Will communicating the health and environmental benefits of
meat substitute products replace meat on the Dutch plate?
A study on the persuasiveness of health and environmental product claims on
meat substitute products among meat-eaters, and the moderating roles of food
choice motives and meat resemblance
Master Thesis, Graduate School of Communication Master’s Program Persuasive Communication
Author: Fenna van der Kruit Student number: 11865709
Supervisor: Mw. dr. E. (Ester) de Waal Date of completion: July 25th 2018
1
Abstract
The consumption of meat severely harms both human health and the environment. However, Dutch consumers’ meat consumption is vastly increasing and they seem little aware of the consequences of their meat-intensive diets. Consequently, the urge to understand how to stimulate the acceptance of meat substitute products among meat-eaters has grown along with the question to what extent information provision could facilitate a change in their dietary behavior. This research answers these questions by studying the role of health and environmental product claims on consumer responses to meat substitute products. Whereas these claims can enhance attitudes and purchase intentions towards various food products, their potential effectiveness on meat substitute products remained unexplored. Hence, the following research question was addressed: To what extent do health and environmental
product claims on meat substitute products influence Dutch meat-eaters’ attitudes and purchase intentions towards these products, and is this affected by consumers’ food health and environmental motives, and the products’ meat resemblance? Among Dutch meat-eaters
(N = 380), a 3 (Product claim: health claim, environmental claim, no claim) x 2 (Meat resemblance: resembles meat, does not resemble meat) between-subjects online survey experiment was employed. Contrary to the expectations, both product claims had no significant effect on meat-eaters’ attitudes and purchase intentions towards meat substitute products. However, meat-eaters’ food environmental and health motives significantly influenced their attitudes towards the products with the environmental claim. Compared to the other product claim conditions, those with high food environmental motives showed more positive attitudes and those with high health motives more negative attitudes towards the product with the environmental claim. The products’ meat resemblance appeared the most powerful asset as it not only suggested increased effectiveness of the product claims, but also directly significantly influenced meat-eaters’ attitudes and purchase intentions. In order to successfully reach meat-eaters, marketers should therefore offer those products that resemble meat and target exclusively environmentally-friendly consumers with an environmental claim. Health claims should be disregarded.
KEYWORDS: food marketing, meat substitute acceptance, product claims, health claims, environmental claims, food choice motives, information provision, health and sustainability communication.
3
Introduction
Despite a slight decrease since 2010, meat consumption in the Netherlands is rising again (NOS, 2017). This is worrisome as excessive meat intake correlates with cardio vascular diseases, high cholesterol, cancer, and obesity (Troy & Kerry, 2010). Meat consumption additionally accounts for substantial environmental damage as its production process strongly contributes to all greenhouse gas emissions (GHGE) and biodiversity loss (Zur & Klöckner, 2014). Consequently, there is an urgent need for meat reduction interventions (de Boer, Schösler, & Aiking, 2017).
Promoting the consumption of meat substitutes among meat-eaters has been raised as a viable option (Apostolidis & McLeah, 2016). Meat substitutes are food products primarily made from processed plant ingredients that replace meat in the meal (Hoek, Luning, et al., 2011). Relative to meat, they are lower in fat and have lower ecological footprints regarding carbon, energy, and land use (Apostolidis & McLeay, 2016). Stimulating meat-eaters to substitute meat for these products is challenging as they show an overall negative attitude towards consuming more plant-based foods and less animal products (Hoek, Pearson, James, Lawrence, & Friel, 2016). This is likely due to the poorly acknowledged relationship between meat consumption and human health as well as the environment (Macdiarmid, Douglas, & Campbell, 2016). Westerners still believe vegetarian diets are less healthy than meat-based diets (Stoll-Kleeman & Schmidt, 2017) and only 12 percent of the Dutch consider meat reduction as a means to mitigate climate change (de Boer, de Witt & Aiking, 2016).
Nevertheless, information provision endeavors can enhance consumers’ knowledge, shape their attitudes and redirect their decision-making regarding food choices (Verbeke, 2008). Thereby, academia point to insufficient sustainability information at the point of purchase which hinders environmentally-friendly meat choices (Cho, 2015; Dibb & Fitzpatrick, 2014) and propose that increased awareness of the impact of livestock production could expand the market for meat alternatives (de Boer, de Witt & Aiking, 2016; Slade,
4
2018). Similarly, Arora, Bradford, Arora, and Gavino (2017) believe marketeers could spark meat-free consumer choices through utilizing health claims. Informing meat-eaters about the benefits of meat substitutes might thus form a steppingstone in reducing meat consumption.
Hence, this research explores health and environmental product claims which inform consumers by thoroughly describing the relationship between the food product and its benefit (Hoek et al., 2017). Health claims transmit information about nutritional or positive health benefits of food products or their components (Nocella & Kennedy, 2012). Environmental claims are statements about the environmentally beneficial characteristics of products (OECD, 2011). By emphasizing product attributes, these claims have proven to enhance attitudes and purchase intentions towards different consumer goods (Fernqvist & Ekelund, 2014; Randheer, Al-aali, & Al-ibrahim, 2014; Shan et al., 2017). Since the latter are predictive of behavior (Ajzen, 1991), health and environmental claims can generate behavioral change which has been evinced by the products’ increased market share (Nielsen, 2015; Williams, 2005).
Literature on these claims regarding meat remains scarce and has either focused on their effects on choice for different meat products and/or only studied one claim type (Hoek et al., 2017; Hung, Kok, & Verbeke, 2016; Koistinen et al., 2013; van Loo, Caputo, Nayga, & Verbeke, 2014; Shan et al., 2017). While these studies show claims can influence meat product preferences, none of them considered the effects of the two claim types on attitudes and purchase intentions for one specific product. Nonetheless, they consistently address claim effectiveness depends on the meat product. Health claims for example stimulated consumers to select pork, but not beef (Koistinen et al., 2013).
Literature on meat substitutes has, however, focused primarily on consumer acceptance, based on the products’ perceived appropriateness in a dish; appearance; sensory appeal; and consumers’ familiarity with the product (Elzerman, van Boekel, & Luning, 2013; Hoek, Luning et al., 2011; Hoek, van Boekel, Voordouw & Luning, 2011). Thus far, advertising
5
efforts have largely been ignored. Just one study combined advertising and consumer acceptance by including a vegetarian option in their investigation of the effect of labeling on minced-meat products (Apostolidis & McLeay, 2016). Though this product appeared the least preferred compared to its meat reference products, interestingly, low fat content and carbon footprint attributes were highly valued. Despite this promising outlook, health and environmental claims have not been studied in relation to meat substitutes as a separate product category, leaving their potential to enhance attitudes and purchase intentions among meat-eaters unexplored.
Nonetheless, it is known that product claims are especially effective when they match consumers’ motivations (Stancu, Grunert, & Lähteenmäki, 2017). Since motivations in food orientation affect food choices (de Boer, Schösler, & Boersema, 2013), claim effectiveness may be subject to how meat-eaters value making healthy or environmentally-friendly food choices. Literature on product claims further suggests that claim evaluation is contingent on how consumers categorize the product (e.g. healthy/unhealthy) and what associations the category evokes (Nocella & Kennedy, 2012; Olsen, Slotegraaf, & Chandukala, 2014). Similarly, resembling meat substitutes to real meat stimulates consumers to categorize the products into a familiar product category, enhancing its acceptance among meat-eaters (Hoek et al., 2011). However, how claims on meat substitute products affect meat-eaters with different food choice motives and interact with products that do (not) resemble meat remains unstudied. To close these literature gaps, the following research question is proposed: To
what extent do health and environmental product claims on meat substitute products influence Dutch meat-eaters’ attitudes and purchase intentions towards these products, and is this affected by consumers’ food health and environmental motives, and the products’ meat resemblance? The scientific relevance of the research is two-fold. First, directly comparing the effects of a health and environmental claim will extend the literature
6
on education through information provision by clarifying whether and which of these appeals will be of influence to a segment as complex as meat-eaters, which until today remains highly debatable (Apostolidis & McLeay, 2016; Hoek, Luning et al., 2011). Secondly, by investigating information provision in relation to meat substitutes, the research contributes to meat substitute acceptance literature that calls for experimental studies on how to motivate consumers to minimize their meat consumption (Hartmann & Siegrist, 2017). The findings are of additional practical and social relevance as they could assist marketers in crafting the most suitable packaging-design for meat-eaters, and could possibly contribute to meat mitigation in the Netherlands.
Theoretical background
Knowledge Deficit Theory
Knowledge is a precondition for changing behavior (Stoll-Kleemann & Schmidt, 2017). Hence, knowledge about the consequences of food consumption is essential to motivate consumers to change their dietary patterns and agree with meat reduction interventions (Hartmann & Siegrist, 2017). As such, awareness of the impacts of meat consumption on health and climate change is strongly related to support for meat reduction and the adoption of vegetarian diets (Mullee et al., 2017; Pohjolainen, Tapio, Vinnari, Jokinen, & Räsänen, 2016). It is thus arguable that unawareness rather than unwillingness takes center stage in impeding meat reduction intentions and meat substitute acceptance among meat-eaters.
Knowledge deficit theory (KDT) by Schultz (2002) supports this assumption. It postulates that educating individuals through the provision of information induces behavior change. It builds on the notion that individuals are principally willing to contribute – e.g. to themselves or the environment – but do not understand why and how. KDT’s solution to inaction is therefore the provision of information about why and how an individual must
7
attempt action. Information will motivate behavior change when a lack of knowledge originally formed an internal barrier for performing the behavior (Nolan, 2010).
In stimulating meat reduction, the theory proved true by revealing advantages of meat mitigation. Informing consumers about lowered CO2 emissions increased environmental
concern and intentions to cut on meat intake (Graham & Abrahamse, 2017). Similarly, information about the health and sustainability benefits of a plant-based diet led to intentions to decrease the consumption of red meat and increase plant-based alternatives (Vaino, Irz, & Hartikainen, 2018).
The role of product claims on meat substitutes
Information alone does not, however, guarantee behavioral intentions and change (Ratner et al., 2008). Campbell-Arvai, Arvai, and Kalof (2014) demonstrated that providing information about the environmental benefits of a novel meatless dish did not increase consumers’ choices for this dish. This is, however, not fortuitous as for information provision to be optimally effective, it should be accompanied by efforts that remove external barriers to the promoted behavior (Gardner & Stern, 1996; Schultz, 2002). The main external barriers to the adoption of more plant-based diets lie in what consumers should eat instead of meat and in the reluctance to shift to unfamiliar dishes (Corrin & Papadopoulus, 2017; Mullee et al., 2017). These barriers could be removed by offering consumers a direct solution: a meat substitute that replaces the meat in their usual dish. Therefore, providing information on the beneficial attributes of meat substitutes through product claims – the why consumers should change their behavior – on a meat substitute product – the how – should increase attitudes towards and intentions to purchase these products.
Prior studies on health and environmental claims embody these expectations. Health claims positively influenced attitudes towards functional foods consumers were unfamiliar with (Tudoran, Olsen, & Dopico, 2009) and a low-fat product that
8
was negatively perceived (Kähkönen, Tuorila, & Rita, 1996). Despite consumers’ poor awareness of meat products containing nitrite, studies on processed meat further showed that attitudes and purchase intentions increased when the products were promoted as containing reduced levels of nitrite (Hung et al., 2016; Shan et al., 2017). Additionally, environmental claims enhanced consumers’ attitudes by altering consumers’ perception about the functional aspect of products (Carlson, Grove, & Kangun, 1993) and increased intentions to make green purchases (Randheer et al., 2014). Attitudes also increased for utilitarian products when a claim emphasizing an environmental benefit was present (Steinhart, Ayalon, & Puterman, 2013). Since meat substitutes are functional alternatives to meat (Elzerman et al., 2013), similar effects are expected. It is therefore hypothesized that among Dutch meat-eaters:
H1: Exposure to meat substitute products with a claim (either health or environmental)
will lead to a) more positive attitudes and b) higher intentions to purchase the products compared to meat substitute products without a claim.
Claim type
Scholars anticipate information concerning personal health to be most effective in increasing meat substitute usage (de Boer et al., 2014; Hoek, 2010; Hartmann, 2017). This also applies to meat consumption as fat content arose as major factor influencing meat choice and willingness to pay, whereas a carbon footprint label played a secondary role (Apostolidis & McLeay, 2016; Koistinen et al., 2013). Scholars believe this is due to priority given to personal motives in making food choices (Hoek et al., 2017), which could be explained by food consumption being an individually-controlled behavior. Consumers feel that by eating foods with health claims they can – themselves – directly influence a health function (Dean et al., 2012; Weiner, 2010). Alternately, environmental benefits are only perceivable on the long-term and the establishment is highly dependent on efforts of others (Hyland, Henchiona, McCarthy, & McCarthy, 2017). Regarding food consumption, one’s self-confidence to reach
9
the acclaimed health benefit – self-efficacy beliefs – might therefore exceed one’s collective-efficacy beliefs, the belief in the collective power to establish the pro-environmental benefit (Bandura, 1977; Bandura, 1986). Since efficacy beliefs predict attitudes and behavior (Chen, 2015), it is expected that:
H2: Exposure to meat substitute products with a health claim will lead to a) more
positive attitudes and b) higher intentions to purchase the products compared to meat substitute products with an environmental claim.
Food Choice Motives
The extent to which product claims influence attitudes and purchase intentions is further attributed to consumers’ motivation or the fit with consumers’ health or environmental goals (van Buul & Brouns, 2015; Thøgersen, 2000). These goals are reflected in consumers’ Food
Choice Motives (FCMs). FCMs are shaped by the importance consumers attach to certain
food attributes and explain individual food preferences (Steptoe, Pollard, & Wardle, 1995). As such, they predict attitudes towards (healthy) eating (Sun, 2008) and dietary intake (Pollard, Steptoe, & Wardle, 1998). Whereas health motives pertain to the importance individuals attach to food being healthy and nutritious, environmental motives pertain to animal welfare and environmental protection to be reflected in food choices (Lindeman & Väänänen, 2000; Steptoe, Pollard & Wardle, 1995). These FCMs could be vital in the persuasiveness of health and environmental claims as they determine whether consumers perceive an advantage from engaging in active reasoning (Verbeke, 2008) and to what information consumers are sensitive (Verain, Sijtsema, & Antonides, 2016). As outlined by the elaboration likelihood model (ELM), this is crucial to induce persistent attitudinal and behavioral change.
10
The Elaboration Likelihood Model
Following the ELM, one’s motivation to process a message is determined by the degree to which a consumer considers an attitude object as personally important (Petty & Cacioppo, 1990). Motivated consumers follow a central route to attitude change whereby they attentively evaluate the object based on the attributes they consider relevant (Hartmann, Ibáñez, & Sainz, 2005). Consequently, they form strong attitudes which are more predictive of behavior (Petty & Cacioppo, 1986). Since attitudes towards a product become more favorable when the product corresponds with consumers’ goals (Hoek, 2010), meat-eaters’ giving importance to health motives are likely to be receptive to enduring, and positive, attitudinal and behavioral change after exposure to a health claim; those giving importance to environmental motives after exposure to an environmental claim. However, consumers with low motivation do not elaborate on the message and will be more influenced by external peripheral cues surrounding the information (Hartmann et al., 2005; Verbeke, 2008), making product claims less relevant. Compared to the absence of the respective claim, the presence of a health claim should therefore be more effective for meat-eaters with high health motives and an environmental claim for those with high environmental motives. Accordingly:
H3: Exposure to meat substitute products with a health claim will lead to a) more
positive attitudes and b) higher intentions to purchase the products compared to meat substitute products without a claim, which effect becomes stronger the more importance meat-eaters attach to food health motives.
H4: Exposure to meat substitute products with an environmental claim will lead to a)
more positive attitudes and b) higher intentions to purchase the products compared to meat substitute products without a claim, which effect becomes stronger the more importance meat-eaters attach to food environmental motives.
11
Consumers can, however, attach importance to multiple food attributes simultaneously (Steptoe et al., 1995). This is expected to be valuable in predicting attitudes and purchase intentions towards the environmental claim as consumers tend not to differentiate between self-centered health motives and altruistic motives when making sustainable product choices (Aschemann-Witzel, 2015; O’Rourke & Ringer, 2016). Thereby, environmentally-friendly products are greatly selected for health reasons and perceived as more healthy than their traditional alternatives (Leire & Thidell, 2005; Sörqvist et al., 2015). Hence, it is expected that:
H5: Exposure to meat substitute products with an environmental claim will lead to a)
more positive attitudes and b) higher intentions to purchase the products compared to meat substitute products without a claim, which effect becomes stronger the more importance meat-eaters attach to food environmental and health motives.
Meat Resemblance
The categorization of a product further determines the persuasiveness of health and environmental claims and the acceptance of meat substitutes. Herein, the physical appearance of the meat substitute plays a dominant role (Elzerman et al., 2013). When this product resembles meat, consumers categorize it as similar to its meat reference product, which is especially evident for products that mimic processed meat products (e.g. hamburgers) (Hoek et al., 2011). This induces an assimilation effect in which the associations evoked by the meat reference product are allocated to the meat substitute (Hoek, 2010). Consequently, the products share similar sensory and convenience attributes, increasing meat-eaters’ perceived familiarity with, and hereby the acceptance of, these meat substitute products (Hoek, Luning et al., 2011).
12
Schema Theory and moderate incongruity
This shared product category also determines the effectiveness of product claims as the perceived (mis)match between the claims and product category affects product evaluations (Nocella & Kennedy, 2012; Olsen et al., 2014). According to schema theory, associations related to a product category, defined as schemas, are activated when consumers are exposed to a new product and function as reference points to which its advertised attributes are compared (Cheong & Kim, 2011). The consumer compares the claim with the schema activated from the product category and evaluates its perceived similarity. This evaluation influences the extent to which consumers process the information and are subject to attitudinal change (Stoltman, 1991). Specifically, a moderate level of incongruity between the claim and schema is most effective in affecting consumers’ attitudes because consumers extensively process the message driven by an urge to solve the incongruity. This process induces feelings of satisfaction which results in most favorable product evaluations. When there is congruency between the claim and activated schemas, however, the message is shallowly processed which could only lead to a modest positive evaluation. Extreme incongruity results in the least positive evaluations as the incongruity cannot be resolved (Cheong & Kim, 2011; Mandler, 1982). Bech-Larsen and Grunert (2003) confirmed the effectiveness of moderate incongruity for products with health claims. Promoting the presence of vitamins and minerals increased consumers’ attitudes towards processed foods but not towards healthier, natural foods. Regarding environmental claims, Olsen et al. (2014) similarly showed consumers’ brand attitudes became less favorable the more the product was perceived as exclusively healthy.
To establish this moderate incongruity and facilitate successful substitution of consumer products from the same product category, Wansink (1994) demonstrated advertisers must focus on distinct product attributes. The more the substituted product looks similar to the reference product, the more its functional dissimilarities should be expressed. Corroborating
13
these findings, Hoek and colleagues (2017) showed consumers’ willingness to shift to healthy- and sustainably-labelled food products increased when there was close similarity with the original reference products. On the basis of these incongruency effects, product claims should be more persuasive in conjunction with meat substitute products that resemble meat. A product that does not resemble meat generates a greater disconnection between the original and alternative product, and rather evokes associations of naturalness by revealing its vegetable-based ingredients. Since consumers associate naturalness with health (Sijtsema et al., 2007) and sustainability attributes (Verhoog, Matze, van Bueren, & Baars, 2003), the presence of health and environmental claims on products that do not resemble meat are prone to result in congruity. Processed products that resemble meat evoke associations related to the meat product such as taste and meal-context (Hoek et al., 2011) and are unlikely to evoke health, environmental or naturalness associations. Due to this moderate incongruity, it is finally hypothesized that:
H6: Exposure to meat substitute products with a claim (either health or environmental)
will lead to a) more positive attitudes and b) higher intentions to purchase the products compared to meat substitute products without a claim, which effect will be stronger for products that resemble meat than for products that do not resemble meat.
14
Figure 1. Conceptual model
Methods
Design
A 3 (Product claim: health claim, environmental claim, no claim) x 2 (Meat resemblance: resembles meat, does not resemble meat) factorial between-subjects survey experiment was designed in Qualtrics. This design minimized the chance of carryover effects expected to occur when exposing participants to multiple conditions. Dutch meat-eaters were randomly assigned to one of the six conditions in which they were exposed to a meat substitute product containing either a health claim (HC), environmental claim (EC), or no claim (NC) on a product that resembles meat (MRproduct) or not (Non-MRproduct). Figure 2
provides a design diagram.
Figure 2. Experimental design including sample distribution (N = 380) Sample
Through the use of non-probability sampling, all participants were recruited online by making use of snowball sampling within the researcher’s e-mail network and Facebook groups. The recruitment text reached out to Dutch meat-eaters by stating the inclusion criteria (not adhering to a vegetarian/vegan diet, >18 in age, Dutch nationality).
After considering missing values, response time and zero variation in answers, 30 participants were removed from the sample. Consequently, a final sample size of 380 was used. Participants ranged from 18 to 87 in age (M = 38.49, SD = 17.34) and 64% was female.
Product claim
Meat resemblance HC EC NC
MRproduct Condition 1 (n = 63) Condition 3 (n = 67) Condition 5 (n = 65)
15
41.6% of the sample had completed a university degree, 69.7% consumed meat at least 4 times a week during the meal, and 53.2% had never tried or rarely ever consumed meat substitutes.
Procedure
All participants were presented with an informed consent form, outlining the study’s purpose, relevant ethical information and inclusion criteria. Following consent, participants were randomly assigned to one of the six conditions and asked to attentively observe the product they were respectively exposed to. A 10-second time-lock ensured attention was devoted to the stimuli. Thereafter, participants evaluated the products through purchase intentions and product attitude scales. Manipulation check questions followed.
Subsequently, statements about daily food choices were presented, assessing participants’ health and environmental food choice motives, followed by scales measuring meat substitute familiarity; brand familiarity; and meat substitute consumption attitudes. Next, meat consumption statements assessed participants’ level of meat-attachment. Demographics were collected at last to avoid untruthful or socially-desirable answers, due to feeling stereotyped by reported meat consumption frequency, education, gender or age (Babbie, 2010). Participants were debriefed and thanked for participating (Appendix A).
Stimulus material
Product claim was manipulated at three levels. Firstly, the HC was crafted by
evaluating the effectiveness of the official claims allowed within the EU: general health claims and reduction of disease risk claims (RDRCs). General claims outperformed RDRCs on product evaluations across a Belgian sample where a bias prevailed towards more highly educated people (Verbeke, Scholderer, & Lähteenmäki, 2009). Since consumer responses to health claims depend on country and level of education (van Trijp & van der Lans, 2007), the
16
current sample was expected to equally prefer general health claims over RDRCs due to cultural proximity and snowball sampling among highly-educated participants. Additionally, the effectiveness of RDRCs in the Netherlands is largely restricted to individuals perceiving to be at risk (van Assema, Glanz, Burg, & Kok, 1996). Hence, based on European Regulation 1924/2006 the following general health claim was selected: “Arm aan verzadigde vetten. Dit
product is goed voor de cholesterolspiegel” (Low in saturated fats. This product is beneficial
for cholesterol levels). This claim was selected as meat substitutes’ low saturated fat contents have been identified as its main health benefit (Kumar et al., 2017) and cholesterol-lowering claims are highly valued among Dutch consumers (van Assema et al., 1996). In composing the EC, a Type II self-declared environmental claim was selected (OECD, 2011). This claim type allows businesses to construct their own claim to encourage the demand for products causing less stress on the environment (ISO, 2012), whereby similarity with the HC and internal validity could be achieved. The following claim was constructed: “Laag in de uitstoot
van broeikasgassen. Dit product is milieuvriendelijk” (Low in greenhouse gas emissions.
This product is environmentally-friendly). This claim was selected as reduced GHGE is the greatest environmental benefit meat substitutes provide relative to meat products (van Mierlo, Rohmer, & Gerdessen, 2017). To pursue external validity, the claim was constructed following rules set by ISO 14021 (ISO, 2012). Lastly, the NC condition functioned as control condition. Herein, participants were presented to products not containing any claim.
Meat resemblance was manipulated at two levels: MRproduct was
established by selecting a meat substitute identical to a hamburger, Non-MRproduct by one
clearly deviating in color (green) and ingredients (vegetables). To maintain internal validity, both products contained the internationally recognized V-logo signifying a vegetarian product; were displayed in a meal-context illustrating its usage; and were shaped as and named ‘Burger’. All additional information was removed.
17
The carrier-products were selected from the brand ‘Garden Gourmet’ in order to further maintain external validity and minimize potential effects of brand familiarity. Recently, the brand underwent a name alteration (from ‘Tivall’ in 2017), expecting to remain relatively unfamiliar to consumers. The stimuli are presented in Appendix B.
Pretests
An online pretest was employed among Dutch meat-eaters (N = 35) ranging from 20 to 79 in age (M = 46.15, SD = 16.38). Participants were exposed to the three within-subjects conditions (HC, EC, NC) to assess whether both claims and products were perceived as intended and necessary for this relatively small sample to reduce the chance of error variance. Participants rated whether they believed the meat substitute product conveyed a health benefit (Q1) and environmental benefit (Q2) on a self-devised 7-point scale (1 = not at all, 7 = very
much so). Subsequently, they indicated the extent to which the products looked like meat
(Q3). Paired samples t-tests showed significant differences across the conditions and products (table 1, Appendix C). Health benefit scores were significantly higher for the HC than the other claim types, as were environmental benefit scores for the EC. Additionally, the MRproduct
was perceived as more similar to meat than the Non-MRproduct. Lastly, results indicated 85.7%
of the sample was unfamiliar with Garden Gourmet (M = 1.60, SD = 0.98). The stimulus material was thus successfully designed.
Another pretest tested for ambiguities within the complete survey instrument (N = 3). Herein, the manipulation check appeared unsuccessful due to the between-subjects design that did not allow for direct comparisons between the claims and products, and the survey length by which participants failed to remember the claim. Manipulation check questions were therefore moved to the beginning of the survey and directly asked which claim the participant had been exposed to (HC/EC/NC/“I don’t remember”).
18
Measurements
All existing scales were translated from English into Dutch. Appendix D provides a complete overview of the original and translated items of the multi-item scales, including reversed items. Appendix E shows the factor loadings generated from the principle component factor analyses (PCA) conducted with varimax rotation.
Dependent variables. Product attitudes were measured by Kozup, Creuer and Burton’s
(2003) measurement which assesses consumers’ overall product attitude on a 3-item 7-point semantic differential scale (unfavorable-favorable; bad-good; negative-positive). The items
unpleasant-pleasant, wise-foolish and unenjoyable-enjoyable were added to the scale as these
also proved relevant to evaluation of healthier food products (Solheim & Lawless, 1996). The PCA indicated the extended scale was unidimensional, explaining 65.61% of the variance. The new scale showed good reliability (α = .89), (M = 4.60, SD =1.20).
Purchase intentions were measured by two items from the three-item scale by Kozup et
al. (2003) proven successful in assessing purchase intentions for products with health claims (Küster & Vila, 2017): “How probable is it that you would consider the purchase of this product in the near future?” on a 7-point scale (1 = very unprobeable, 7 = very probable) and “Given the information on the package, how likely would you be to purchase this product in the near future?” (1 = very unlikely, 7 = very likely). These two items were selected to minimize respondent fatigue (Bergkvist, 2015) but to still account for possible differences between participants predominantly influenced by their prior attitudes and those more influenced by the product claim. The PCA indicated the two-item scale was unidimensional, explaining 90.72% of the variance. The new scale showed good reliability (α = .90), (M = 4.60, SD = 1.20).
Independent variables. Food health motives were measured by the original health
19
participants to evaluate the FCQ-statement “It is important for me that the food I eat on a typical day… ” for 6 items. Examples include: “Keeps me healthy” and “Is nutritious”. Recommended by Chen (2011), a 7-point category scale was adopted instead of the original 4-point scale, allowing for more discrimination (1 = strongly disagree, 7 = strongly agree). The item “Is high in protein” was changed in “Is low in saturated fat” to optimally align the motives with the health claim. The PCA indicated the scale was unidimensional, explaining 59.52% of the variance. The scale showed good reliability (α = .86), (M = 5.00, SD = 1.03). Higher scores suggest meat-eaters attach high importance to health motives in making food choices, i.e. they have high health motives/are health motivated.
Food environmental motives were assessed using
Lindeman and Väänänen’s ecological welfare scale (2000), which embodies animal welfare and environmental protection and was developed as extra measurement within the original FCQ. On 7-point scales, participants rated the FCQ-statement for 5 items, including: “Is packaged in an environmentally-friendly way” and “Has been produced in a way which has not shaken the balance of nature”. The scale proved highly reliable (α = .94), (M = 4.91, SD
=1.42). Higher scores suggest meat-eaters attach high importance to environmental motives
in making food choices, i.e. they have high environmental motives/are environmentally motivated.
Covariates. Participants’ level of Meat-attachment was assessed by two factors of the
Meat Attachment Questionnaire (MAQ) (Graça, Calheiros, & Oliveira, 2015). The factors hedonism and dependence were selected as these inhibit reduced meat consumption (Verbeke, 2008). Participants rated 9 meat-related items on 7-point Likert scales (1 = strongly disagree, 7 = strongly agree) instead of the original 5-point scales, allowing for more discrimination within the variable. Example items include: “I love meals with meat” and “Meat is irreplaceable in my diet”. Several items were reverse-coded to correct for response
20
acquiescence and straight-line responding (Wong, Rindfleisch, & Burroughs, 2003). Indicated by the Eigenvalue-criterion of >1, the PCA indeed revealed two components representing
Hedonism (component 1, eigenvalue of 4.27) and Dependence (component 2, eigenvalue of
1.10). Together these components explain 59.69% of the variance. Reliability analysis for component 1 showed this scale was reasonably reliable (α = .74). The item “I’m not a big fan of meat” was however removed to yield greater reliability (α =.80). Adding the remaining items, the Hedonism Scale was created (M = 4.59, SD = 1.37). Reliability analysis for component 2 showed the Dependence Scale was reliable (α =. 82), (M = 3.63, SD = 1.30). Higher scores suggests meat-eaters very much enjoy meat (Hedonism) or are very dependent on meat consumption (Dependence).
Meat substitute consumption attitude was measured by combining items from meat
consumption attitude scales (Berndsen & Van der Pligt, 2004; Graham & Abrahamse, 2017). Participants responded to the statement: “Within my diet, I consider the consumption of meat substitutes as” on a 5-item 7-point semantic differential scale (unimportant-important;
unfavorable-favorable; dispensable-indispensable; bad-good; harmful-beneficial). The PCA
showed the scale was unidimensional, explaining 67.37% of the variance. The new scale showed good reliability (α = .88), (M = 4.15, SD = 1.28).
Meat substitute familiarity was assessed by a 5-point category scale derived from
Tuorila, Lähteenmäki, Pohjalainen and Lotti (2001) which includes both awareness of and experience with a food product. It presents participants with an explanation of the product and asks which answer is most applicable to them (1 = I had never heard of meat substitute
products, 5 = I regularly eat meat substitute products), (M = 3.52, SD = 0.98).
Brand familiarity was measured along a self-devised 5-point category scale to ensure
internal validity with the meat substitute familiarity scale. Inspired by Tuorila et al. (2001), participants were introduced to the brand and asked which answer is most applicable to them
21
(1 = I had never heard of Garden Gourmet, 5 = I know Garden Gourmet, and regularly eat a
product of this brand), (M = 2.18, SD = 1.24).
Meat consumption frequency was assessed through a self-devised 7-point category
scale asking “How many days a week do you on average consume meat during the main meal?” (1 = once or less, 7 = every day), (M = 4.17, SD =1.55).
Results
Prior checks
Randomization check. Chi-square tests of independence revealed no significant
differences in gender, χ2 (10, N = 380) = 15.72, p = .108, nor educational degree, χ2 (22, N = 380) = 16.51, p = .684, between the conditions. The results of a one-way ANOVA moreover showed no significant differences for age (F(5, 374) = 0.98, p = .432) nor the covariates hedonism (F(5, 374) = 0.67, p = .650), dependence (F(5, 374) = 0.50, p = .774), meat substitute consumption attitude (F(5, 374) = 0.87, p = .499), meat substitute familiarity (F(5, 374) = 1.40, p = .222), brand familiarity (F(5, 374) = 1.15, p = .334), and meat consumption frequency (F(5, 374) = 1.44, p = .208). Hence, randomization was successful.
Manipulation check. Another Chi-square test of independence was conducted to test
whether participants recalled the correct claim. Results revealed a significant difference between the conditions exposed to the HC, EC and NC χ2 (6, N = 380) = 617.45, p < .001, as such that 91.9% (N = 114) of participants exposed to the HC, 91.3% (N = 116) to the EC, and 76.7% (N = 99) to the NC remembered the correct claim. Moreover, an independent sample t-test revealed those exposed to the MRproduct rated the question to which extent the product
looked like meat significantly higher (M = 5.63, SD = 1.43) than participants exposed to the Non-MRproduct (M = 2.59, SD = 1.54), t(378) = 19.89, p < .001, 95% CI [2.74, 3.34].
22
Preparation prior to the analyses
New categorical variables were constructed and dummy coded with values of 0 and 1 (table 1). Additionally, a bivariate Pearson Correlation revealed all covariates significantly related to one or both of the dependent variables. Hence, all these variables were controlled for in the analyses. Consider Table 1, Appendix F, for the respective correlation matrix.
Table 1. Constructed categorical variables
Hypothesis testing
Main effects. Hypothesis 1 and 2 were tested by conducting One-Way MANCOVAs,
allowing to test relationships on both dependent variables whilst reducing the chance of Type I error (Field, 2013). All assumptions were validated and Product attitudes and Purchase
intentions were selected as dependent variables.
Hypothesis 1 predicted exposure to meat substitute products with a claim would result in more positive attitudes (H1a) and higher purchase intentions towards the products (H1b) compared to exposure to products without a claim. Claim Presence was selected as the independent variable. Results showed no significant difference on the combined dependent variables based on exposure to a product with or without a claim, F(2, 368) = 0.17, p = .842, Wilks’ Λ = .999. H1a and H1b were not accepted.
Hypothesis 2 proposed that exposure to a product with a HC would lead to more positive attitudes (H2a) and higher purchase intentions (H2b) than exposure to an EC. Claim
Variable name 0 1 N
Claim Presence A claim (n = 251) NC (n = 129) 380 Claim Type HC (n = 124) EC (n = 127) 261 Health Claim HC (n = 124) NC (n = 129) 263 Environmental Claim EC (n = 127) NC (n = 129) 256 Meat Resemblance Resembles meat
(n = 195)
Not Resembles meat (n = 185)
23
Type was selected as the independent variable. Results showed no significant difference on
the combined dependent variables based on the type of claim, F(2, 239) = 0.38, p = .681, Wilks’ Λ = .997. H2a and H2b could again not be supported.
Interaction effects. Hypotheses 3 and 4 were tested using PROCESS Model 1 (Hayes,
2013). Hypothesis 3 proposed that relative to NC, exposure to a HC would result in more positive attitudes (H3a) and higher intentions to purchase the products (H3b), which effect would become stronger the more importance meat-eaters attached to food health motives. Results revealed no significant interaction effect between Health Claim and Food health
motives on Product attitudes, b = 0.07, SE B = 0.13, t(240) = 0.50, p = .616, 95% CI [-0.20,
.33]. Similarly, no significant interaction effect on Purchase intentions was found, b = -0.08,
SE B = .18, t(240) = -0.42, p = .673, 95% CI [-0.44, 0.28]. Hence the importance attached to
health motives did not moderate the effect of exposure to a HC (vs. NC) on attitudes and purchase intentions. H3a and H3b were not confirmed.
Hypothesis 4 investigated whether exposure to an EC (vs. NC) would result in more positive attitudes (H4a) and higher intentions to purchase the products (H4b), and whether this effect would become stronger the more importance meat-eaters attached to food environmental motives. Results revealed a significant interaction effect between
Environmental Claim and Food environmental motives on Product attitudes, b = -0.19, SE B
= 0.09, t(243) = -2.05, p = .042, 95% CI [-0.37, -.01]. Displayed in figure 3, the more importance meat-eaters attached to environmental motives, the more positive attitudes they showed towards the products with the EC (vs. NC). H4a was supported. However, results revealed no significant interaction effect between Environmental Claim and Food
environmental motives on Purchase intentions, b = -0.16, SE B = .12, t(243) = -1.26, p = .209,
24
moderate the effect of exposure to an EC (vs. NC) on purchase intentions. H4b could not be supported.
Figure 3. Interaction effect of Environmental Claim and Food environmental motives on Product attitudes for EC
25
To test hypothesis 5, PROCESS Model 3 was utilized (Hayes, 2013). Exposure to an EC (vs. NC) should result in more positive attitudes (H5a) and higher intentions to purchase the products (H5b), which effect would become stronger the more importance meat-eaters attached to food environmental and health motives. The three-way interaction of
Environmental Claim × Food environmental motives × Food health motives on Product Attitude neared significance, b = -0.14, SE B = 0.08, t(239) = -1.80, p = .073, 95% CI [-0.30,
0.01]. Contrary to the hypothesis, figure 4 displays that when meat-eaters attached equal importance to both motives, an EC (vs. NC) was unlikely to induce differences in product attitudes. However, the figure suggests a positive effect of the EC when environmental motives were high but
health motives relatively low. These differences were, however, insignificant. Similarly, no significant three-way interaction of Environmental Claim ×
Food environmental motives × Food health motives on Purchase intentions was found, b =
-0.18, SE B = 0.11, t(239) = -1.65, p = .101, 95% CI [-0.39, 0.03]. Both high environmental and health motives did not result in more positive attitudes nor higher purchase intentions for the product with the EC. H5a and H5b could not be supported.
Figure 4. Three-way Interaction between Food environmental motives and health motives under low, medium and high importance on Product attitudes for EC
26
A 2x2 Factorial MANCOVA was conducted to test hypothesis 6, proposing that exposure to meat substitute products with a claim would generate more positive attitudes (H6a) and higher intentions to purchase the products (H6b) compared to products without a claim, but that this effect would be stronger for products that resemble meat products (vs. not resemble). No statistically significant interaction effect was found between Condition Type and Meat Resemblance on the shared dependent variables, F(2, 366) = 2.64, p = .073, Wilks’ Λ =.986. Nonetheless, changes in mean scores do denote a claim induces more positive attitudes (Mdifference = 0.62) and slightly higher purchase intentions (Mdifference = 0.25) towards the MRproduct than Non-MRproduct (table 2, Appendix F). Due to insignificance, hypothesis 6a
and 6b could still not be supported.
However, a significant weak main effect for Meat Resemblance on the shared dependent variables was found, F(2, 366) = 8.82, p < .001, Wilks’ Λ = .954, η2 = .05. Univariate tests showed significant differences exist between the MRproduct and Non-MRproduct for both Product
attitude, F(1, 367) = 17.56, p < .001, η2 = .05, and Purchase intentions, F(1, 367) = 3.96, p = .047, η2 = .01. A simple effects tests revealed that meat-eaters showed significantly more positive attitudes towards the MRproduct (M = 4.82, SD = 1.05) than the Non-MRproduct (M =
4.36, SD = 1.32), Mdifference = 0.46, p < .001. They also showed significantly higher intentions to purchase the MRproduct (M = 3.33, SD =1.65) than the Non-MRproduct (M = 3.02, SD =1.66),
Mdifference = 0.31, p = .047.
Additional analyses
Since Meat consumption frequency, Meat substitute familiarity, Hedonism and
Dependence influence meat substitute acceptance and meat reduction intentions (Hoek,
27
whether they would exert similar influence on the dependent variables. Meat substitute familiarity significantly, positively effects Product attitudes and Purchase intentions, while Dependence negatively effects both variables. Table 3, Appendix F, provides the results.
Additionally, moderation hypotheses 3 and 4 compared the effects of HC and EC, respectively, to that of NC. In each analysis, one of the claim types was thus left unconsidered. However, claims that match meat-eaters’ particular motives can be expected to be more effective than both of the other claim types, which are not necessarily relevant to them. Furthermore, figure 4 suggests that meat-eaters with high health motives and low environmental motives show less positive attitudes towards the EC than NC, indicating that high health motives may influence the effectiveness of the EC. To simultaneously examine these possible effects, the full model was tested using PROCESS Model 2 on Product
attitudes with the multi-categorical independent variable Conditions Combined (HC, EC, NC)
and moderating variables Food health motives and Food environmental motives. Confirming H3a and H4a, health motives did not moderate the effect of exposure to a HC (vs. NC and EC) on Product attitudes (p = .252) while environmental motives did moderate the effect of exposure to an EC (vs. NC and HC) on Product attitudes. The more importance meat-eaters attached to environmental motives, the more positive attitudes they showed towards the EC (vs. NC and HC) (p < .001). Additionally, the more importance meat-eaters attached to health motives the more negative attitudes they showed towards the EC (vs. NC and HC) (p = .002). Table 4, Appendix F, shows the results.
Conclusion and discussion
In an attempt to encourage the consumption of meat substitute products, this research investigated the extent to which health and environmental product claims could influence attitudes and purchase intentions of Dutch meat-eaters. The moderating roles of consumers’ food choice motives and the products’ meat resemblance were additionally considered.
28
The use of health and environmental product claims on meat substitute products alone is an inadequate strategy to induce attitudinal and intentional behavioral change among general Dutch meat-eaters. While health claims were expected to exert greater persuasiveness than environmental claims, both types appeared ineffective. These outcomes refute recent predictions that these claims and information provision would expand the meat substitute market (Arora et al., 2017; de Boer, de Witt & Aiking, 2016; Slade, 2018). Rather, they confirm Apostolidis and McLeay’s (2016) and Hoek, Luning et al.’s (2011) point that raising meat-eaters’ interest in meat substitutes by communicating health and ethical arguments could be challenging for marketers. The sample’s frequent meat consumption and additional finding that meat dependence negatively affects attitudes and purchase intentions towards meat substitutes, suggests meat consumption has become a habitual behavior, which could counter the persuasiveness of product claims by impeding information processing mechanisms (van ‘t Riet, Sijtsema, Dagevos, & de Bruijn, 2011). Specifically, consumers with strong habits appear less open to information concerning alternative behaviors and tend to minimally use the information available to them (Aarts, Verplanken, & van Knippenberg, 1997; Betsch, Haberstroh, Glockner, Haar, & Fiedler, 2001). Characterized by the process of cognitive
dissonance, these meat-consumers deny arguments that contradict their existing beliefs as
they evoke distressing feelings (Stoll-Kleemann & Schmidt, 2017) and are therefore unlikely to change their behavior (Bergmann, von der Heidt, & Maller, 2010).
Nonetheless, meat-eaters include a small segment for which product claims form a promising avenue for attitudinal change. As postulated by the ELM and consumers’ FCMs, those with strong motives to protect the environment through one’s food choices formed more positive attitudes towards the product with the environmental claim, compared to the absence of a claim and the absence of a claim combined with the health claim. For these consumers, no claim and a health claim are thus of lesser relevance than an
29
environmental claim, whose long-term environmental benefit and dependence on others does not impede their preference. However, no significant results were found for the health claim for consumers with high health motives. Given consumers more often acknowledge the link between meat consumption and health than between the environment (Hoek et al., 2016), those health motivated consumers might have already been aware of this relationship. Opposed to other food and sustainability issues like excessive food packaging, the link between meat consumption and the environment is barely considered (Garnett et al. 2015; Lea & Worsley 2008). Since KDT proposes unawareness is a pre-requisite for information provision to be effective, this might explain why health motivated consumers remained indifferent to the health claim. Their inaction despite them being aware, could additionally imply they hold low self-efficacy beliefs regarding the behavior. Considering their habitual meat consumption, it is feasible that their beliefs to maintain healthy cholesterol-levels through substituting meat, are low. Their belief the behavior will effectively lead to the acclaimed benefit – their response efficacy (Rogers, 1975) – could also be marginal because the products may not be considered healthy/low-fat. Alternately, environmentally motivated meat-eaters expectedly do believe meat substitute consumption could contribute to reduced GHGE. They presumably also hold a strong belief in the collective power to establish this benefit through meat substitution, as collective-efficacy beliefs predict pro-environmental attitudes and behavior (Chen, 2015). Since these beliefs are constituted through social modeling and communication (Bandura, 1997), they might inhibit their habitual meat consumption barrier by sensing others are able to commit to the behavior as well.
Moreover, having both high health and environmental motives did not lead to increased persuasiveness of the environmental claim. Though insignificant, a trend was apparent showing attitudes towards the environmental claim especially increased when health motives were low.
30
Inconsistent with Aschemann-Witzel (2015) and O’Rourke & Ringer (2016), meat-eaters thus do differentiate between self-centered and altruistic motives when making sustainable product choices. Hereby, ambivalence seems to play an important role. Ambivalent attitudes and beliefs are more susceptible to change than those that are not (Povey, Wellens, & Conner, 2001). As such, meat-eaters with ambivalent beliefs about meat are more willing to reduce their meat consumption (Berndsen & van der Pligt, 2004). This possibly also holds true for one’s FCMs, justifying positive attitudinal change was only apparent for those with high environmental food motives, not for those with high health motives. In fact, health motivated meat-eaters showed more negative attitudes towards the environmental claim compared to the absence of a claim and the absence of a claim combined with the health claim. Hence, the effectiveness of environmental claims on meat substitute products seems to be driven by collective food concerns and undermined by individual health concerns. This confirms Kim’s (2011) finding that individual concerns inhibit environmentally-concerned consumption attitudes while collective concerns motivate them.
After all, the greatest driver for meat substitute acceptance seems to be the products’ meat resemblance, which directly enhances meat-eaters’ attitudes and purchase intentions. As the success of food substitutes is determined by fulfillment of consumer needs (van Trijp & Steenkamp, 1998; van Trijp & van Kleef, 2008), health and sustainability benefits are evidently not primarily desired in a meat substitute. Evinced by the ELM, the peripheral cues surrounding the claims – particularly relevant for meat-eaters with strong habits – therefore receive greater attention. Given meat-like products evoke familiar, positive associations, meat resemblance thus remains the overarching driver for meat-eaters’ meat substitute acceptance (Hoek et al., 2011; Elzerman et al., 2013). Nonetheless, claims should not be completely disregarded. Although insignificant, a claim yielded more positive attitudes and higher
31
purchase intentions for a product mimicking meat, suggesting proof for moderate incongruity in line with the findings of Wansink (1994) and Hoek et al. (2017).
Implications
These results imply that along with the products’ meat resemblance, consumers’ FCMs form novel additions to predicting meat substitute acceptance directly or indirectly through product claims. Hereby, the validity of the FCQ, ELM and schema theory has been extended.
The KDT, however, appeared to hold less predictive power. When applying the theory to healthy and sustainable consumption behaviors it might therefore benefit from including individuals’ efficacy beliefs, common among health information processing models. Additionally, habitual behavior should be acknowledged as a boundary condition to KDTs validity. Hereby, the research extends the literature on education through product claims by corroborating Rothman, Sheeran and Wood’s (2009) finding that for habitual dietary behaviors, the effectiveness of information provision is limited. In applying KDT to meat substitute acceptance, it is therefore essential to also address the barriers that inhibit ceasing meat consumption as a habitual behavior (e.g. its exclusive taste). This affirms de Boer and Aiking’s (2017) proposal that in pursuing low meat diets, frames should be developed focusing on both push factors away from meat-eating and pull factors towards meat alternatives.
As such, to reach consumers with strong meat-eating habits, marketers should focus on peripheral cues by enhancing meat resemblance and ignore claims. Nonetheless, for specifically environmentally-friendly consumers the use of environmental claims is strongly recommended. Retailers should, however, ensure health motivated meat-eaters remain unexposed to these claims. Given they are – environmentally motivated meat-eaters alike –
32
not specifically appealed to meat substitutes’ health benefit, investing in health claims is ill-advised.
Limitations and future research
Due to its limited scope and experimental design, the research is subject to limitations. Firstly, only one type of health and environmental claim was studied, and taste claims remained unconsidered. Future research should investigate the potential greater persuasiveness of meat substitutes’ meat-like taste and protein and animal welfare benefits, as many meat-eaters believe meat is fundamental to their protein-intake (Arora et al., 2017) and animal-related arguments strongly influence meat consumption reduction (Cordts et al., 2014).
Given brand familiarity enhances brand trust (Herbst, Hannah, & Allan, 2013; Gretry, Horváth, Belei, & van Riel, 2017) and information is disregarded with untrusted sources (Verbeke, 2008), the unfamiliarity of Garden Gourmet may have affected the results. Future research may therefore benefit from considering source trust through manipulation of stimuli including hallmark-logos of verified, trusted, third-parties/sources.
Furthermore, the research assumed participants were unaware of the consequences of meat consumption, yet this was not verified through measurement nor was it analyzed as dependent variable. Consequently, no conclusions can be drawn concerning product claims’ potential of enhancing awareness, proposing valuable venues for future research. Similarly, participants were only exposed to the claims once. Given repeated exposure to a message can enhance one’s attitude towards it (Zajonc, 1968), longitudinal studies are indispensable to make valid inferences about the claims’ persuasiveness.
Lastly, future interventions should focus on stimulating meat-eaters to try the products as the research confirmed familiarity with meat substitutes is key in establishing meat-eaters’
33
acceptance (Hoek et al., 2011). If manufacturers subsequently succeed in meeting meat-eaters taste expectations, the major barriers to meat reduction and the acceptance of meat substitutes are eliminated. This would offer a fruitful first step in breaking their habitual behavior, being more receptive to product claims and eventually establishing meat consumption reduction in the Netherlands.
34
References
Aarts, H., Verplanken, B., & Knippenberg, A. V. (1997). Habit and information use in travel mode choices. Acta Psychologica, 96, 1-14.
Abrahamse, W., Steg, L., Vlek, C., & Rothengatter, J. A. (2005). A review of intervention studies aimed at household energy consumption. Journal of Environmental Psychology,
25(1), 273-291. doi:10.1016/j.jenvp.2005.08.002
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human
decision processes, 50(2), 179-211. doi:10.1016/0749-5978(91)90020-T
Apostolidis, C., & McLeay, F. (2016). Should we stop meating like this? Reducing meat consumption through substitution. Food Policy, 65, 74-89.
doi:10.1016/j.foodpol.2016.11.002
Arora, A. S., Bradford, S., Arora, A., & Gavino, R. (2017). Promoting Vegetarianism through Moralization and Knowledge Calibration. Journal of Promotion Management, 23(6), 889-912. doi:10.1080/10496491.2017.1323263
Aschemann-Witzel, J. (2015). Consumer perception and trends about health and
sustainability: trade-offs and synergies of two pivotal issues. Current Opinion in Food
Science, 3(1), 6-10. doi:10.1016/j.cofs.2014.08.002
Babbie, E.R. (2010). The Practice of Social Research. Belmont, CA: Wadsworth Cengage. Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191-215. doi:10.1037/0033-295X.84.2.191
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: PrenticeHall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: WH Freeman and Company.
35
conjoint study of Danish, Finnish and American consumers' perception of functional foods. Appetite, 40(1), 9-14. doi:10.1016/S0195-6663(02)00171-X
Bergkvist, L. (2015). Appropriate use of single-item measures is here to stay. Marketing
Letters,
26(3), 245-255. doi:10.1007/s11002-014-9325-y
Bergmann, I., von der Heidt, T., & Maller, C. (2010, July 15). Cognitive dissonance and
individuals' response strategies as a basis for audience segmentation to reduce factory farmed meat consumption. Paper presented at International Nonprofit and Social
Marketing conference (INSM): conference proceedings, Brisbane. Griffith, Griffith University.
Berndsen, M., & Van Der Pligt, J. (2004). Ambivalence towards meat. Appetite, 42(1), 71-78. doi:10.1016/S0195-6663(03)00119-3
Betsch, T., Haberstroh, S., Glöckner, A., Haar, T., & Fiedler, K. (2001). The effects of routine strength on adaptation and information search in recurrent decision
making. Organizational behavior and human decision processes, 84(1), 23-53. doi:10.1006/obhd.2000.2916
Campbell-Arvai, V., Arvai, J., & Kalof, L. (2014). Motivating sustainable food choices: The role of nudges, value orientation, and information provision. Environment and
Behavior, 46(4), 453-475. doi: 10.1177/0013916512469099
Carlson, L., Grove, S.J., Kangun, N. (1993). A content analysis of environmental advertising claims: a matrix method approach. Journal of Advertising, 22(1), 27-39.
Chen, M. F. (2011). The joint moderating effect of health consciousness and healthy lifestyle on consumers’ willingness to use functional foods in Taiwan. Appetite, 57(1), 253-262. doi:10.1016/j.appet.2011.05.305
36
model: Which more effectively explains people's self-reported proenvironmental behavior?. Journal of Environmental Psychology, 42, 66-75.
doi:10.1016/j.jenvp.2015.02.002
Cheong, Y., & Kim, K. (2011). The interplay between advertising claims and product categories in food advertising: A schema congruity perspective. Journal of Applied
Communication Research, 39(1), 55-74. doi:10.1080/00909882.2010.536845
Cho, Y. N. (2015). Different shades of green consciousness: The interplay of sustainability labeling and environmental impact on product evaluations. Journal of business
ethics, 128(1), 73-82. doi:10.1007/s10551-014-2080-4
Cordts, A., Nitzko, S., & Spiller, A. (2014). Consumer response to negative information on meat consumption in Germany. International Food and Agribusiness Management
Review, 17(1), 83-106.
Corrin, T., & Papadopoulos, A. (2017). Understanding the attitudes and perceptions of vegetarian and plant-based diets to shape future health promotion
programs. Appetite, 109(1), 40-47. doi:10.1016/j.appet.2016.11.018
Dagevos, H., & Voordouw, J. (2013). Sustainability and meat consumption: is reduction realistic? Sustainability: Science, Practice and Policy, 9(2), 60-69.
doi:10.1080/15487733.2013.11908115
Dean, M., Lampila, P., Shepherd, R., Arvola, A., Saba, A., Vassallo, M., Lähteenmäki, L. (2012). Perceived relevance and foods with health-related claims. Food Quality and
Preference, 24(1), 129–135. doi:10.1016/j.foodqual.2011.10.006
de Boer, J., Schösler, H., & Boersema, J. J. (2013). Climate change and meat eating: an inconvenient couple?. Journal of Environmental Psychology, 33, 1-8.
doi:10.1016/j.jenvp.2012.09.001
