Master’s Thesis Ludo Penning de Vries 10712429 Graduate School of Communication Master’s programme Communication Science S.C.M. Welten
Food for thought
Examining the effects of involvement and
pride- and guilt-appeals in persuasive
Index
Introduction 4
Theoretical background 6
Methods 11
Results 21
Abstract
The aim of this study was to provide insight into the effects of pride- and guilt-appeals on organic food purchase intention of highly and lowly involved consumers. This was done through a
between-subjects experimental design with three conditions: pride, guilt, and a control condition. Involvement with organic food served as a moderator. Participants were exposed to one of three pretested advertisements and indicated their purchase intention for organic food, attitude towards the advertisement, and attitude towards buying organic food. Actual behavior was measured through a prize selection item, in which participants chose either an organic or a non-organic chocolate bar. Significant effects were found from the level of participants’ involvement with organic food on the level of their purchase intention, their attitude towards the advertisement and to buying organic food. More involved participants have a higher purchase intention, a more positive attitude towards the advertisement, and a more positive attitude towards buying organic food than less involved participants. The emotion condition a participant was in affected the attitude they held towards the advertisement: participants liked the guilt advertisement significantly less than the pride and control advertisements. No significant interaction effect of emotional condition and level of involvement on purchase intention, attitude towards the advertisement, attitude towards buying organic, or prize selection was found. Conclusions are drawn and implications for future research as well as practice are discussed.
Introduction Main problem
How to advertise organic food? This is an interesting question that is, considering current sales trends, likely to be on the minds of many a marketer. One way to stimulate buying behavior for organic products is to use emotional appeals in advertising. Onwezen, Antonides and Bartels (2013) researched consumers’ decision-making process regarding environmentally friendly behaviors. They explored the effects of anticipated emotions, specifically pride and guilt, on behaviors such as purchasing environmentally friendly food and using “green” means of transport. While this research provides useful insight into emotions as anticipated by the receiver of a
message, the researchers take a solely observing role. Actually felt emotions as induced by a persuasive message or advertisement are not researched and participants were not exposed to a persuasive message evoking pride or guilt. Taking their research a step further and looking into the effects of exposure to emotional appeals in advertising and thereby taking an inducing role, rather than an observing one, could generate valuable information for advertisers of organic food.
In the research by Onwezen et al. (2013) it was not specified which emotion, pride or guilt, was most effective in creating a certain behavioral intention. Both emotions had an effect on intention, with beta regressive weights for anticipated pride and guilt on a person’s intention of .242 and . 274, respectively, but to increase efficiency in targeting consumers, more specific information about what emotional strategy to employ to reach and influence the target audience is necessary.
A factor that has not been taken into account in many studies about pride, guilt and behavior, including that of Onwezen et al. (2013), is prior involvement with organic food. It may be expected that a consumer who frequents an organic-only supermarket will respond differently to an
advertisement than a consumer who does not value or cannot afford “green” products. For advertisers, it would be insightful to know what influence involvement can have on purchase behavior, so they can design and adapt advertisements accordingly. Knowledge about how to best target people who are not very involved yet can be very lucrative.
Organic food is food that has not been treated with synthetic fertilizers, synthetic pesticides or sewage sludge-based fertilizers. It is not exposed to radiation, and in the case of livestock, has not
been given growth hormones or preventative antibiotics. It must be certified under the National Organic Program (NOP). This type of food is in increasing demand, generating 20 billion US dollars in sales in the United States in 2007, compared to 3.6 billion dollars in 1997 (Food
Marketing Institute, 2008). In The Netherlands, the growth in the past year of all organic groceries was 5.4 per cent, generating 980 million euros in sales (monitor duurzaam voedsel, 2013). This rise is a positive development, as according to Smith (1993), several organic foods contain over 90 per cent more nutritional elements than “regular” food, while non-organic food contains about 22 per cent more bad elements such as aluminum, mercury and lead. Research by Pummeling et al. (2008) showed that when children consumed organic milk instead of conventional milk, levels of eczema, asthma and wheezing decreased significantly. By increasing consumption of organic foods, the occurrence of these diseases could be reduced.
Although the benefits are clear, still only a small portion of all the consumed foods is organic. In The Netherlands, only 2.4 per cent of all groceries bought is organic. Moreover, while the total percentage of organic purchases is rising, some categories are falling steeply. Last year, the organic meat sector experienced a drop of 13.6 per cent in sales when compared to the year before, while organic dairy was consumed 1.6 per cent less than the year before (monitor
duurzaam voedsel 2013). An explanation for these low figures could be the uncertainty about how to effectively advertise organic foods. More effective advertisements could aid consumers in making a transition to organic food and boost these figures.
In the current study, the findings by Onwezen et al. (2013) will be empirically tested to provide insight in the effects of pride and guilt induced by advertisements. The aim of this study is to provide insight into the effects of pride- and guilt-appeals on organic food purchase intention of highly and lowly involved consumers. In order to do this, the following research question will be answered: “To what extent do pride- and guilt-appeals influence the purchase intention for organic products for highly and lowly involved consumers?”
Theoretical background Pride and guilt
Tracy and Robins (2004) researched emotions and made a distinction between basic emotions such as fear and joy, and self-conscious emotions such as shame, pride and guilt. The authors note five differences between the types of emotions. The most important difference between basic and conscious emotions is that the latter require the presence of awareness and self-representations, which means that people need to be able to think about how they present themselves in order to feel these emotions. Other differences include that they develop later in childhood, do not have universally recognized facial expressions and that they are more complex than basic emotions. This makes them slightly more difficult to research and to use in advertising, one reason being that they differ from person to person. Unlike advertisements using basic emotions like fear or happiness, where a happy image could simply be shown, advertisements using self-conscious emotions must be targeted towards the receiver of the advertisement in order to address their self-awareness and stimulate thought about self-representation.
Two self-conscious emotions that are often associated with environmental behavior are feelings of pride and guilt. These emotions have to do with following or not following a personal standard one sets for him- or herself (Onwezen et al., 2013) and can, according to Antonetti and Maklan (2014), lead to increased intentions to buy ethical products. To provide insight into the way in which these emotions are related to environmentally friendly behavior, Onwezen et al. (2013) performed a literature study and a survey. They investigated the relation between pride, guilt and the Norm Activation Model (NAM). This model was developed by Schwartz (1977) and adapted as a
mediator model by De Groot and Steg (2009). It holds that the moral obligations an individual feels are determined by two factors: the awareness that his or her actions have specific consequences on one hand, and the responsibility he or she feels for performing the behavior in question on the other. A graphical representation of the model can be seen in figure 1.
Figure 1
Graphical representation of the NAM taken from Onwezen et al. (2013).
Onwezen et al. (2013) expanded this model with the levels of anticipated pride and guilt felt by consumers. Seven of these extended NAM versions were tested, each of which featured pride and guilt in a different relation to behavior and/or personal norm. The model with the best fit was then combined with the Theory of Planned Behavior (TPB)-model by Ajzen (1991), improving the model fit even further. The Theory of Planned Behavior holds that behaviors can be best predicted by behavioral intentions. These intentions are in turn determined by a person’s attitude toward performing the behavior, social norms and the level of control the person has to execute the behavior. The integrated NAM-TPB model shows that a person’s intention is indeed the best predictor of their behavior and that anticipated pride and anticipated guilt both were significant determinants of behavioral intention, approximately of equal strength. However, it is unknown whether these effects will hold in an empirical testing by inducing these emotions using a persuasive message.
Harth, Leach and Kessler (2013) researched the effect pride, guilt and anger have on people’s intentions to repair or protect the environment. Participants were asked to read two types of newspaper articles, one about harm done to the environment by people’s country and one about steps taken by their country to protect the environment. Participants then indicated what emotions they felt. They found that when people feel responsible for environmental damage, they feel guilt, and when people feel responsible for environmental protection, they feel pride. In the guilt
condition, people felt a greater intention to repair the damage done, while in the pride condition, people favored intentions of environmental protection. An example of an intention to repair the damage done is: “It is our duty to take care of the environmental problems we have caused” (Harth et al., 2013, p. 23). An example of an intention to protect the environment is: “We should only invest in companies that promote environmental technological progress” (Harth et al., 2013, p. 23). In the context of consumer buying behavior, buying pro-environmental products such as organic food does not directly restore harm already inflicted to the environment. However, it will prevent more harm from being done. For example, switching to organically produced vegetables supports an industry that does not pollute the soil with synthetic fertilizers. Therefore, this could be
considered a protective behavior. Both pride and guilt can be related to pro-environmental
behavior, but because pride-feelings are associated with protective intentions, while guilt feelings lead to repairing intentions (Harth et al, 2013), it is expected that:
H1: Pride appeals will lead to a) a greater purchase intention for organic food, b) a more positive ad attitude, and c) a more positive attitude towards buying organic food than guilt appeals.
Involvement with organic food
One can look at many factors of influence when considering effectiveness of emotional appeals for buying organic food. For example, level of income or the distance to the closest shop stocking organic food could determine one’s purchase intention for organic foods. Another factor of influence could be the interest people have and value they attach to organically produced food. It may be expected that people who are interested in organic food and attach great value to it have a different level of purchase intention and hold different attitudes towards buying it than people who are not interested in organic food at all. In the current research, the interest people have and value they attach to organically produced food will be referred to as their level of involvement. Meijers and Van Dam (2012) suggest exploring in what way information about involvement may be used in order to make sustainable products more successful. This information could be of great value for advertisers of sustainable products, allowing them to target individuals at specific levels of involvement. This in turn could increase popularity of sustainable consumption, the benefits of which have been discussed.
Much research does not take a person’s level of involvement into account when looking at organic food buying behavior. In a study performed by Schifferstein and Oude Ophuis (1998) determinants of organic food consumption in The Netherlands were researched. They collected part of their sample from a store specialized in health and natural foods and compared the results to those of a nationwide representative sample. It must be noted that customers of a specialized health and nature store are likely to be biased about the source of the products they choose, as something has led them to prefer these stores over “regular” supermarkets. Therefore, they could be considered highly involved consumers. For marketers, determinants of organic food buying behavior for average consumers (e.g., people who do not buy their groceries at specialized health stores) would be of more informative value as these consumers form the majority of the population. Effectively targeting a larger audience would mean a greater impact.
However, some research about involvement and organic food has been conducted. Zanoli and Naspetti (2002) looked at the effects of consumer knowledge and perceptions of organic food on purchasing behavior. It was found that people who buy organic food regularly have a higher level of involvement with it than people who purchase organic food only occasionally. Consistent with this result are the findings in Meijers and Van Dam (2012), who found a link between an
individual’s sustainability concerns and sustainable behavior (such as the purchasing of sustainable products). An individual with more concerns for sustainability attaches more importance to it than an individual who does not think twice about it, and therefore could be considered a more involved individual. It was found that product-specific concerns proved a better predictor for sustainable consumption than general sustainability concerns. For example, concerns about the origin of meat in a supermarket would be a better predictor for its purchase than
concerns about environmentally friendly ways of transport. This indicates a relationship between involvement with these products and their consumption.
Because it was found that regular consumers of organic foods have a higher level of involvement (Zanoli & Naspetti, 2002) and Meijers and Van Dam (2012) found that people with sustainability concerns are more likely to purchase environmentally friendly products, it is expected that:
H2: People who are more involved with sustainable products will have a) a greater purchase intention for organic food, b) a more positive ad attitude, and c) a more positive attitude towards buying organic food than people who are less involved.
Interaction of pride and guilt with involvement
As discussed, pride- and guilt-appeals can affect people’s purchase intention positively. However, an individual’s level of involvement in this relationship has not been widely researched yet. The current research will be focused on exploring the relationship between pride- and
guilt-appeals, and the purchase intention of people with high and low levels of involvement. Knowledge about this could provide advertisers of sustainable products insight into which appeal is most effective when targeting consumers at different levels of involvement. For the scientific field of persuasive communication, the underexposed intersection between involvement and pride- and guilt appeals might contain valuable information about persuasion techniques that people can use
and may or not already use. The discovery of an interaction between involvement and pride- and guilt appeals may shed new light on results of prior research and have implications for future research.
For example, it could be argued that Harth et al. (2013), as well as Antonetti and Maklan (2014) unintentionally created a high involvement sample by asking people to read texts about the subject. Had the participants attached relatively low importance to sustainability and organic food before, reading about the subject may have increased their involvement. Results from Harth et al. (2013) show that pride-appeals generated greater protecting intentions (e.g., supporting pro-environmental industries and buying sustainable products) than did guilt-appeals.
In the case of the supposed high involvement, results by Harth et al. (2013) suggest that pride-appeals have a greater impact on behavioral intentions of highly involved people. Also assuming a highly involved sample, results by Antonetti and Maklan (2014) suggest that pride appeals in the high involvement condition correspond with a greater purchase intention than guilt appeals in a high involvement condition. Therefore, it is expected that:
H3: When involvement is high, it may be expected that the effect that pride appeals will generate a) a greater purchase intention for organic food, b) a more positive positive ad attitude, and c) a more positive attitude towards buying organic than guilt appeals will be stronger than when involvement is low (H3).
This research will look into the relation between pride- and guilt-appeals and purchase intention, attitude towards the advertisement, attitude towards buying organic food, as well as actual
behavior, while looking at an individual’s involvement with the subject. A schematic representation or conceptual model of this research is provided in figure 2.
Methods
This research is a between-subjects experimental design with three conditions: pride, guilt, and neutral (control). Involvement with organic products served as a moderator. Participants were approached in the library of the University of Amsterdam and recruited on a Facebook page designed for participant recruiting (“Respondenten gezocht”). As a stimulus, participants who were approached in the library were offered a biscuit in return for their participation. This did not
significantly affect the results (all r’s (164) ≤ .075, all p’s ≥ .282).
Pretest
In order to select the best materials for the main study a pretest was run. In this pretest, 92 people aged 18 or over took part (Mage = 24, SDage = 6.63). Of this sample, 59% was female. 92% of the
sample has attended or currently attends university. Three participants (3.3 %) indicated to be parents. For the pretest, three sets of advertisements were created, each set consisting of three versions of the same poster: a pride version, a guilt version and a neutral (control) version. The advertisements can be found in Appendix A. The three sets of advertisements focused on the most important reasons for people to buy organic food (Schifferstein & Oude Ophuis, 1998):
wholesomeness (nutritional value), chemicals or the absence thereof and environment friendliness.
Purchasing organic products is not synonymous with environmentally friendly behavior. For organic beef, for instance, the animal’s welfare is considered and the cow is not preventatively medicated, yet more food is needed for the cow that lives longer and more carbon dioxide is produced
(Meijers, 2014). In this research however, organic food buying behavior is considered an H1
H2 H3
Figure 2
Conceptual model of the current study
IV DV’s
Emotional appeals: pride versus guilt
Purchase intention
Attitude towards the advertisement Attitude towards buying organic Actual behaviour
Low versus high involvement with organic products
environmentally friendly behavior because consumers perceive it as such (Sparks & Shepherd, 1992; Tacken, de Winter, & Wertheim-Heck, 2007; Thøgersen & Ölander, 2003).
The first set of advertisements featured a young girl taking a bite out of an apple. The title for the pride version of the advertisement read "Thank you for buying organic and protecting her from 400+ chemicals," whereas the title for the guilt condition read "You're poisoning her with 400+ chemicals found in non-organic food." The control condition featured no title. This set of advertisements will be called the girl-advertisement from now on. The second set of
advertisements featured a young boy picking an apple from a supermarket shelf. The pride title read "Thank you for buying organic and giving him 22 per cent extra nutritional value," whereas the guilt advertisement read "By not buying organic you are denying him 22 per cent extra nutritional value." The control condition did not have a title. This set of advertisements will henceforth be referred to as the boy-advertisement. The third set of advertisements featured a chicken. In the pride condition, the chicken eats grain from a set of hands and the title read "Thank you for supporting the environment and animal welfare by buying organic food." In the guilt condition, the helping hands feeding the chicken were removed from the image and the title read "By not buying organic you are supporting an industry harming animals and the environment.” The control
condition did not feature a title, but did show the hands feeding the chicken. This set of advertisements will be referred to as the chicken-advertisement.
All advertisements featured a direct appeal: “Buy organic food.” To increase the advertisements’ credibility, they all featured a logo of a made-up company, Walkers Organics. This fictitious brand was chosen because research has shown that logo’s for commercial organizations are more persuasive than non-profit organization logo’s such that of the European Union. Within each version, as much ad characteristics as possible was kept constant across the conditions. This includes background image, word layout, logo, logo placement and punchline.
Participants viewed three advertisements from the same condition (either pride, guilt or control). A minimum exposure of five seconds was ensured by means of a timing mechanism embedded in the questionnaire. Participants first indicated which of five emotions (pride, happiness, guilt, sadness, anger) would best describe their reaction to the advertisement. These emotions were chosen as options because they were likely emotions a participant may feel in response to the
advertisement. Participants then indicated how strongly they felt each of the five emotions on a seven-point Likert scale, 1 = “Not at all”, 7 = “A lot.” Participants also indicated whether they found the advertisement professional, credible and trustworthy using a three-item seven-point semantic scale ranging from unprofessional to professional, not credible to credible and untrustworthy to trustworthy, respectively. Because of the guilt-inducing nature of the material, a debriefing was necessary. Research has shown that humorous positive stimuli work better in lightening the mood than non-humorous positive stimuli (Strick, Holland, Van Baaren & Van Knippenberg, 2009). For this reason, a funny short video clip (30 seconds) was shown at the end of the pretest in which participants saw the way animals would shop if they had supermarkets. Additionally, participants were given the opportunity to leave comments in a text box at the end of the survey. This item did not reveal any negative effect caused by the research.
Upon running the pretest, first, the advertisements’ professionalism, credibility and trustworthiness are checked. As Table 1 shows, all advertisements were found to be of sufficient professionalism, credibility and trustworthiness, scoring above midpoint. Two significant differences were found: the reported credibility and trustworthiness of the guilt condition of the boy-advertisement were
significantly lower than that of the control condition (F(2, 89) = 5.03, p = .008), (F(2, 89) = 5.24, p = .007), respectively.
Table 1
Professionalism, credibility and trustworthiness for all advertisements across conditions
Advertisement Condition M (SD)
Girl Professional Pride 4.27 (1.34) a
Guilt 4.47 (1.66) a Control 4.56 (1.37) a Credible Pride 4.20 (1.52) a Guilt 4.27 (1.66) a Control 4.34 (1.73) a Trustworthy Pride 4.30 (1.34) a Guilt 4.07 (1.55) a Control 4.69 (1.64) a
Boy Professional Pride 4.93 (1.31) a Guilt 4.50 (1.61) a Control 5.31 (1.18) a Credible Pride 4.53 (1.41) a, b Guilt 4.17 (1.32) a Control 5.22 (1.26) b Trustworthy Pride 4.80 (1.16) a, b Guilt 4.20 (1.35) a Control 5.22 (1.21) b
Chicken Professional Pride 5.17 (1.44) a
Guilt 5.03 (1.22) a Control 4.88 (1.43) a Credible Pride 5.27 (1.51) a Guilt 4.83 (1.32) a Control 4.72 (1.59) a Trustworthy Pride 5.20 (1.27) a Guilt 4.70 (1.29) a Control 4.88 (1.34) a
Note. Means per advertisement with different letters differ
Next, to select the set of advertisements most effective in evoking pride and guilt, the mean scores of emotions reported by the participants are compared for the pride and guilt advertisement, respectively. In Tables 2, 3, and 4, the advertisement most effective in evoking the targeted emotions (shown in bold) is the chicken-advertisement. When participants had to choose one emotion they felt when looking at the ad, this advertisement had the most “proud” responses in the pride condition and the most “guilty” responses in the guilt condition.
In Table 5, it can be seen that the girl-advertisement evoked a low amount of pride that did not differ significantly across the conditions. The pride condition did evoke significantly more happiness than the guilt condition. The guilt condition evoked significantly more guilt than the pride condition, but not significantly more than the control condition. It also evoked more sadness and anger than the other conditions.
Table 4
Percentages for forced-choice emotion for chicken-ad
Pride (%) Guilt (%) Control (%) Proud 33.3 0 15.6 Guilty 3.3 46.7 21.9 Sad 6.7 30 18.8 Angry 0 20 3.1 Happy 56.7 3.3 40.6 Table 3
Percentages for forced-choice emotion for boy-ad
Pride (%) Guilt (%) Control (%) Proud 13.3 0 0 Guilty 6.7 36.7 0 Sad 6.7 40 3.1 Angry 3.3 13.3 0 Happy 70 10 96.9 Table 2
Percentages for forced-choice emotion for girl-ad
Pride (%) Guilt (%) Control (%) Proud 6.7 0 3.1 Guilty 26.7 26.7 18.8 Sad 13.3 43.3 25 Angry 16.7 23.3 3.1 Happy 36.7 6.7 50
Table 6 shows that in the boy-advertisement, the pride condition evoked significantly more pride than the guilt condition, but not more than the control condition. The guilt condition evoked
significantly more guilt than the other conditions. Like in the girl-advertisement, the pride condition evoked more happiness than the guilt condition. Also like in the girl-advertisement, the guilt condition evoked more sadness and anger than the other conditions.
In Table 7 it can be seen that the pride condition in the chicken-advertisement did evoke
significantly more pride than the other conditions, in contrast to the first two advertisements. The guilt condition also evoked more guilt than the other conditions. Again, the pride condition evoked more happiness than the guilt condition, and the guilt condition evoked more sadness and anger than the other conditions.
Table 5
MANOVA results for emotions towards the girl-advertisement
Pride Guilt Control F(2, 89) p η2
M (SD) M (SD) M (SD) Proud 2.40 (1.81) a 1.63 (0.89) a 2.28 (1.53) a 2.40 .097 0.05 Happy 3.40 (1.87) a 2.20 (1.52) b 3.50 (1.90) a 5.06 .008 0.10 Guilty 3.20 (1.86) a, b 3.90 (1.37) b 2.72 (1.76) a 3.85 .025 0.08 Sad 2.77 (1.76) a 4.73 (1.34) b 2.44 (1.44) a 20.39 < .001 0.31 Angry 2.57 (1.61) a 4.07 (1.28) b 1.56 (1.05) c 27.75 < .001 0.38
Note. Means per row with different letters differ significantly with p < .05 in Tukey Post Hoc tests; bold
means targeted variable.
Table 6
MANOVA results for emotions towards the boy-advertisement
Pride Guilt Control F(2, 89) p η2
M (SD) M (SD) M (SD) Proud 3.33 (1.67) a 1.80 (1.21) b 3.09 (1.67) a 8.69 < .001 0.16 Happy 4.27 (1.53) a 2.07 (1.26) b 4.72 (1.63) a 27.87 < .001 0.39 Guilty 2.10 (1.30) a 4.00 (1.58) b 1.72 (1.05) a 26.14 < .001 0.37 Sad 2.00 (1.39) a 4.03 (1.43) b 1.75 (1.24) a 26.08 < .001 0.37 Angry 1.87 (1.31) a 3.57 (1.28) b 1.47 (.95) a 27.05 < .001 0.38
Note. Means per row with different letters differ significantly with p < .05 in Tukey Post Hoc tests;
After comparing the results, the advertisement that was selected for the main study was the chicken-advertisement. It was the only advertisement to significantly evoke a feeling of pride significantly more in the pride condition than both other conditions, while evoking guilt more in the guilt condition than the other conditions. Additionally, the means for the targeted emotions are closest to each other, meaning the intensity of the emotion evoked is the most similar of the three advertisements.
The chicken-advertisement was shown with a minor adaptation. Although pride and happiness are strongly correlated (r(92) = 0.8, p < .001), it is expected that by placing more emphasis on the gratitude for buying organic food, the recipient of the message feels more pride. This emphasis was achieved by slightly enlarging the words “Thank you” in the advertisement. This adaptation can be seen next to the original in Appendix A. The adaptation was made in an attempt to create more equal conditions. Otherwise, the advertisements were shown unaltered.
Main test
For the main test, 164 participants (89 males, 75 females) aged 18 years or over (Mage = 27, SDage
= 9.70) took part. Of the participants, 86% attended or attends a university, while 9.1% attended or attends a university of applied sciences (HBO). The high education level of the sample did not affect the dependent variables (all r’s(164) ≤ .030, all p’s ≥ .335).
Table 7
MANOVA results for emotions towards the chicken-advertisement
Pride Guilt Control F(2, 89) p η2
M (SD) M (SD) M (SD) Proud 4.17 (1.60) a 1.67 (1.24) c 2.81 (1.62) b 20.94 < .001 0.32 Happy 4.30 (1.42) a 1.80 (1.32) b 3.53 (1.54) a 23.96 < .001 0.35 Guilty 2.30 (1.42) a 4.63 (1.97) b 3.06 (1.68) a 14.61 < .001 0.25 Sad 2.00 (1.26) a 4.30 (1.73) b 2.94 (1.78) a 15.52 < .001 0.26 Angry 1.57 (.94) a 3.93 (1.53) c 2.44 (1.66) b 21.38 < .001 0.33
Note. Means per row with different letters differ significantly with p < .05 in Tukey Post Hoc tests;
Involvement. Interest people have in, and value they attach to organic food was measured using an organic involvement questionnaire adapted from Tarkiainen and Sundqvist (2009). The items can be found in Appendix B. In the research by Tarkiainen and Sundqvist (2009), four scales were used to measure involvement with organic food, with a total of twelve items. Self-perceived
environmental friendliness was measured using a five-item seven-point Likert scale. An example of this would be: “With my own consumption habits I can influence the state of the
environment.” (Cronbach’s α = .835). The importance participants give to their health was measured using a three-item seven-point Likert scale. An example would be: “I choose food carefully to ensure good health.” (Cronbach’s α = .853). Participants’ attitude towards organic products was measured using a three-item seven-point Likert scale. An example of this is: “I am interested in organic products.” (Cronbach’s α = .922). One seven-point Likert item measured the current purchase behavior of organic products, with answers ranging from “Never” to “Always.” Although all scales had a high reliability, as participants’ attitude towards organic food best reflects involvement with organic food as defined in the present research, this scale was chosen to
measure involvement. A principal component analysis (PCA) was performed on the three items measuring involvement. As Table 8 shows, the items strongly loaded on one component. Reliability for this scale was high (Cronbach’s α = .922) and could not be enhanced by omission of one or more of the items. The items were combined into one variable called involvement. A median split was performed on this variable, resulting in a low involvement group (n = 84) and a high
involvement group (n = 80).
Table 8
PCA results for items measuring involvement.
Item Component
1 “I think that organic products are very meaningful.” 0.911
“I am interested in organic products.” 0.947
The participants were then exposed to one of the previously selected advertisements. They were asked to look at it and continue with the survey when ready. Again, a minimum exposure of five seconds was ensured using a timing mechanism in the questionnaire. On the next page, participants saw the same advertisement again, this time with a number of questions displayed directly under the image.
Attitude towards the ad. Firstly, they were asked to indicate their attitude towards the
advertisement using a three-item seven-point semantic scale (Coulter & Pinto, 1995). Reliability for this scale was very high (Cronbach’s α = .940). As Table 9 shows, a PCA revealed that all items in the scale loaded strongly onto one component. For this reason, the items were combined into a single variable called attitude towards the advertisement.
Attitude towards buying organic food. Next, participants’ attitude towards buying organic food was measured using a five-item seven-point semantic scale taken from Knabe (2012) (Cronbach’s α = .925). As Table 10 shows a PCA revealed that the items loaded strongly onto one component. The items measured the same variable: attitude towards buying organic food. Therefore, the items were combined into a single variable.
Table 9
PCA results for items measuring attitude towards the advertisement
Item Component
1
Favorable / unfavorable 0.927
Negative / positive 0.951
Emotions. Emotions participants felt when looking at the advertisement were measured using the same items as in the pretest. Participants indicated which of five emotions best described how they felt (pride, happiness, guilt, sad, anger). In the pride condition, participants indicated happiness most (50%), in the guilt condition, participants indicated guilt most (36.4%), and in the control condition, happiness was also chosen by most participants (58.5%). Subsequently they indicated how strongly they felt each of these emotions. Table 11 shows that the pride advertisement evoked significantly more pride than the guilt condition, but not more than the control condition. The guilt advertisement evoked significantly more guilt than the control condition, but not more than the pride condition.
Table 10
PCA results for items measuring attitude towards buying organic food
Item Component 1 Harmful / beneficial 0.885 Unpleasant / pleasant 0.908 Bad / good 0.910 Worthless / valuable 0.866 Unenjoyable / enjoyable 0.825 Table 11
MANOVA results for emotions towards the advertisements in the main test
Pride Guilt Control F(2, 161) p η2
M (SD) M (SD) M (SD) Proud 2.89 (1.72) a 1.98 (1.53) b 2.85 (1.85) a 4.98 .008 0.58 Happy 3.66 (1.70) a 2.24 (1.53) b 4.23 (1.85) a 19.92 < .001 0.20 Guilty 2.93 (1.70) a, b 3.65 (1.88) a 2.47 (1.78) b 6.02 .003 0.07 Sad 2.54 (1.55) a 3.87 (1.75) b 2.36 (1.57) a 14.15 < .001 0.15 Angry 1.89 (1.29) a 3.24 (1.83) b 1.64 (1.02) a 19.85 < .001 0.20
Note. Means per row with different letters differ significantly with p < .05 in Tukey Post Hoc tests; bold
Purchase intention. Upon continuing, participants indicated how likely they would be to purchase organic food, granted its availability in the store. This was done using a single-item seven-point semantic scale (Hsu & Chen, 2014). Answer options ranged from “definitely avoid it” to “definitely buy it.”
Prize selection. Additionally in this survey, actual choosing behavior of organic food was measured through a prize selection item; participants could indicate their choice of prize that was given to three randomly selected participants. The options were two similar bars of chocolate; one being organic, the other being conventional. The bars can be seen in Appendix C. This item allows for measuring actual behavior instead of only behavioral intention. Contact information for the prize draw was collected in a separate survey as to ensure the anonymity of the responses. A random number generator was used to select three email addresses to which an email was sent informing the participants they had won the raffle and asking for an address to ship the prize to.
Debriefing. A similar debriefing method as in the pretest, using a humorous element, was used to ensure that the survey, and in particular the guilt appeal, does not cause lasting negative
consequences. For brevity reasons, participants viewed a comic instead of a video. The comic showed the two options consumers have: a popular and affordable, but unhealthy product, or a natural and safe, but expensive one, showing that the consumer is always the deceived one. The comic was shown at the end of the survey prevent it from influencing the results.
Results
Purchase intention
To see the effects of involvement with organic food, the emotional condition the participants were in and the interaction of these variables on participants’ purchase intention for organic food, a two-way analysis of variance was run with purchase intention as dependent variable and condition and involvement as independent variables. A significant result was found: people who are highly involved (Minvolved = 5.58, SDinvolved = 1.02) had a significantly higher purchase intention for organic
food than people who are lowly involved (Muninvolved = 4.33, SDuninvolved = 1.01) (F(1, 158) = 59.35, p
involved, was above midpoint, meaning that lowly involved people do not intend to avoid organic food. No significant effect was found from the condition participants were in (F(2, 158) = .215, p = . 806, η2 = .003), and no interaction effect was found (F(2, 158) = 1.408, p = .248, η2 = .018).
Attitude towards the advertisement
To examine the effects of the level of involvement with organic food, the emotional condition the participants were in and the interaction of these variables on the attitude participants had towards the advertisement, a two-way analysis of variance was run with ad attitude as dependent variable and condition and involvement as independent variables. A significant effect was found from involvement (F(1, 158) = 5.661, p = .019, η2 = .035). Participants who are involved with organic
food liked the advertisements more than participants who are less involved (Minvolved = 4.82,
SDinvolved = 1.44, and Muninvolved = 4.42, SDuninvolved = 1.24, respectively). A significant effect was
also found from the condition participants were in (F(2, 158) = 11.631, p <.001, η2 = .128).
Participants in the guilt condition (Mguilt = 3.98, SDguilt = 1.64) liked the advertisement significantly
less than participants in the other conditions (Mpride = 4.90, SDpride = 0.94, and Mcontrol = 4.97
SDcontrol = 1.18). No significant interaction effect was found from involvement and emotional
condition on the attitude towards the advertisement (F(2, 158) = .653, p = .522, η2 = .008).
Attitude towards buying organic
Participants also indicated their attitude towards buying organic food. A two-way analysis of variance was performed to measure the effects of emotional condition, involvement and
interaction effects. A significant effect was found from involvement (F(1, 158) = 67.386, p < .001, η2 = .299): highly involved people had a more positive attitude towards buying organic food
(Minvolved = 5.86, SDinvolved = .74) than lowly involved people (Muninvolved = 4.72, SDuninvolved = .98). No
significant effect was found from the emotional condition a participant was in (F(2, 158) = .332, p = .718, η2 = .004) and no interaction effect was found (F(2, 158) = .931, p = .396, η2 = .012).
Prize selection
A logistic regression was performed to measure the effects of the condition the participant was in, their involvement with organic food, and the interaction between these variables on the likelihood that participants choose organic chocolate as their prize. A binary logistic regression was chosen
because the answer options for the dependent variable were binary (either organic, or non-organic chocolate). The logistic regression model was statistically significant (χ2(3) = 9.461, p = .
024). The model explained 9.5% (R2 = .095) of the variance in prize choice and correctly classified
81.6% of cases. The variables condition (Wald(1) = .021, p = .886) and involvement with organic food (Wald(1) = 1.374, p = .241) had no significant effect on the prize a participant chose. No significant interaction effect from involvement and condition on prize selection was found (Wald(1) = .026, p = .871).
Conclusions and discussion
The aim of this study was to provide insight into the effects of pride- and guilt-appeals on organic food purchase intention of highly and lowly involved consumers. This was done through a
between-subjects experimental design with three conditions: pride, guilt, and a control condition. Involvement with organic food served as a moderator.
It was expected that pride appeals will lead to a) a greater purchase intention for organic food, b) a more positive ad attitude, and c) a more positive attitude towards buying organic food than guilt appeals (H1). Such an effect was not found for all parts of the hypothesis: purchase intention for organic food and attitude towards buying organic food did not differ across the pride, guilt or control conditions. The emotional condition a participant was in did significantly affect their attitude towards the advertisement. Participants who were exposed to the guilt-inducing advertisement liked it significantly less than participants who were exposed to the pride and control
advertisements. They felt significantly more guilt than the control condition and significantly more sadness and anger than both other conditions. As guilt, sadness and anger are unpleasant emotions to experience, this result was to be expected.
A significant effect was found for all parts of the second hypothesis. It was expected that people who are more involved with sustainable products will have a) a greater purchase intention for organic food, b) a more positive ad attitude, and c) a more positive attitude towards buying organic food than people who are less involved (H2). Participants who were more involved with organic food had a purchase intention that was 1.28 times higher than that of participants who were less involved. This was an expected result and is in line with findings from, among others, Meijers and
Van Dam (2012). Participants who were highly involved with organic food also had an attitude towards the advertisement that was slightly higher, and an attitude towards buying organic food that was 1.24 times higher than that of participants who were lowly involved.
Lastly, the expectation that when involvement is high, the effect that pride appeals will generate a) a greater purchase intention for organic food, b) a more positive positive ad attitude, and c) a more positive attitude towards buying organic than guilt appeals will be stronger than when involvement is low (H3), could not be confirmed in the present study. No significant interaction effect on
purchase intention, attitude towards the advertisement or prize choice was found from the type of appeal used in the advertisement and the participant’s level of involvement. From these results, no emotional strategy can be selected as most effective for different levels of involvement.
This study contributed to the scientific knowledge available about pride, guilt and organic food by furthering the research started by Onwezen et al. (2013). The anticipated emotions they
researched were brought into real life settings by exposing people to advertisements and taking an inducing role instead of an observing one. Although Onwezen et al. (2013) found a significant effect of anticipated pride and anticipated guilt on people’s intention, such an effect was not found in the present research. This difference may be caused by the fact that in the present research, participants in the pride condition did experience pride more than the guilt condition, but
participants in the guilt condition did not necessarily experience guilt more than participants in the pride condition. If the emotions had been evoked in only the targeted condition, the effects of the emotional appeal would have been clearer. This may also explain why the results differ from those of Antonetti and Maklan (2014), who found that both feelings of pride and guilt increase purchase intention. Another explanation for this difference however, may be a difference in stimulus material. In the present study, the emotions were evoked by a persuasive message in the form of an
advertisement, whereas in the study by Antonetti and Maklan (2014), the emotions were evoked by reflection on a choice made by the participant. It would be interesting to research whether different causes of the emotions generate different outcomes. For example, emotions induced by an
external stimulus (such as an advertisement) may have a different impact on the intention to purchase organic than emotions induced or felt the person him- or herself.
Like all studies, the present research leaves room for discussion and improvements. One limitation of the present study was that participants’ reasons for buying or not buying organic food have not been taken into account. Participants indicated their purchase intention without any motivation for it, meaning a low purchase intention score may not have been caused by exposure to the
advertisement, but simply because of lacking funds or another reason. In future studies, including one or more items in which participants can indicate the reasoning behind their choice could generate new knowledge and understanding about consumer choices. Secondly, the samples used in this research (both for the pretest and main test) were not representative for the whole population. Although approximately equally distributed across genders, both samples were highly educated, likely to be a result of recruiting participants in a university library. A more diverse, more nationally representative sample featuring representative percentages of people with a basic and higher education, would have made the results more generalizable for the whole population. In the present study however, education did not significantly affect the dependent variables. A third and last limitation of the present research was that organic products have been treated as one
category, while, in fact, there are many subcategories such as dairy, meat, and non-food items. As was stated in the introduction, the overall sales of organic products are rising but some categories such as meat and dairy are stagnating or decreasing. A separation of these categories may be as simple as distinguishing bread, coffee, flour and fruit like Tarkiainen and Sundqvist (2009) and could improve clarity of results from future research. Future research might find that for certain product categories, some emotional strategies boost the purchase intention more than others. For example, it might be found that some strategies influence purchase intention differently for very frequently bought items like bread and vegetables than for non-food items like plastic cups and shampoo, which are bought less regularly.
As suggested by Meijers and Van Dam (2012), the focus in the present research was mainly on how to make sustainable products successful, while also considering consumer characteristics such as involvement. The present study contributed to this area of research by bringing the
research of Onwezen et al. (2013) into an emotion-inducing experiment and testing and comparing attitudes, behaviors and intentions of people at varying levels of involvement with organic food. Future studies could focus on exploring the way pride and guilt in advertisements do affect people’s intentions and behavior.
The present study may contain information valuable to advertisers and marketers of organic products. For example, it was found that involved people are more likely to buy organic food and therefore may be persuaded more easily. This knowledge about consumer behavior could aid advertisers in targeting their intended consumers more efficiently. It was also found that overall purchase intention, even for people who are lowly involved, was above midpoint. Future research could investigate the determinants for the significant difference between the involvement groups, such as food pricing or knowledge about organic food. “What makes an individual an involved one?” would be interesting to look into. Knowledge about this could increase advertising efficiency by allowing marketers to target consumers at varying levels of involvement.
The present study produced some insightful results about organic food. If the sales trend pushes on and organic food becomes more mainstream, this might prove to be an interesting and fruitful area for advertisers, researchers and consumers alike.
References
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.
Antonetti, P., & Maklan, S. (2014). Exploring Postconsumption Guilt and Pride in the Context of Sustainability. Psychology & Marketing, 31(9), 717-735.
Coulter, R. H., & Pinto, M. B. (1995). Guilt Appeals in Advertising: What Are Their Effects Journal of Applied Psychology, 80(6), 697-705.
Groot, J. I. M. D., de & Steg, L. (2009). Morality and prosocial behaviour: The role of awareness, responsibility, and norms in the norm activation model. Journal of Social Psychology, 149(4), 425–449.
Harth, N. S., Leach, C. W., & Kessler, T. (2013). Guilt, anger, and pride about in-group environmental behaviour: Different emotions predict distinct intentions. Journal of Environmental Psychology, 34, 18-26.
Hsu, C., & Chen, M. (2014). Explaining consumer attitudes and purchase intentions toward organic food: Contributions from regulatory fit and consumer characteristics. Food Quality and Preference, 35, 6-13.
Knabe, A. (2012). Applying Ajzen's Theory of Planned Behavior to a Study of Online Course Adoption in Public Relations Education (dissertation). Available from http://
epublications.marquette.edu/dissertations_mu/186.
Kummeling, I., Thijs, C., Huber, M., Vijver, L. P. L van de, Snijders, B. E. P., … Dagnelie, P.C. (2008). Consumption of organic foods and risk of atopic disease during the first 2 years of life in the Netherlands. British Journal of Nutrition, 99, 598-605.
Meijers, M. H. C. (2014). On Justifying Eco-Unfriendly Behaviors (unpublished doctoral dissertation). The Amsterdam School of Communication Research (ASCoR), Amsterdam, The Netherlands.
Meijers, M. H. C., & Dam, Y. K. van (2012). Sustainable food purchases in the Netherlands: The influence of consumer characteristics. Journal on Chain and Network Science, 12, 181-198. Ministerie van Economische Zaken (2013). Monitor Duurzaam Voedsel 2013:
Consumentenbestedingen aan duurzaam gelabelde producten. Retrieved from Rijksoverheid website http://www.rijksoverheid.nl/bestanden/documenten-en-publicaties/rapporten/
Onwezen, M. C., Antonides, G., & Bartels, J. (2013). The Norm Activation Model: An exploration of the functions of anticipated pride and guilt in pro-environmental behaviour. Journal of
Economic Psychology, 39, 141-153.
Schifferstein, H. N. J., & Oude Ophuis, P. A. M. (1998). Health-related determinants of organic food consumption in the Netherlands. Food Quality and Preference, 9(3),119-133.
Schwartz, S. H. (1977). Normative influence on altruism. In L. Berkowitz (Ed.). Advances in experimental social psychology (Vol. 10, pp. 221–279). New York:Academic Press. Smith, B. L. (1993). Organic Foods vs Supermarket Foods: Element Levels. Journal of Applied
Nutrition, 45(1), 35-39.
Sparks, P., & Shepherd, R. (1992). Self-identity and the theory of planned behaviour. Social Psychology Quarterly, 55(4), 388-399. doi: 10.2307/2786955
Strick, M., Holland, R. W., Baaren, R. B. van, & Knippenberg, A. van (2009). Finding Comfort in a Joke: Consolatory Effects of Humor Through Cognitive Distraction. Emotion, 9(4), 574–578. Tacken, G. M. L., Winter, M. A., de & Wertheim-Heck, S. C. O. (2007). De invloed van meerwaarde
van biologische producten op consumentenaankopen: Consumentenperceptie van
biologische producten in de supermarkt (The influence of surplus value of organic products on consumer purchases: Consumers' perceptions of organic products available in
supermarkets). Den Haag: LEI
Tarkiainen, A., & Sundqvist, S. (2009). Product Involvement in Organic Food Consumption: Does Ideology Meet Practice? Psychology & Marketing, 26(9), 844-863
Thøgersen, J., & Ölander, F. (2003). Spillover of environment-friendly consumer behavior. Journal of Environmental Psychology, 23, 225-236. doi: 10.1016/S0272-4944(03)00018-5
Tracy, J. L., & Robins, R. W. (2004). Putting the Self into Self-Conscious Emotions: A Theoretical Model. Psychological Inquiry, 15(2), 103-125.
Tropp, D., Ragland, E., & Barham, J. (2008). The Dynamics of Change in the U.S. Food Marketing Environment. Retrieved from U.S. Department of Agriculture, Agriculture Marketing Service website http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5070995
Zanoli, R., & Naspetti, S. (2002). Consumer motivations in the purchase of organic food. A means-end approach. British Food Journal, 104(8), 643-653.
Appendix A
Advertisements created for the present study Girl-advertisement: pride condition
Girl-advertisement: guilt condition
Boy-advertisement: pride condition
Boy-advertisement: guilt condition
Chicken-advertisement: original pride condition (left) and adapted text layout (right)
Chicken-advertisement: guilt condition
Appendix B
Consumer involvement with organic food adapted from Tarkiainen & Sundqvist (2009) Self-identification as green consumer
“I always buy products that are friendly to the environment.” 1 = strongly disagree, 7 = strongly agree
“Buying environmentally friendly products makes me feel good.” 1 = strongly disagree, 7 = strongly agree
“With my own consumption habits I can influence the state of the environment.” 1 = strongly disagree, 7 = strongly agree
“I think of myself as environmentally conscious.” 1 = strongly disagree, 7 = strongly agree
“I am willing to compromise on my own standard of living, in order to prevent environmental problems.”
1 = strongly disagree, 7 = strongly agree Health consciousness
“I choose food carefully to ensure good health.” 1 = strongly disagree, 7 = strongly agree
“I think of myself as health conscious consumer.” 1 = strongly disagree, 7 = strongly agree
“I think often about health issues.” 1 = strongly disagree, 7 = strongly agree Attitude towards organically produced products “I think that organic products are very meaningful.” 1 = strongly disagree, 7 = strongly agree
“I am interested of organic products.” 1 = strongly disagree, 7 = strongly agree
“I think that organic products are important for me.” 1 = strongly disagree, 7 = strongly agree
Buying frequency
“How often have you purchased/do you purchase organic products?” 1 = never, 7 = always
Appendix C
Prize draw options for participants Option 1: organic chocolate bar
