The effect of positively and negatively framed moral appeals on students filling out course evaluations, moderated by response history
University of Groningen Faculty of Economics and Business
Master Thesis
Msc Human Resource Management
27 June 2012 Sanne Carabain Nicolaihof 40 8031 PC Zwolle 06-12589188 sanne_carabain@hotmail.com Student number: 1738380 Supervisor/university: L.B. Mulder Second supervisor/university: F.A. Rink
ABSTRACT
INTRODUCTION
Nonresponse to course evaluations is a big problem for the Faculty of Economics and Business at the University of Groningen, since a lot of students do not fill out these evaluations. According to the ‘Handboek kwaliteitszorg onderwijs FEB’ (2008) the course evaluations are an important source for improving the quality of the University. The faculty pursues an average response rate of 40%, and for having a representative picture of the course, a minimum of 20% is required. Having an optimal evaluation system can result in improving the education, more satisfied students, and meeting the different requirements concerning the education level. With these results, the University wants to register which outcomes are being achieved and, if necessary, they can take corrective measures. The goal of this research is to examine if an appeal to the morality of students can improve the response to course evaluations, and if there is a difference between students who hardly ever fill out the course evaluations and students who often fill out the course evaluations.
Filling out a course evaluation or not, is a decision that is also called a social dilemma. In social dilemma situations, people have to choose between two options; one that may represent your own interests (defection), and one that represents the group’s interest (cooperation). Each individual always receives a higher payoff for defecting than for cooperating, but all are better off if all cooperate than if all defect (Dawes and Messick, 2000: p. 111). Comparing this with the course evaluations, it is in students’ self interest not to fill out these evaluations as it costs some time and effort. However, if all students do not fill out the course evaluations, a course cannot be evaluated, and subsequently not be improved. This may result in a quality decrease of a course, which concerns all students. When all the individuals decide to cooperate and fill out the evaluations, all students are better off by receiving high quality education, than when all students do not.
affect tax compliance (e.g., Bosco and Mittone, 1997; Schwartz and Orleans, 1967; Webley, Cole and Eidjar, 2001; Wenzel, 2004; Schwartz, 1977). Schwartz and Orleans (1967) further found that appeal to personal conscience of taxpayers could increase their tax compliance.
In the same way, personal norms can play a role for students filling out the course evaluations. By addressing to personal norms of people, moral appeals may increase moral behaviour. In current paper, the definition of Beniers (2009) will be used for moral appeals; ‘moral appeals are directed to the audience’s sense of what is right and proper’. According to Chen, Pillutla, and Yao (2009: p. 247), moral appeals may direct people’s attention to the greater purpose of the group, and the non-instrumental value of cooperation. Moral appeals are more likely to be seen as an appeal to one’s better nature, instead of an external pressure to cooperate. They can be used to increase moral behaviour, which was examined by the use of appeals to increase tax compliance (Schwartz and Orleans, 1967). These researchers used a field experiment to test the effect of appeals, compared to sanctions, on compliance. Participants of the appeal condition were shown moral reasons for compliance with tax law, which included questions that emphasised noncontroversial uses for which tax money is employed, citizen obligation to government, and the value of personal integrity (Schwartz and Orleans, 1967: p. 274). The results of one of their studies showed that participants who read the appeal increased their tax payment more than participants who were threatened with the sanction, which were increases of respectively 804 dollar and 181 dollar. So based on this, it can be stated that both a sanction threat and an appeal increased moral behaviour, however, the effect of an appeal is larger.
Moral appeals might also work the other way around. The emotions people feel about an appeal can be negative too (e.g. fear, guilt, and shame), which in turn might affect their future actions (Batra and Ray, 1986). Earlier research of Brehm (1981) introduced the term psychological reactance, which explained why individuals might react aversive against persuasive messages, like an appeal to one’s morality. The essence of this theory assumes that individuals cherish their ability to choose between different options, and that they do not want to be eliminated in their choice. When a moral appeal leads to reactance, this can to lead people doing the forbidden act (Dillard and Shen, 2005).
In addition, the way people will react to a moral appeal might depend on previous (moral) behaviour. Falk and Zimmermann (2011) stated that an important determinant of people’s behaviour is the need to be consistent. According to Khan and Dhar (2006), people make often a series of equal decisions. If people behaved morally in the past, they will continue to do that in the future. However, it might also happen that people do not behave morally all the time, which may lead to a threatened moral self-concept. When this occurs, motivation will accrue to reaffirm the self and try to restore their moral self-concept (Zhao and Nan, 2010). This self-affirmation theory states that people have the need to perceive themselves as moral. Controversially, licensing behaviour is the opposite. It means that one’s behaviour in one direction, can license behaviour in the contrary direction (Monin and Miller, 2001; Khan and Dhar, 2006). Someone who behaves morally may feel liberated to behave immorally in the future.
THEORETICAL BACKGROUND
Human behaviour
As Falk and Zimmermann (2011) stated, a powerful determinant of human behaviour is the desire to be and appear consistent. Bénabou and Tirole (2006) also discussed that one’s behaviour is influenced by a strong need to maintain conformity between a person’s actions, feelings and values. Once people have made a decision or taken a stand, they feel pressure to live up to that commitment. According to Khan and Dhar (2006), initial actions can activate specific goals that guide subsequent choices in the direction of attaining an ultimate personal goal. Following this, students who want to behave morally (as their ultimate personal goal) behave consistently in their actions and continue to show moral behaviour. Additionally, Khan and Dhar (2006) said that people often make a series of equal decisions in which one choice follows another. So if people act in a moral way, they act in a similar way next time. Based on previous mentioned theory, it can be expected that there will also be some consistency in the behaviour of students. When they are filling out the course evaluations, they will continue to do this in the future.
Moreover, people have the need to obtain a moral self-concept, but people do not behave morally all the time and they do not always have the same feeling about their global self-worth and self-integrity. If the positive moral image of the self is threatened, people become motivated to reaffirm the self and subsequently try to restore the perceived integrity of the self (Zhao and Nan, 2010). This is called self-affirmation. The fundamental assumption of self-affirmation theory is that people have an innate need to perceive themselves as morally adequate (Steele, 1988). Since this positive moral self-concept derives from many aspects of the self, people have flexibility in how to cope with a specific threat (Cohen, Aronsom, and Steele, 2000). People can, as was described above, reaffirm the self directly and compensate their previous immoral behaviour. For instance, Carlsmith and Gross (1969) stated that when moral values had been violated, compliance to requests for help increased. An immoral action may stimulate future moral behaviour in order to restore the positive moral self-concept (Sachdeva, Iliev, and Medin, 2009). People can also reaffirm indirectly, by affirming other positive personal values or characteristics of the self (Cohen, Aronsom, and Steele, 2000), in this way, the global image of the self remains also positive.
one’s previous moral behaviour, can license behaviour in the opposite (immoral) direction. Someone who behaves prosocially may feel liberated to subsequently behave antisocially in the future. Khan and Dhar (2006) stated that a prior choice of someone, of which one can increase their positive self-concept and the believe that they behave morally, can lead to licensing a following choice of behaving immorally. They conducted a study where participants in the license condition have to make two following decisions. Firstly, the participants were asked to imagine that they donated $100, of an income tax rebate of $500, to a charity organization. Secondly, they were asked to choose between a cheap, necessary pair of sunglasses, and an expensive, luxury pair of sunglasses. Participants of the control condition did not donate the $100 to a charity organization, but they only were asked to choose between the cheap or luxury pair of sunglasses. The results of this study demonstrated that 56.5% of the participants in the licensing condition chose to buy the expensive pair of sunglasses, as opposed to the participants in the control condition, of which only 27.7% chose the expensive pair of sunglasses. So, the choice for a more self-indulgent option increases after a prior virtuous choice.
Framing the moral appeals and the influence of previous behaviour
As was described in the review of Levin, Schneider, and Gaeth (1998), framing a message can influence the decision-making process of people. For instance, many health-related messages use a specific framing, namely gain framing (emphasizing the positive) and loss framing (emphasizing the negative). A clear example is given in the paper of Shamaskin, Mikels, and Reed (2010: p. 746); a positive framed message regarding skin cancer might be ‘The earlier it is detected, the better the person’s chances are for full recovery’. In contradiction, the negative message might be ‘The later it is detected, the poorer the person’s chances are for full recovery’. The anti-smoking campaign in the Netherlands also uses positive and negative framed messages. For instance, a positive message is ‘stop smoking reduces the risk of fatal heart- and lung diseases’, and the negative message is ‘smoking causes fatal lung cancer’.
However, Kahneman and Tversky (1979) found that there is a stronger effect of negative framing on people’s behaviour. Baumeister, Bratslavasky, Finenauer, and Vohs (2001) and Rozin and Royzman (2001) showed also that negative messages have more impact than positive messages. Other studies suggested that negative activities have more impact than positive activities. Kahneman and Tversky (1979) stated, for instance, that gains are twice as powerless as losses. Baumeister, Bratslavasky, Finenauer, and Vohs (2001) found that negatively rated events (e.g. receiving criticism) have a greater impact on people than positively rated events of the same type (e.g. receiving a compliment). This may also apply to negative feedback or information. Negative feedback has more impact than positive feedback, and negative information is processed more thoroughly than positive information.
Low response history
A moral appeal framed in a negative way sounds more forced, and less friendly than a moral appeal framed in a positive way. The negative moral appeal can lead to psychological reactance, which is an aversive motivational state that occurs when a freedom is eliminated or threatened with elimination (Brehm and Brehm, 1981). Psychological reactance occurs when any persuasive message arouse a motivation to reject the advocacy (Dillard and Shen, 2005). When psychological reactance is high, people are more likely to produce judgments of message attributes such that the message is rated as overstated, exaggerated, or overblown (Shen et al., 2009; Witte, 1994; in Zhao and Nan, 2010: p. 496). According to Quick and Stephenson (2007) psychological reactance may result in boomerang effects and will lead individuals to be motivated to re-establish their freedom, which involves doing the forbidden act (Dillard and Shen, 2005). An example is shown in the research of Pennebaker and Sanders (1976). They put one of two messages on college bathroom walls. On one wall students read: ‘Do not write on these walls under any circumstances’, and on the other wall students read: ‘Please do not write on these walls’. A few weeks later, the walls with the first message had far more graffiti than the walls with the second message.
theory, one can say that reactance will decrease the response rate to course evaluations, and therefore it is assumed that a positively framed moral can better be used for increasing the response rate to course evaluations. This positive moral appeal sounds less forced and friendlier, which results in less reactance among students with low response history.
High response history
Students who often fill out the course evaluations may react differently on a moral appeal than students who hardly ever fill out the course evaluations. Monin and Miller (2001), and Khan and Dhar (2006) studied licensing behaviour of people. If someone behaves morally, which results in an increased positive self-concept, he or she might feel liberated to behave in an immoral way in the future. As Khan and Dhar (2006) concluded, the choice for a more self-indulgent option increased after a prior virtuous choice. For students, this means that if they often filled out the course evaluations, they might feel liberated and licensed to fill out an evaluation in the future.
If a negatively framed moral appeal is shown to these students, who might consider not to fill out a course evaluation, their immoral behaviour is emphasised and the appeal confirms that they will act in an immoral way. In addition, according to the self-affirmation theory, these negative messages threaten the individual’s sense of self-integrity (Steele, 1988). This leads students to compensate their immoral behaviour and therefore, using a negative moral appeal leads students with high response history to reaffirm the self, and tries to restore the perceived integrity and the positive image of the self (Zhao and Nan, 2010). So, students who often fill out course evaluations will continue to do this when a negatively framed moral appeal is shown to them. Based on this, the following hypothesis can be formulated:
Hypothesis: The effect of framing of moral appeals on a student’s fill out behaviour is moderated by previous history of filling out the course evaluation in the following way:
a: Using a negatively framed moral appeal will decrease the student’s fill out behaviour to course evaluations among students who hardly ever fill out the course evaluations, compared to the use of a positively framed moral appeal.
METHOD
Participants and design
The participants were students of the Faculty of Economics and Business at the University of Groningen, who followed a course during semester 2.1. They were randomly assigned to the positively framed moral appeal condition or negatively framed moral appeal condition. Twenty students (10 males, 7 females, 3 unknown; Mage = 22.29, SDage = 2.37) who were in the negative moral appeal condition completed the questionnaire, and fourteen students (4 males, 6 females, 4 unknown; Mage = 21.20, SDage = 2.53) who were in the positive moral appeal condition completed the questionnaire. The response rate for the negative moral condition was 9.1%, and the response rate for the positive moral appeal condition was 6.4%.
Independent variables were response history (measured) and the manipulation of the moral appeal (positively framed moral appeal versus negatively framed moral appeal).
Procedure
All students of the faculty received an invitation e-mail after semester 2.1, with a link to the course evaluations.
In both conditions, students first read the following:
“Course evaluations are essential to improve the quality of education, which has high priority in the Faculty of Economics and Business. The more students evaluate the courses that they have followed, the better picture we get of the quality of education and the more tools the faculty has for improving it. Therefore, students have, according to the faculty, the moral obligation to fill out the course evaluations”.
Subsequently, in the negative moral appeal condition, students read:
And in the positively moral appeal condition, students read:
“If you fill out the course evaluations, you underline the faculty’s values, and contribute to the education improvement program. In this way, you enable the improvement of the education and educational programs. Therefore, the faculty counts on students feeling responsible for filing out the evaluations”.
About 2 weeks after the invitation e-mail for the course evaluations was sent, all students received an invitation by e-mail to fill out a questionnaire about the way they were invited to fill out the course evaluations of semester 2.1. in year 2011/2012. At the beginning of this questionnaire, they were presented the same text by which they were invited to fill out the course evaluations. Students were asked if they recognised the text and if they read the text when they received the invitation e-mail. Subsequently, the text in the invitation e-mail was shown again and students were asked to read it carefully before they proceed with the questionnaire.
After that, participants were asked a number of questions to measure the manipulation, to measure licensing and compensation, some general questions about the course evaluations and if they filled out this time and their past response behaviour. Next, participants were asked questions about the extent of reactance they felt when they read the moral appeal and questions about future intentions to fill out course evaluations. Finally, the participants were able to write down any comments about the course evaluations or the questionnaire. And after that, the participants were thanked for their cooperation.
Measures and manipulations
Moral appeal. The moral appeal was manipulated as described in the procedure section. The negative moral appeal condition was coded as -1, and the positive moral appeal condition was coded as 1.
Check. To check whether students find the message in the invitation e-mail an appeal to their morality, participants indicated on a 7-point scale (1 = strongly disagree, 7 = strongly agree) whether they agree with the following statements: (1) I experience the message as an appeal to my morality, (2) the message aims to call upon my sense of morality. These items formed a reliable scale (α = .78).
= strongly agree) to what extent they agree with the following statements: After receiving and reading this invitation to fill out the course evaluation…. (1) My belief of being a moral person was strengthened, and (2) I felt good about myself. These two items formed no reliable scale, and will be measured apart from each other (α = .58).
Need for compensation. To examine if participants felt that they had to compensate the self, they needed to indicate on a 7-point scale (1 = strongly disagree, 7 = strongly agree) whether they agreed with the following statements: After receiving and reading this invitation to fill out the course evaluation… (1) I wanted to improve my behaviour in the past, (2) I wanted to make more effort than before, (3) I felt that I had to compensate something. These items were averaged to determine the need for compensation (α = .86).
Current fill out behaviour. To see whether or not students filled out the course evaluations this time, they needed to answer the question “did you fill out the course evaluations this time?” on a 3-point scale (1) no, (2) partly, (3) yes.
Response history. Students were asked to indicate on a 5-point scale to what extent they filled out course evaluations in the past, (1) never, (2) rarely, (3) sometimes, (4) often, (5) always.
Reactance. Reactance was measured using the scale of Dillard and Shen (2005). The developer of the psychological reactance theory, Brehm, stated that reactance cannot be measured, however, Dillard and Shen rejected this. Their data showed that reactance could be operationalized as a composite of self-report indices of anger and negative cognitions. Participants had to indicate on a 7-point scale (1 = strongly disagree, 7 = strongly agree) on what extent they agreed with the following statements: the message as presented in the invitation email (1) irritated me, (2) made me feel angry, (3) annoyed me, (4) made me aggravated, (5) threatened my freedom to choose, (6) tried to make a decision for me, (7) tried to manipulate me, and (8) tried to pressure me. These eight items were averaged to determine the reactance score for each participant (α = .86)
RESULTS
Descriptive statistics
Table 1 presents the means, standard deviations, and correlation coefficients between all variables of current study. Most important variables according to the hypothesis are the framed moral appeal, response history, and students’ current fill out behaviour. There was no relation between the moral appeal and a student’s current fill out behaviour (r = .11, p = n.s.). There neither was a relation between the moral appeal and response history (r = -.18, p = n.s.). So, the moral appeal conditions did not correlate with the current fill out behaviour, and did not correlate with response history. However, there was a significant relation between response history and a student’s current fill out behaviour (r = .51, p < .01). This indicates that the response history correlates with current fill out behaviour.
Table 1
Means (M), Standard Deviations (SD), and correlations
Variable M SD 1 2 3 4 5 6 7 8
1 Moral appeal -.18 1.00
2 Response history 3.62 1.21 -.18
3 Students’ fill out behaviour 2.35 .92 .11 .51**
4 Reactance 4.14 1.18 -.29+ .11 -.09
5 Self-affirmation1 3.35 1.65 .44** .04 .06 -.45** 6 Self-affirmation2 3.94 1.41 .47** .09 .20 -.49** .41*
7 Compensation 3.04 1.34 .31+ -.31+ -.14 -.40* .73** .42** 8 Behavioural intentions 4.75 1.51 .05 .54** .63** -.17 .21 .27 .00 +. Correlation is marginally significant at the .10 level
*. Correlation is significant at the .05 level **. Correlation is significant at the .01 level
Check
A one-sample t-test was conducted to examine if the message in the invitation e-mail was an appeal to students’ morality, with the average test value of 4 on the scale from 1 (don’t receive the e-mail as an appeal to morality) to 7 (receive the e-mail as an appeal to morality). The results showed that students who read the message in the invitation e-mail received it as an appeal to their morality (M = 4.82, SD = 1.36, p = .00).
Hypothesis testing
It was hypothesised that the effect of framing the moral appeals on a student’s current fill out behaviour is moderated by previous history of filling out the course evaluations, and that the use of a negatively framed moral appeal, in comparison to a positive moral appeal, will increase the fill out behaviour among students who often filled out the course evaluations in the past, but will decrease among students who hardly ever filled out the course evaluations in the past. In order to test this hypothesis, a linear regression was conducted. In the first step, I entered moral appeal and the response history (standardised). In the second step, I added the moral appeal x response history of students. The results are presented in Table 2.
In this linear regression, the 7 participants (20%) who indicated that they did not read the moral appeal, or do not remember if they read it when they received the invitation e-mail were not included in the analysis, because this might bias the results. The table shows that, in step 1, there was no main effect of the framing of a moral appeal on a student’s current fill out behaviour (B = -.04, p = .80) and there was a significant effect of the response history on a student’s fill out behaviour (B = .34, p = .02). After entering the interaction in step 2, there was again no significant main effect of the framing of the moral appeal on a student’s current fill out behaviour (B = -.03, p = .82). There was still a significant effect of response history on student’s fill out behaviour (B = .34, p = .03). The moderating effect of response history on the relationship between moral appeal and a student’s current fill out behaviour was not significant (B = -.06, p = .66; ∆R2 = .01, p = .66). Based on these results, the hypothesis cannot be supported.
Table 2
Linear regression of the relation between the framing of a moral appeal and student’s current fill out behaviour, with response history as moderator (N = 27)
Model
Variable 1 2
Moral appeal Response history
Moral appeal x Response history R2 ∆R2 -.04 .34* -.03 .34* -.06 .21 .21 .22 .01 *: p < .05 **: p < .01
Table 3
Linear regression of the relation between the framing of a moral appeal and students’ future intentions to fill out course evaluations, with response history as moderator (N = 34)
Model
Variable 1 2
Moral appeal Response history
Moral appeal x Response history R2 ∆R2 .23 .86** .24 .86** .08 .31 .31 .31 .00 *: p < .05 **: p < .01 Reactance
In the theoretical background was stated that a negatively framed moral appeal could result in an aversive attitude of students against the message, which might lead to boomerang effects. A linear regression analysis was conducted to see if there is an effect between the framing of a moral appeal and the extent of reactance. Furthermore, the moderating effect of response history on the relationship between the moral appeal and reactance was also tested.
Figure 1
RH = Response history NMA = Negative moral appeal
PMA = Positive moral appeal
This interaction was plotted in Figure 1. As you can see in the graph, the students with high response history felt more reactance in the negative moral appeal condition than in the positive moral appeal condition. For students with low response history, there was hardly any difference between the negative moral appeal condition and the positive moral appeal condition.
Self-affirmation
Two linear regression analyses were performed to test the influence of the moral appeal on the need for self-affirmation. Since cronbach’s alpha was .58, the two questions about the need for self-affirmation were tested apart from each other.
First, students needed to indicate if their belief of being a moral person was strengthened after reading the moral appeal. With the linear regression analysis, the effect of the moral appeal on student’s belief of being moral was tested, with the moderator response history. In the first step, the framing of the moral appeal and student’s belief of being a moral person were entered, and in the second step, moral appeal x response history was added. The results showed that in step 1, there was a main effect of the moral appeal on the belief of being a moral person (B = .77, p = .01) and there was no significant effect of response history on the belief of being a moral person (B = .21, p = .45). After adding the
interaction in the second step, there was still a significant main effect of the framing of the moral appeal on students’ belief of being a moral person (B = .79, p = .00), and there was no significant effect of response history on the belief of being a moral person (B = .25, p = .34). There was a marginally significant moderating effect of response history on the relation between the moral appeal and students’ belief of being a moral person (B = .49, p = .07; ∆R2 = .08, p = .07).
This marginally significant interaction was plotted in Figure 2. As you can see in the graph, for both students with low response history and high response history their belief of being a moral person was strengthened more in the positive moral appeal condition than in the negative moral appeal condition. However, for students with high response history this belief was stronger than for students with low response history.
Second, students could also indicate to what extent they felt good about themselves when they read the moral appeal. Again with the linear regression analysis, the effect of the moral appeal on the extent students felt good about themselves was tested. Furthermore, the moderating effect of response history on this relationship was examined. In step 1, the
framing of the moral appeal and the extent students felt good about themselves were entered Figure 2
RH = Response history NMA = Negative moral appeal
and in step 2, moral appeal x response history was added. These results showed that in the first step, there was a main effect of the framed moral appeal on the extent students felt good about themselves (B = .71, p = .00) and there was no significant effect of the response history on the extent students felt good about themselves (B = .26, p = .25). After adding the interaction in step 2, there was still a significant main effect of the moral appeal on the extent students felt good about themselves (B = .70, p = .00), and there was no significant effect of response history on the extent students felt good about themselves (B = .26, p = .27). There was no significant moderating effect of response history on the relationship between moral appeal and the extent students felt good about themselves (B = -.06, p = .79; ∆R2 = .00, p = .79).
Based on the results of both linear regression analyses, it can be stated that a positive moral appeal leads to a greater need for self-affirmation among students. After reading the positive message, students’ belief of being a moral person was strengthened and they felt good about themselves. However, only with one item of self-affirmation, students’ belief of being a moral person, there was a moderating effect of response history.
Compensation
To examine the influence of the moral appeal on compensation behaviour, again a linear regression analysis was conduced. Moreover, the moderating effect of response history on the relationship between the moral appeal and the need for compensation was tested. In the first step of the linear regression analysis, the moral appeal and the need for compensation were entered and in the second step, moral appeal x response history of students was added. The results showed that there was no significant main effect of the moral appeal on the need for compensation (B = .35, p = .13) and there was no significant effect of response history on the need for compensation (B = -.35, p = 13). After entering the interaction in step 2, there was again no significant main effect of the framing of the moral appeal on the need for compensation (B = .36, p = .12) and still no significant effect of response history on the need for compensation (B = -.34, p = .15). There was no significant moderating effect of response history (B = .19, p = .42; ∆R2 = .02, p = .42).
Consistency
To indicate if students show consistent behaviour regarding filling out the course evaluations, two different linear regression analyses were conducted. First, the relationship between a student’s response history and current behaviour were tested, with the framed moral appeal as moderator. In step 1, the response history and current fill out behaviour were entered and in step 2, response history x moral appeal was added. The results of this first linear regression analysis showed that there was a significant main effect of response history on current fill out behaviour (B = .45, p = .00) and there was no significant effect of the moral appeal on current fill out behaviour (B = -.10, p = .51). After entering the interaction in the second step, there was again a significant main effect of response history on current fill out behaviour (B = .45, p = .01) and no significant effect of the moral appeal on current fill out behaviour (B = -.10, p = .51). There was no significant moderating effect of the moral appeal on the relationship between response history and the current fill out behaviour (B = -.03, p = .83; ∆R2 = .00, p = .83).
Second, the relationship between students’ current response behaviour and their future intentions, moderated by the moral appeal, were tested. In the first step, current fill out behaviour and future intentions to fill out the course evaluations were entered, and in the second step, current fill out behaviour x moral appeal was added. The results of the second linear regression analysis showed that there was a significant main effect of a student’s current fill out behaviour on the intention to fill out evaluations in the future (B = 1.00, p = .00) and that there was no significant effect of the framed moral appeal on future intentions (B = .27, p = .21). After entering the interaction in the second step, there was still a significant main effect of the current fill out behaviour on the future intentions (B = .98, p = .00) and no significant of the moral appeal on future intentions (B = .28, p = .23). There was no significant moderating effect of the framing of a moral appeal on the relationship between current fill out behaviour and future intentions (B = -.22, p = .31; ∆R2 = .02, p = .31).
DISCUSSION
The purpose of this study was to investigate if there is an effect of the framing of a moral appeal on a specific type of moral behaviour of students, namely filling out the course evaluations. Simultaneously, the moderating effect of students’ response history was tested. It was hypothesised that the effect of framing the moral appeals on student’s current fill out behaviour is moderated by previous history of filling out the course evaluations. More specifically, the use of a negatively framed moral appeal, in comparison to a positively framed moral appeal, will increase the fill out behaviour among students who have a high response history, but will decrease the fill out behaviour among students who have a low fill out history. This hypothesis was tested in a field experiment in which students read either a negative moral appeal or a positive moral appeal in the invitation e-mail for filling out course evaluations.
The hypothesis was tested in two different ways, namely first without participants who indicated that they did not read the moral appeal when they received the invitation e-mail, and after that, including all participants and their intentions to fill out course evaluations in the future. However, the results of both analyses showed that this hypothesis could not be supported. In both analyses, there was no significant main effect of the framing of a moral appeal on current fill out behaviour or future intentions to fill out evaluations. There was neither a significant moderating effect of response history. This suggests that it may not matter what kind of moral appeal, negative or positive, students read when they are invited to fill out the course evaluations. Their decision to fill out the course evaluation or not, does not depend on the way the message was framed in this research.
evaluations than other. Therefore, future research should try to enlarge the group of participants.
Since the hypothesis could not be supported, it should be asked to what extent the argumentation in the theoretical part aligns with the students of the University of Groningen. First of all, did the students who read the negative moral appeal really felt reactance? And does this differ from students who read the positive moral appeal? As Brehm and Brehm (1981) and Dillard and Shen (2005) found, psychological reactance is an aversive motivational state that occurs when a freedom is eliminated, and when any persuasive message arouses a motivation to reject the advocacy. Furthermore, according to Quick and Stephenson (2007) psychological reactance may result in boomerang effects, which in turn leads people doing the forbidden act (Dillard and Shen, 2005).
The results of this study showed that there was no significant main effect of the framing of a moral appeal on the extent of reactance. However, when the moderator response history was added, there was a significant effect of the framing of a moral appeal on reactance. There was also a significant moderating effect of response history in the following way: students with high response history felt more reactance when they read the negative moral appeal than when they read the positive moral appeal. For students with low response history, there was hardly any difference in the extent of reactance between the negative moral appeal condition and the positive moral appeal condition. These results confirmed the expectation that response history moderates the relationship between the moral appeal and reactance. However, as stated in the theoretical background, it was expected that especially students with low response history felt a higher extent of reactance to a negative moral appeal, compared to students with high response history.
Although the framing of a moral appeal has a significant relationship with the extent of reactance, moderated by response history, it appears that this relation does not have any influence on the decision whether or not filling out the course evaluations. However, the negative moral appeal arouses negative feelings among all students, and for students with a high response history these negative feelings became stronger when they read the negative moral appeal
reaffirm the self and try to restore moral image (Steele, 1988; Zhao and Nan, 2010). Students needed to indicate if their belief of being a moral person was strengthened and if they felt good about themselves, after reading the moral appeal. In a study of Homer and Yoon (1992) it was found that people’s attitude becomes positive when messages are framed positively and contains credible information. This statement can also be applied to this research, since a significant positive relation was found between the moral appeal and the need for self-affirmation. This means that the more positive a moral appeal is framed, the more need for self-affirmation exists among students. Students, who read the positive moral appeal, had a stronger belief of being a moral person and felt good about themselves.
Response history moderates only the relationship between the framing of a moral appeal and the believe students have of being a moral person, and not the relationship between the framing of a moral appeal and the extent students felt good about themselves. Students’ their believes of being a moral person was stronger in the positive moral appeal condition than in the negative moral appeal condition, and this effect was stronger for students with high response history than for students with low response history. It appears that this is in line with the expectations, since people have the need to obtain a positive self-concept. Students who have a high response history, already behave in a moral way which is confirmed by the positive moral appeal.
People can also reaffirm the self by compensate previous behaviour. An immoral action stimulates future moral behaviour in order to restore the positive self-concept (Sachdeva, Iliev, and Medin, 2009). In alignment with the course evaluations, it can be expected that students with high response behaviour felt a need for compensation when they read the negatively framed moral appeal. Since they behaved morally in the past, they might feel liberated to behave immorally this time, and not fill out the course evaluations. However, a negative moral appeal emphasises the immoral behaviour, which will probably lead to compensating their (potential) immoral behaviour.
Finally, to see if there is any consistency in students’ fill out behaviour, the relationship between response history and the current fill out behaviour was tested, and the relationship between current fill out behaviour and future intentions to fill out the evaluations, both moderated by the framing of the moral appeal. It was expected that people prefer to behave consistently, and it appears that this also counts for the students of this research. The results showed that students in behave consistently. However, there was no moderating effect of the framing of the moral appeal. The more a student filled out the course evaluations in the past, the more he/she filled out this time, and the more a student have the intention to fill out course evaluations in the future, regardless of the negative or positive moral appeal students read in the invitation e-mail.
Practical implications
The results of this study can be used by the University of Groningen that tries to increase the response rate to the course evaluations after every half-semester. It was suggested that, the framed moral appeals in the invitation e-mail did not influence the cooperative behaviour of students. It can therefore be concluded that there is no difference in influence of a negative moral appeal or positive moral appeal, on students’ decision filling out the course evaluations or not. However, the results showed that a negative moral appeal leads to more reactance among students. This research did not found any influence on the behaviour of students, but since the negative moral appeal leads to negative feelings, the faculty could better use a positive moral appeal in the invitation e-mail. This positive moral appeal results in a more positive self-concept and positive feelings of the students. Also, the positive moral appeal leads to more self-affirmation, which in turn diminishes immoral behaviour and improves the moral behaviour
there is more openness in what the faculty does with the results, students know what their effort is used for.
Limitations and future research
As every research has limitations, the limitations of this research also need to be acknowledged. First, the sample size is too small and therefore the statistical power of this research is limited. The participants might not represent the majority of the students, since there were only 34 participants, and there are thousands of students studying at the faculty of Economics and Business. More participants would help to make the results more reliable, so future research should try to enlarge the group of participants. Moreover, there was a relatively large difference in the number of students who completed the questionnaire about the negative moral appeal and students who completed the questionnaire about the positive moral appeal. Although both conditions were equally divided between 440 students, the students of the negative moral appeal condition filled out the questionnaire more than students of the positive moral appeal condition. However, this difference is not that large anymore, when you compare this to the total of 440 students who received the questionnaires, instead of the 34 students who filled out the questionnaire.
Second, in this study no accurate data is used to test the current response behaviour and response history. Students were asked to indicate by themselves if they filled out the evaluations this time, and to what extent they filled out the evaluations in history. This might bias the results, due to the fact that when first year students have indicated that they ‘never’ filled out the evaluations, they missed only two times until semester 2.1. They can still change their behaviour in the forthcoming years. A last year student who indicated that he/she ‘never’ filled out the course evaluations, did that obviously more times than the first year student. Future research could use the accurate data, with the number of times a students filled out course evaluations.
does not matter anymore for them since they will graduate soon, and therefore the moral appeal may have less influence.
The variables ‘current fill out behaviour’ and ‘response history’ were each measured by only one question. For current response behaviour students needed to answer the question whether they fill out the course evaluations this time with yes, partly, or no. For response history, students needed to indicate on a 5-point scale (never – always) to what extent they filled out the evaluations in the past. A shortcoming of this is that it may be biased by social desirability answers, since participants might felt that they had to answer in a politically correct way. The participants needed to fill out their student number in the beginning of the questionnaire, and it could be that they thought this data would be used for other things too, beside this research. Despite the fact that students read that their anonymity was guaranteed, future research may need to make this more clear and repeat this message of guaranteed anonymity during the questionnaire.
Conclusion
Allingham, M., and Sandmo, A. (1972). Income tax evasion: A theoretical analysis. Journal
of Public Economics. 1: 323-338
Batra, R., and Ray, M.L. (1986). Affective responses mediating acceptance of advertising.
Journal of Consumer Research. 13: 234-249
Baumeister, R.F., Bratslavsky, E., Finkenauer, C., and Vohs, K.D. (2001). Bad is stronger than good. Reviews of General Psychology. 5: 323-370
Bénabou, R., and Tirole, J. (2006). Incentives and prosocial behaviour. American Economic
Review. 96 (5): 1652-1678
Beniers, C.J.M. (2009). Cultural differences in television advertising (3)
http://blog.beniers-consultancy.de/archives/tag/moral-appeals [29-02-2012]
Bosco, L., and Mittone, L. (1997). Tax evasion and moral constraints: Some experimental evidence. Kyklos. 50: 297-324
Braun, K.A., Geath, G.J., and Levin, I.P. (1997). Framing effects with differential impact: The role of attribute salience. Advances in Consumer Research. 24: 405-411
Brehm, S. S. B., and Brehm, J. W. (1981). Psychological reactance: A theory of freedom
and control. New York: Academic Press.
Carlsmith, J.M., and Gross, A.E. (1969). Some effects of guilt on compliance. Journal of
Personality and Social Psychology. 11: 232-239
Cohen, G.L., Aronson, J., and Steele, C.M. (2000). When beliefs yield to evidence: Reducing biased evaluations by affirming the self. Personality and Social Psychology
Bulletin. 26: 1151-1164
Dawes, R.M., and Messick, D.M. (2000). Social Dilemmas. International Journal of
Psychology. 35 (2): 111-116
Dillard, J.P., and Shen, L. (2005). On the nature of reactance and its role in persuasive health communication. Communication Monographs. 72: 144-168
Falk, A., and Zimmermann, F. (2011). Preferences for consistency. CESifo Working Paper
Series: 3528
Feenstra, H., and Le Rutte, M. Handboek kwaliteitszorg onderwijs FEB. Een integraal kwaliteitszorgsysteem voor het onderwijs binnen de faculteit Economie en Bedrijfskunde. Versie 2008
Kahneman, D. and Tversky, A. (1979). Prospect theory: an analysis of decision under risk.
Econometrica. 47 (2): 263-292
Khan, U., and Dhar, R. (2006). Licensing effect in consumer choice. Journal of Marketing
Research. 43 (2): 259-266
Levin, I.P., Schneider, S.L., and Gaeth, G.J. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behaviour and Human
Decision Processes. 76: 149-188
Monin, B., and Miller, D.T. (2001). Moral credentials and the expression of prejudice.
Journal of Personality and Social Psychology. 81 (1): 33-43
Quick, B. L., & Stephenson, M. T. (2007). The Reactance Restoration Scale (RRS): A measure of direct and indirect restoration. Communication Research Reports, (24): 131-138 Sachdeva, S., Iliev, R., and Medin, D.L. (2009). Sinning saints and saintly sinners: The paradox of moral self-regulation. Psychological Science. 20 (4): 523-528
Schwartz, R., and Orleans, S. (1967). On legal sanctions. University of Chicago Law
Review. 34: 274-300
Schwartz, S.H. (1977). Normative influences on altruism. In L. Berkowitz (Ed.), Advances
in experimental social psychology. 10: 221-279. New York: Academic Press
Shamaskin, A.M., Mikels, J.A., and Reed, A.E. (2010). Getting the message across: Age differences in the positive and negative framing of health care messages. Psychology and
Aging. 25 (3): 746-751
Steele, C.M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.). Advances in Experimental Social Psychology. 261-302. New York: Academic Press
Webley, P., Cole, M., and Eidjar, O.P. (2001). The prediction of self-reported and hypothetical tax-evasion: Evidence from England, France and Norway. Journal of
Economic Psychology. 22: 141-155
Wenzel, M. (2004). An analysis of norm processes in tax compliance. Journal of Economic
Psychology. 25 (2): 213-228
Wenzel, M. (2007). The multiplicity of taxpayer identities and their implications for tax ethics. Law & Policy. 29 (1): 31-50
Yu, H. (2012). Producing more persuasive anti-violence messages for college students: Testing the effects of framing, information sources, and positive/negative fact appeal.
Journal of Interpersonal Violence. 27 (9): p. 1631-1654
