, 1-9. doi: 2048/10.1080/10410236.2016.1234536

Baldwin, A.S., Rothman, A.J., Vander Weg, M.W., & Christensen, A.J. (2013). Examining causal components and a mediating process underlying self-generated health arguments for exercise and smoking cessation. Health Psychology, 32(12), 1209-1217. doi:10.1037/a0029937

Bernritter, S. F., van Ooijen, I., & Müller, B. C. N. (2017). Self-persuasion as marketing technique: The role of consumers' involvement. European Journal of Marketing, 51(5/6), 1075-1090. doi:10.1108/EJM-04-2015-0213

Bogt, T. T., Engels, R., Hibbel, B., Van Wel, F., & Verhagen, S. (2002). “Dancestasy”: Dance and MDMA use in Dutch youth culture. Contemporary Drug Problems, 29(1), 157-181. doi:10.1177/009145090202900107

Brehm, J.W. (1966). A theory of psychological reactance. New York: Academic Press.

Briñol, P., McCaslin, M. J., & Petty, R. E. (2012). Self-generated persuasion: effects of the target and direction of arguments. Journal of personality and social psychology, 102(5), 925-940. doi:10.1037/a0027231

Burgoon, M., Alvara, E., Grandpre, J., & Voulodakis, M. (2002). Revisiting the theory of psychological reactance: Communicating threats to attitudinal freedom. In J.P. Dillard & M. Pfau (Eds.), The Persuasion Handbook: Developments in Theory and Practice. California: Sage Publications.

Burns, C. P. (2006). Cognitive Dissonance Theory and the induced‐compliance paradigm: concerns for teaching religious studies. Teaching Theology & Religion, 9(1), 3-8. doi:10.1111/j.1467-9647.2006.00255.x

Centraal Bureau voor de Statistiek (2017, May 19). Leefstijl en (preventief) gezondheidsonderzoek; persoonskenmerken. Retrieved from:

http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=83021ned&D1=34-36&D2=0-13,30-42&D3=0&D4=l&HDR=T&STB=G1,G2,G3&VW=T

Cialdini, R.B. (2014). Influence: science and practice (5th ed., Pearson new international ed. Ed.). Harlow: Pearson Education Limited.

Cobuild, C. (2006). Collins Cobuild advanced learner's English dictionary. Glasgow, UK: HarperCollins Publishers.

(23)

De Bruijn, G. J., Spaans, P., Jansen, B., & van‘t Riet, J. (2016). Testing the effects of a message framing intervention on intentions towards hearing loss prevention in adolescents. Health education research, 31(2), 161-170. doi:10.1093/her/cyw006

Damen, T.G.E. Muller ,B.C.N., van Baaren, R.B. & Dijksterhuis, A. (2015). Re-Examining the Agentic Shift: The Sense of Agency Influences the Effectiveness of (Self)Persuasion. PLoS ONE 10(6): e0128635. doi:10.1371/journal.pone.0128635 Dillard, J. P., & Peck, E. (2000). Affect and persuasion: Emotional responses to public service

announcements. Communication Research, 27, 461–495.

Dillard, J. P., Plotnick, C. A., Godbold, L. C., Freimuth, V. S., & Edgar, T. (1996). The multiple affective consequences of AIDS PSAs: Fear appeals do more than scare people. Communication Research, 23, 44–72.

Dillard, J.P. & Shen, L. (2005). On the Nature of Reactance and its Role in Persuasive Health Communication. Communication Monographs, 72(2), 144-168,

doi:10.1080/03637750500111815

Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press.

Fransen, M.L., Verlegh, P.W.J., Kirmani, A. and Smit, E.G. (2015). A typology of consumer strategies for resisting advertising, and a review of mechanisms for countering them. International Journal of Advertising, 34(1), 6–16.

doi:2048/10.1080/02650487.2014.995284

Fitzsimons, G.J. & Williams, P. (2000). Asking questions can change choice behavior: Does it do so automatically or effortfully?. Journal of Experimental Psychology: Applied, 6(3), 195–206. doi: 10.1037/1076-898X.6.3.195

Hamilton, W.L., Biener, L. & Brennan, R.T. (2008). Do local tobacco regulations influence perceived smoking norms? Evidence from adult and youth surveys in Massachusetts. Health Education Research , 23(4), 709–722. doi: 10.1093/her/cym054

Harrington, N. G., & Kerr, A. M. (2016). Rethinking risk: prospect theory application in health message framing research. Health Communication, 32(2), 131-141.

doi: 10.1080/10410236.2015.1110004

Hayes, A. F. (2013). An introduction to mediation, moderation, and conditional process analysis: A regression- based approach. New York: The Guilford Press.

Jacks, J.Z. & Cameron, K.A. (2003). Strategies for Resisting Persuasion. Basic and Applied Social Psychology, 25(2), 145-161. doi: 10.1207/S15324834BASP2502_5

(24)

Jans, K. S. H., Ketelaar, P. E., & Van Gisbergen, M. S. (2008). Merkkennis verhoogt waardering van advertentie-openheid. Merk & Reputatie, 14(2), 48-51

Jellinek (2016). Hoeveel mensen overlijden er door alcohol, tabak en andere drugs?. Retrieved from: https://www.jellinek.nl/vraag-antwoord/hoeveel-mensen-overlijden- er-door-alcohol-tabak-en-drugs/

Jellinek (2017). Hoeveel mensen gebruiken XTC?. Retrieved from:

https://www.jellinek.nl/vraag-antwoord/hoeveel-mensen-gebruiken-xtc/

Jenkins, M., & Dragojevic, M. (2013). Explaining the process of resistance to persuasion: A politeness theory-based approach. Communication Research, 40(4), 559-590. doi:10.1177/0093650211420136

Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1991). Anomalies: The Endowment Effect, Loss Aversion, and Status Quo Bias. The Journal of Economic Perspectives, 5(1), 193-206. doi: 10.1257/jep.5.1.193

Klompenhouwer, L. (2014, October 20). Derde dode tijdens ADE, vrouw overleed door xtc-gebruik. NRC.nl. Retrieved from https://www.nrc.nl/nieuws/2014/10/20/derde-dode-na- amsterdam-dance-event-a1421620

Knowles, E. S., & Linn, J. A. (Eds.). (2004). Resistance and persuasion. Mahwah, New Jersey: Lawrence Erlbaum Associates.

Kreuter, W. M., Bull, F. C., Clark, E. M., & Oswald, D. L. (1999). Understanding how people process health information: A comparison of tailored and nontailored weight-loss materials. Health Psychology, 18, 487 - 494. doi:10.1037/0278-6133.18.5.487

Krischler, M., & Glock, S. (2015). Alcohol warning labels formulated as questions change alcohol-related outcome expectancies: A pilot study. Addiction Research &Theory, 23(4), 343-349. doi:10.3109/16066359.2015.1009829

Latimer, A. E., Salovey, P., & Rothman, A. J. (2007). The effectiveness of gain-framed messages for encouraging disease prevention behaviour: is all hope lost? Journal of Health Communication, 12, 645-649. doi: 10.1080/10810730701619695

Meyerowitz, B. E., & Chaiken, S. (1987). The effect of message framing on breast self- examination attitudes, intentions, and behavior. Journal of personality and social psychology, 52(3), 500 - 510. doi: 10.1037/0022-3514.52.3.500

Miron, A.M., & Brehm, J.W. (2006). Reactance theory – 40 years later. Zeitschrift für Sozialpsychologie, 37(1) 9-18. doi: 10.1024/0044-3514.37.1.9

Morton, T. A., Rabinovich, A., Marshall, D., & Bretschneider, P. (2011). The future that may (or may not) come: How framing changes responses to uncertainty in climate change

(25)

communications. Global Environmental Change, 21(1), 103–109. doi:10.1016/j.gloenvcha.2010.09.013

Müller, B. C., van Baaren, R. B., Ritter, S. M., Woud, M. L., Bergmann, H., Harakeh, Z., . . . Dijksterhuis, A. (2009). Tell me why... The influence of self-involvement on short term smoking behaviour. Addict Behavior, 34(5), 427-431.

doi:10.1016/j.addbeh.2008.12.016

Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation is moderated. Journal of personality and social psychology, 89(6), 852. doi: 10.1037/0022-3514.89.6.852

Munger, M.C. (2011). Persuasion, psychology and public choice. Journal of Economic Behavior and Organization, 90(2), 290-300. doi: 10.1016/j.jebo.2011.07.012

Mussweiler, T., & Neumann, R. (2000). Sources of Mental Contamination: Comparing the Effects of Self-Generated versus Externally Provided Primes. Journal of Experimental Social Psychology, 36(2), 194-206. doi:10.1006/jesp.1999.1415

Naughton, P., McCarthy, S., & McCarthy, M. (2013). Healthy eating attitudes and healthy living: An examination of the relationship between attitudes, food choices and lifestyle behaviours in a representative sample of Irish adults. Proceedings of the Nutrition Society(2013), 72 (OCE4), E221. doi:10.1017/S0029665113002462

Niesink, R. J. M., Brunt, T. B., & Croes (2015). Ecstasy: PMMA, MDMA en Hooggedoseerde pillen. Verslaving, 1-14. doi:10.1007/s12501-015-0020-3.

Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions. Psychological bulletin, 133, 673-693. doi:10.1037/0033-2909 .133.4.673

O’Keefe, D. J., & Jensen, J. D. (2007). The relative persuasiveness of gain- framed and loss framed messages for encouraging disease prevention behaviors: A meta-analytic review. Journal of Health Communication, 12, 623–644.

doi:10.1080/10810730701615198

O’Keefe, D. J., & Jensen, J. D. (2009). The relative persuasiveness of gain- framed and loss framed messages for encouraging disease detection behaviors: A meta-analytic review. Journal of Communication, 59, 296–306. doi:10.1111/j.1460 2466.2009.01417.x

O’Keefe, D. J., & Jensen, J. D. (2011). The relative effectiveness of gain- framed and loss framed persuasive appeals concerning obesity-related behaviors: Meta-analytic evidence and implications. In R. Batra, P. A. Keller, & V. J. Strecher (Eds.), Leveraging

(26)

consumer psychology for effective health communications: The obesity challenge (pp.171–185). Armonk, NY: M. E. Sharpe.

Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of personality and social psychology, 37(10), 1915-1926. doi: 10.1037/0022-3514.37.10.1915

Pornpitakpan, C., & Francis, J. (2001). The Effect of Cultural Differences, Source Expertise, and Argument Strength on Persuasion: An Experiment with Canadians and Thais. Journal of International Consumer Marketing, 13(1), 77-102.

doi:10.1300/J046v13n01_06

Rothman, A. J., Bartels, R. D., Wlaschin, J., & Salovey, P. (2006). The strategic use of gain- and loss-framed messages to promote healthy behavior: How theory can inform practice. Journal of Communication, 56, 202-220. doi:10.1111/j.1460- 2466.2006.00290.x

Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: the role of message framing. Psychological bulletin, 121(1), 3-19.

doi:10.1037/0033-2909.121.1.3

Trimbos: Netherlands Institute of Mental Health and Addiction (2016a). Drugs Informatie Monitoring Systeem Jaarbericht 2016. Retrieved from: https://www.trimbos.nl/kerncijfers/nationale-drug-monitor

Trimbos: Netherlands Institute of Mental Health and Addiction (2016b). Nationale drug monitor. Jaarbericht 2016. Retrieved from https://assets.trimbos.nl/docs/3fdeab39-f34e-4aa0-97af-5dbf111c05a0.pdf

Trimbos: Netherlands Institute of Mental Health and Addiction (2017). Red alert. Retrieved from: https://www.drugsredalert.nl

Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211(4481), 453-458.

Unity (2017b). Homepage. Retrieved from: http://www.unity.nl

Unity (2017a, may 24 ). Global Drug Survey 2017: de Nederlandse situatie. Retrieved from: http://www.unity.nl/global-drug-survey-2017-de-nederlandse-situatie/#

Van ‘t Riet, J., Cox, A. D., Cox, D., Zimet, G. D., DeBruijn, J., Van Den Putte, B., De Vries, H., Werrij, M. Q., & Ruiter, R. A. C. (2016). Does perceived risk influence the effects of message framing? Revisiting the link between prospect theory and message framing. Health Psychology Review, 10(4), 447-459.

(27)

Wouters, M., Nabben, T., Benschop, A., & Korf, D. J. (2014). Drug use trends in Amsterdam nightlife, 2013. Jellinekreeks, 25. Amsterdam: Rozenberg Publishers. Retrieved from: http://dare.uva.nl/document/2/15025

Yi, Y. (1990). Direct and indirect approaches to advertising persuasion, which is more effective? Journal of Business Research, 20(4), 279-291. doi:10.1016/0148-2963(90)90007-Z

Appendices

(28)
(29)

Appendix 2: Questionnaire Beste deelnemer,

Leuk dat je mee wilt doen met dit onderzoek! Voor je gaat beginnen, is het belangrijk dat je de onderstaande tekst goed doorleest.

Dit onderzoek wordt uitgevoerd onder de verantwoordelijkheid van de School of Communication, Universiteit van Amsterdam. Dit geeft de garantie dat er vertrouwelijk met je antwoorden wordt omgegaan, je antwoorden of gegevens onder geen enkele voorwaarde aan derden wordt verstrekt, je zonder opgaaf van redenen kunt weigeren mee te doen aan het onderzoek en je deelname altijd voortijdig kunt afbreken.

Het onderzoek brengt geen noemenswaardige risico’s of ongemakken met zich mee. Ook zal geen moedwillige misleiding plaatsvinden en zul je niet met expliciet, aanstootgevend materiaal worden geconfronteerd. Voor meer informatie over het onderzoek en de uitnodiging tot deelname kun je ten allen tijde contact opnemen met de verantwoordelijke onderzoeker Sietske door een email te sturen naar sietske_abelmann@hotmail.com. Mocht je interesse hebben in de uitkomsten van het onderzoek, dan kunnen deze worden opgestuurd indien er een email wordt gestuurd naar het bovenstaande emailadres van de onderzoeker.

Mochten er naar aanleiding van je deelname aan dit onderzoek klachten of opmerkingen zijn, dan kun je contact opnemen met het lid van de Commissie Ethiek van de afdeling Communicatiewetenschap, per adres: ASCoR secretariaat, Commissie Ethiek, Universiteit van Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020-5253680;

Referenties

GERELATEERDE DOCUMENTEN

Derhalve kan worden geconcludeerd op basis van de resultaten zoals beschreven in de voorgaande paragrafen en deze paragraaf, dat beursgenoteerde klanten van de Big Four geen

Here, the slope of conditional direct effect appeared to be significant only for the high levels SN, in other words, only for the situations were procrastination was seen as

It can thus be concluded that MNEs headquartered in countries with enhanced health systems are less likely involved in both health and development violations of human

Tissue specific expression was observed in transgenic sugarcane where expression of the reporter gene was regulated by the UDP-glucose dehydrogenase promoter and first

Ceasefire critics accuse military leaders on all sides of oppor- tunism reminiscent of General Ne Win’s controversial Ka Kwe Ye ‘home-guard’ programme, which allowed selected ethnic

So this is - and this is the second thing and going back to something hinted upon earlier - in the way in the Americans in particular, there is a huge disappointment, particular

Hypothesis 3: Status and Pro-Social Associated Products, Strengthened by Costly Signalling Appeals There is a three way interaction between product associations strengthened by

It examined the effect that employees’ perceptions of organizational support, supervisor support, training, development and career opportunities, performance feedback