The role of message framing and construal level on the relationship between health consciousness and purchase intention of organic products
Student name: Roos Rabenberg Student number: 11178728 Version: Final
Date of submission: 27-01-2017 Thesis supervisor: Drs. Frank Slisser
MSc in Business Administration - Marketing Track University of Amsterdam - Amsterdam Business School
Statement of originality
This document is written by Roos Rabenberg who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of contents
Statement of originality 2
Abstract 4
1. Introduction 5
2. Literature review 8
2.1. Health consciousness and purchase intention of organic products 8
2.1.1. Attitude towards the advertisement 10
2.1.2. Attitude towards organic products 11
2.2. Message framing 12
2.3. Construal level 19
2.4. Conceptual framework 23
3. Data and method 24
3.1. Sample 24
3.2. Manipulation 25
3.3. Measurement of the variables 27
3.4. Statistical procedure 29
4. Results 31
4.1. Manipulation checks 31
4.2. Correlation analysis 32
4.3. Serial mediation 33
4.4. Moderation effects of message framing 35
4.5. Moderation effects of construal level 37
5. Discussion 39
5.1. Serial mediation 40
5.2. The moderating role of message framing 41
5.3. The moderating role of construal level 43
6. Conclusion 44
6.1. Summary 45
6.2. Practical and theoretical implications 45
6.3. Limitations and further research directions 46
Acknowledgements 50
References 51
Abstract
In the upcoming years, organic product consumption is expected to grow at an exceptional rate (Meredith & Willer, 2014). Despite this, organic products represent a small fraction of agricultural sales worldwide at present. As a result, researchers and businesses have recognized the need for effective marketing strategies to increase the sales of organic products. However, little research has examined the effectiveness of organic product advertisements. This study aims to contribute to existing literature by investigating the direct and indirect relationship between health consciousness and purchase intention. The indirect relationship is assessed through consumers' attitude towards the advertisement and attitude towards organic products. Specifically, this study adds to existing literature by including two concepts of persuasion strategies, namely message framing and construal level. A cross-sectional experimental Internet-based survey is distributed dividing 258 respondents into five conditions (one control group and four different types of advertisements). The findings do not find the support that construal level influences the relationship directly nor indirectly between health consciousness and purchase intention of organic products. The study does find support for the assumption that message framing influences the relationship indirectly. Specifically, this study finds that when low health-conscious consumers are exposed to a loss-framed message, consumers have more positive attitudes towards organic products. As a consequence, positive attitudes towards organic products can lead to higher purchase intentions of organic products. These results have implications for both researchers and marketing practitioners.
1. Introduction
Today the public scrutiny of the food supply chain is intensified due to several factors. Consumers are concerned about environmental issues, food safety, animal welfare and the ethical conditions involved in the food production process (Opara & Mazaud, 2001; Sheth, Sethia, & Srinivas, 2010). Because of these growing concerns, organic product consumption has grown at an exceptional growth rate of 7.4% in 2014 (Meredith & Willer, 2014).
Consumer motivations to purchase organically grown products indicate that they perceive organic products as a parameter of quality, environmental-friendly (Schifferstein & Oude Ophuis, 1998), safe (Schifferstein & Ophuis, 1998), tasty (Zanoli & Naspetti, 2002; Lea & Worsley, 2005) and satisfying (Zanoli & Naspetti, 2002; Chen, 2009). The main driver of organic product consumption, however, is the consumers' level of health consciousness (Schifferstein & Oude Ophuis, 1998; Vukasovič, 2016; Williamson, 2007). Health consciousness can be defined as ‘the degree of readiness to undertake healthy actions’ (Schifferstein & Oude Ophuis, 1998, p.122) and consumers who are health-conscious are found to be regular organic product buyers (Schifferstein & Oude Ophuis, 1998). Yet, even though organic production is growing rapidly over the past years, it still represents only a small fraction of agricultural sales worldwide (Scheer & Moss, n.d.).
It is generally known that many people forgo organic product consumption due to its relatively higher pricing and its limited availability (Pearson, Henryks, and Jones, 2011). However, Daugherty and Brase (2010) explain an underlying concept that helps to explain why consumers refrain from purchasing organic products. This underlying concept is the combination of the consumers' level of health consciousness and delay discounting (Daugherty & Brase, 2010). This means that, in health related behavior, low-involved consumers often have trouble to forgo immediate benefits over long-term sacrifices
(Daugherty & Brase, 2010). For example, consumers who are not ready to undertake actions to enhance their health may prefer to pay a lower price for non-organic products than to pay a price premium for organic products.
To increase organic product consumption, it may be that consumers with low levels of health consciousness need promotional messages to draw their attention. However, little research has been done on the persuasiveness of organic product advertisements to encourage consumers with different levels of health consciousness to buy organic products. In response to the call of action by Gallagher and Updegraff (2012), this research will add value to existing literature by testing prior assumptions in the context of organic food products of two concepts that are used in persuasive communication, namely message framing and construal level. Message framing states that people’s preferences for certain options are based on the evaluation of the message, such as gain or loss framing (Levin, Schneider, & Gaeth, 1998; Tversky & Kahneman, 1981). Construal level focuses on temporal distance, which describes the consumers’ mental interpretation of distance in time, such as a present- or a future-focused message (Trope & Liberman, 2010). This will be the first study to examine the influence of message framing (gain- or loss-focused message) and construal level (present- or future-framed message) in advertisements of organic products on the relationship between health consciousness and purchase intention of organic products. Specifically, this study tests whether there are differences in purchase intention when low and highly health-conscious consumers are exposed to different persuasive communications. Therefore, this current study is designed to answer the research question: How do message framing and construal level influence the relationship between health consciousness and purchase intention of organic products?
product industry. Theoretically, this study builds on message framing and construal level theory and explores how marketers can enhance communication persuasiveness when targeting consumers with different levels of health consciousness. As the persuasiveness of advertisements may increase the purchase intention of products, this study can increase the understanding of when and how certain messages drive purchasing behavior of organic products. From a managerial perspective, this study encourages managers to implement strategies that increase organic product consumption. More importantly, this study urges businesses to contribute to a consumption conscious-society a protected production environment, and a world with safer and more honest products to reduce the public's scrutiny. Also, this study can guide marketers decisions to create the right consumer mindset through construal level and the message framing to target specific consumers’ with low and high levels of health consciousness.
The first section reviews literature to investigate the relationship between health consciousness and purchase intention. Indirect effects that influence this relationship are discussed, namely the influence of consumers' attitude towards advertisements and consumers' attitude towards organic products. More importantly, the effect of message framing and construal level on the relationship between health consciousness and purchase intention is investigated. After that, the conceptual framework is stated together with an overview of the hypotheses that are derived from the literature review. The second section provides the methodological background for testing the hypotheses. The third section discusses how the advertisements are manipulated and reveals the results of the study. The last section summarizes and discusses the implications and future research directions.
2. Literature review
The following section investigates the relationship between health consciousness and consumers’ purchase intention of organic products. Two mediators influence this relationship, namely consumers’ attitude towards the advertisement and consumers’ attitude towards organic products. More importantly, two moderators identify the conditions in which these effects can occur, namely message framing and construal level. This is investigated to answer the main research question: ‘How do message framing and construal level influence the relationship between health consciousness and purchase intention of organic products?’
2.1. Health consciousness and purchase intention of organic products
According to the European Commission (2014, p.1), organic products are produced by farmers with an ‘agricultural system that seeks to provide you, the consumer, with fresh, tasty and authentic food while respecting natural life-cycle systems’. In line with the European Commission (2014), Lea and Worsley (2005) state that the production involved of organic products needs to respect the environment and soil where genetic modification and artificial chemicals, such as pesticides and fertilizers, are not allowed. In other words, organic products are produced while respecting humans, animal welfare, and the environment to provide quality products with natural additives, which is the definition that is used throughout this study. In this study, products that are not organic are referred to as non-organic products. Previous research finds that the primary motive of organic product consumption is driven by consumers’ health consciousness (Schifferstein & Oude Ophuis, 1998; Vukasovič, 2016; Williamson, 2007). Health consciousness can be defined as ‘the degree of readiness to undertake healthy actions’ (Schifferstein & Oude Ophuis, 1998, p.122). Consumers who are highly conscious of their health are aware of their well-being and express health-promoting
behaviors by putting an effort in their health to enhance their quality of life and to prevent illness (Gould, 1988).
Several studies indicate a positive relationship between being health consciousness and consumers’ purchase intention of organic products (Schifferstein & Oude Ophuis, 1998; Cheung, Lau, & Lam, 2015; Vukasovič; 2016). In this study, purchase intention is referred to as the consumers' purchase intention of organic products specifically. Schifferstein and Oude Ophuis (1998) find that health-conscious consumers are often organic product buyers. According to Vukasovič (2016), the main attributes that drive health-conscious consumers to purchase organic products are freshness, quality, nutritional value, safety and whether or not the product is organically grown. However, health conscious consumers do not always have a higher purchase intention of organic products from all product categories. For example, Van Doorn and Verhoef (2015) found in their study that health motivation only has a significant effect on organic product purchase in certain categories. Specifically, they find that health motivation drives purchase intention of organic products in categories that do not compete frequently using price promotions. This is because when business do compete on price, consumers are more inclined to make decisions based more on the price and discounts, rather than on attributes of organic products (i.e. freshness, quality, nutritional value, safety and whether or not the product is organically grown) (Van Doorn & Verhoef, 2015). As organic products are often available for a price premium, consumers are more inclined to purchase organic products from categories that do not compete extensively on price (Van Doorn & Verhoef, 2015). Thus, consumers who are health-conscious believe that organic products contribute to their health, which drives their purchase intentions of organic products only in categories with low frequencies of price promotions (Van Doorn & Verhoef, 2015).
2.1.1. Attitude towards the advertisement
To better understand the positive relationship between health consciousness and purchase intention, the mechanisms by which health consciousness causally influences purchase intention of organic products are investigated. Whether a consumer decides to purchase a product is not solely dependent on the belief of product attributes as Fishbein and Ajzen (1975) have suggested. Indeed, Mitchell and Olson (1981) and Shimp (1981) have introduced the construct of consumers' attitude towards the advertisement and find that consumers’ purchase intentions are largely determined by their attitude towards the advertisement (Shimp, 1981; Mitchell & Olson, 1981). An attitude, in general, can be defined as the consumers’ evaluation of an entity in favor or disfavor (Cheung et al., 2015). In this study, the consumers’ attitude towards the advertisement is defined as the consumers’ evaluation of a particular advertising message at a specific point in time (Mackenzie & Lutz, 1989; Fishbein & Ajzen, 1975).
The studies of Mitchell and Olson (1981) and Shimp (1981) have investigated the construct of attitude towards the advertisement together with other measures of advertising effectiveness. For example, Shimp (1981) studied the affect transfer hypothesis. This includes consumers' attitude towards the product or brand in the relationship between consumers' attitude towards the advertisement and purchase intention or brand choice (Shimp, 1981). This can be explained in the way that the positive attitude towards the advertisement can transfer towards the consumers' attitude towards the product or brand, which in turn leads to higher purchase intentions (Shimp, 1981). More studies suggest that consumers' attitude towards the advertisement can lead to more positive attitudes towards the product or brand. Therefore, the next section investigates the influence of consumers' attitude towards (organic) products.
2.1.2. Attitude towards organic products
As mentioned, next to consumers’ attitude towards the advertisement, another dimension that influences purchase intention is consumers’ attitude towards the product/brand (Tsai, 2007). Whether a consumer decides to purchase organic products specifically, seems to be influenced by their attitude towards organic products (Iranto, 2015; Cheung et al., 2015). Especially health-conscious individuals tend to believe organic products affect their health. They believe organic products are healthier compared to conventional products (Michaelidou & Hassan, 2008; Lea & Worsley, 2005). This belief suggests that health-conscious consumers are more likely to have positive attitudes towards organic products (Chen, 2009; Schifferstein & Oude Ophuis, 1998). According to Paul and Rana (2012), consumers’ purchase intention of organic products is significantly influenced by the consumers’ positive attitudes towards organic products. In their research, they found that households who perceive organic products as healthy have higher purchase intentions of organic products than non-organic products (Paul & Rana, 2012).
Previous studies find that that consumers' purchase intention is higher when consumers' attitudes towards the advertisements are more positive, which influences consumers to form more positive attitudes towards the advertised product or brand (Gardner, 1985; Sallam & Algammash, 2016). This means that when a consumer evaluates an advertisement positively, he/she is also more likely to evaluate the product more positively, which increases the likelihood of purchase intention (Ajzen, 1991; Chen, 2009; Iranto, 2015). To investigate the serial mediation of consumers’ attitude towards the advertisement and the product, this study hypothesizes:
H1: The positive relationship between health consciousness and purchase intention of organic products is mediated by consumers attitude towards the advertisement and
consumers attitude towards the advertisement in serial.
2.2. Message framing
One method that advertisers use to influence consumer attitudes and behaviors through advertisements is message framing (Chang, Zhang, & Xie, 2015). Message framing is referred to as gain/loss framing and states that people’s preferences for certain options are based on the evaluation of the message (Levin, Schneider, & Gaeth, 1998; Tversky & Kahneman, 1981). There are two types of message framing1. Gain-framed messages concern the pleasant consequences resulting from pursuing an action (Chang, 2008; Tversky & Kahneman, 1981).
Advertisers use gain-framed messages by highlighting positive consequences by promoting the benefits of using a certain product (e.g. showing the happy life of organic chickens) (Harrison & Rutstrom, 2008). Loss-framed messages, however, highlight the negative consequences derived from pursuing an action and is negatively framed (Chang, 2008; Tversky & Kahneman, 1981). A loss-framed message is one where the advertiser demonstrates what the consumer stands to lose if he/she fails to comply with the message (Idson, Liberman, & Higgins, 2000). Through warnings, advertisements can discourage behavior by demonstrating an undesirable outcome, often referred to as a loss (e.g. showing unsanitary conditions of non-organic chickens) (Harrison & Rutstrom, 2008).
2.2.1. The moderating effect on purchase intention
Studies on message framing effects regarding which type of framing is the more persuasive show that results are far from consistent (For meta-analyses see Levin et al., 1998; O’Keefe & Jensen, 2006; and Gallagher & Updegraff, 2012). Some studies show that gain-framed
1
This study focuses on a particular type of framing effect defined by Levin, Schneider, and Gaeth (1998) as ‘goal framing’. Goal framing uses the positive frame to describe the gains associated with performing the
messages are more persuasive to encourage certain behavior (e.g., Maheswaran & Meyers-Levy, 1990; Rothman, Salovey, Keough, & Martin, 1993, Levin & Gaeth, 1988). Other studies show that loss-framed messages have a greater persuasive power (e.g., Block & Keller, 1995; Kalichman & Coley, 1995; Maheswaran & Meyers-Levy, 1990; Meyerowitz & Chaiken, 1987).
Well-known examples that showed a positive effect of message framing on behavior are the studies of Levin and Gaeth (1988) and MacInnis and Jaworski (1989). Levin and Gaeth (1988) showed that when messages are framed positively (versus negatively), more favorable associations are generated in the consumer mindset, which results in more positive product evaluations (Levin & Gaeth, 1988). Specifically, the study showed that when ground beef was described as 25% fat or as 75% lean, consumers seem to prefer the former option (Levin & Gaeth, 1988). MacInnis and Jaworski (1989) explain why positively framed messages enhance the consumers’ attitudes towards the advertisement and, in turn, result in a more positive evaluation of the product. They explain this through a process called ‘pure affect transfer’. When consumers are exposed to messages that make them feel positive, these positive emotions are then transferred to the advertisement message and towards the advertised product or brand (MacInnis & Jaworski, 1989). In turn, negatively framed messages evoke negative emotions, which do not result in enhanced attitudes (MacInnis & Jaworski, 1989).
Several widely known studies have shown the impact of loss-framed messages (Meyerowitz & Chaiken, 1987; Homer & Yoon, 1992). In their research, Meyerowitz and Chaiken (1987) demonstrate that a loss-framed (negative) message was more persuasive in generating more positive attitudes, intentions, and behaviors by showing the risks of not engaging in breast self-examination (BSE) (‘You can lose several potential health benefits by
failing to spend only five minutes each month doing BSE’). In contrary, a message was rated as less persuasive when framing the same information as a gain (‘You can gain several potential health benefits by spending only five minutes each month doing BSE’). Thus, Meyerowitz and Chaiken (1987) find that individuals are generally conservative in seeking gains and are more willing to engage in risky behavior to avoid losses. Homer and Yoon (1992) argue in their study that negatively framed messages are linked to the so-called ‘negativity effect’, i.e. where ‘negative information receives greater weight than positive information’ (p.20). They study this while advertising a mouthwash product and find that the influence of brand cognitions on brand attitudes was stronger for negatively framed messages (Homer & Yoon, 1992). Homer and Yoon (1992) find evidence for the fact that when a message is framed negatively, it perceived as more important and therefore contributes to the positive formation of consumers’ attitude towards the advertisement. In contrast, they argue that positive information is less likely to be remembered and to be translated into positive attitudes towards the advertisement.
According to O’Keefe and Jensen (2006), the persuasiveness of message framing is not only dependent on the product that is being advertised in several studies, it also depends on whether the desired behavior the study investigates is disease-detection behavior or disease-prevention behavior. According to O’Keefe and Jensen (2006), gain- and loss-framed messages do not differ in persuasiveness when encouraging disease-detection behavior (i.e. breast self-examination, see Meyerowitz & Chaiken, 1987). In contrast to disease-detection behavior, when encouraging disease-prevention, gain-framed messages are significantly more persuasive than loss-framed messages (O’Keefe & Jensen, 2006). Disease-prevention programs attempt to reduce the development of chronic diseases and are often addressed together with health promotion programs (Rural Health Information, n.d.). Considering that
consumers’ health consciousness is one of the main drivers of organic product consumption, advertisements encouraging organic product consumption can be considered encouraging disease-prevention like health promotion programs (Schifferstein & Oude Ophuis, 1998).
As mentioned, O’Keefe and Jensen (2006) argue that gain-framed messages are more persuasive when encouraging disease-prevention, such health-related behavior like purchasing organic products. However, rather than categorizing message framing effectiveness on the type of behavior, several studies argue that argue that the effectiveness of message framing depends on the individual differences in the particular belief or motivation. (Latimer, Salovey, & Rothman, 2007, Gerend & Shepherd, 2013).
Gerend and Shepherd (2013) showed that, in the context of calcium-intake, consumers who predominantly prefer to focus on benefits are more responsive to gain-framed messages. In turn, consumers who predominantly focus on costs and negative outcomes are more response to loss-framed messages (Gerend & Shepherd, 2013). Sherman, Mann, and Updegraff (2006) have found similar results in encouraging flossing behavior. Given that health-conscious consumers are more motivated to purchase organic products, it is expected that their purchase intention is higher when exposed to a gain-framed message, as they predominantly prefer the focus on benefits rather than negative outcomes. In turn, low health-conscious consumers are expected not to be specifically motivated or interested in organic products. Building on the findings of Meyerowitz and Chaiken (1987), this study expects that individuals who are exposed to a negative outcome, are willing to reduce, avoid or overcome the framed loss by adapting their behavior. In other words, this study expects that consumers with low levels of health consciousness are encouraged to purchase organic products when exposed to an advertisement with loss-framed messages (Gerend & Shepherd, 2013; Sherman, Mann, & Updegraff, 2006). To confirm this line of reasoning, this study
hypothesizes:
H2: The relationship between health consciousness and consumers’ purchase intention of organic products is moderated by message framing. Specifically,
H2a: Consumers that are low in health consciousness have higher purchase intentions of organic products when the consumer is exposed to a loss-framed message.
H2b: Consumers that are high in health consciousness have higher purchase intentions of organic products when the consumer is exposed to a gain-framed message.
2.2.2. The moderating effect on attitude towards advertisements and organic products
Message framing does not only affect the relationship between health consciousness and purchase intention of organic products, but also the consumers’ attitude towards the advertisement. As mentioned before, whether or not consumers intend to purchase (organic) products, may depend on their attitude towards the advertisement and the advertised product (Shimp, 1981; Gardner, 1985; Sallam & Algammash, 2016).
2.2.2.1. The moderating effect on attitude towards advertisement
This study aims to investigate whether message framing influences the relationship between health consciousness and consumers' attitude towards the advertisement. In other words, this study explores whether consumers with different levels of health consciousness evaluate advertisements promoting an organic product different when the advertisement is gain- or loss-framed. Previous research has shown that consumers who are health conscious have more positive attitudes towards organic products in general (Schifferstein & Oude Ophuis, 1998; Cheung et al., 2015; Vukasovič; 2016). Schifferstein and Oude Ophuis (1998) argue that health conscious consumers are often organic product buyers. Therefore, these
consumers are expected to have a higher understanding of organic products and are expected to have more positive attitudes towards advertisements promoting organic products. Otherwise, consumers with lower levels of health consciousness may be less involved with and have a weaker understanding of organic products. This is in line with Glass (2007), who find that high-involved consumers have more positive attitudes towards the advertisement and that low-involved consumers have more negative attitudes towards the advertisement. However, whether consumers have positive attitudes towards the advertisement does not only depend on whether the consumers are health conscious or not.
As mentioned, message framing can help explain why certain consumers have higher purchase intentions than others. To extend the line of reasoning that low (vs. high) health-conscious consumers have higher purchase intentions when exposed to loss-framed (vs. gain-framed) messages, this study investigates whether the same is true for consumers' attitude towards the advertisement. As Gerend and Shepherd (2013) showed, consumers who predominantly prefer to focus on benefits are more responsive to gain-framed messages. Given that consumers that are health consciousness are more interested in organic products, it is expected that they also have more positive attitudes when exposed to a gain-framed advertisement of an organic product. In turn, consumers who predominantly focus on costs and negative outcomes are more response to loss-framed messages (Gerend & Shepherd, 2013). Thus, low health-conscious consumers are not specifically motivated or interested in organic products but are expected to be willing to reduce, avoid or overcome the negative feeling when exposed to a loss-framed advertisement of an organic product (Gerend & Shepherd, 2013). Specifically, when consumers with low levels of health consciousness are exposed to the disadvantages of non-organic products, they are expected to be willing to reduce, avoid or overcome the negative feeling and alter their behavior by adapting their
attitude towards the advertisement. To confirm this line of reasoning, this study hypothesizes the following:
H3: The relationship between health consciousness and consumers’ attitude towards the advertisement is moderated by message framing. Specifically,
H3a: Consumers that are low in health consciousness have more positive attitudes towards the advertisement when the consumer is exposed to a loss-framed message. H3b: Consumers that are high in health consciousness have more positive attitudes towards the advertisement when the consumer is exposed to a gain-framed message. 2.2.2.2. The moderating effect on attitude towards organic products
According to several studies, when a consumer evaluates an advertisement positively, he/she is also more likely to evaluate the product more positively (Ajzen, 1991; Chen, 2009; Iranto, 2015). In addition, whether consumers have positive attitudes towards the advertised product depends on whether the consumers are interested in the advertised product (Iranto, 2015). In the context of organic products this could translate into the following: consumers who are interested in organic products, such as consumers with high levels of health consciousness, are expected to have a more positive attitude towards organic products. Given that consumers that are health consciousness are more interested in organic products, it is expected that they predominantly prefer to focus on benefits of organic products, and not on the costs or the negative outcomes (Gerend & Shepherd, 2013). In turn, building on the study of Gerend and Shepherd (2013), consumers who have a low-level of health-consciousness are expected to focus on the costs and the negative outcomes of organic products, but are expected to be willing to reduce, avoid or overcome the negative feeling when exposed to a loss-framed advertisement. To extend the line of reasoning that low (vs. high) health-conscious consumers have higher purchase intentions of organic products and more positive attitudes
towards the advertisement when exposed to loss-framed (vs. gain-framed) messages, this study investigates whether the same is true for consumers' attitude towards the organic products. Specifically, this study hypothesizes the following:
H4: The relationship between health consciousness and consumers’ attitude towards organic products is moderated by message framing. Specifically,
H4a: Consumers that are low in health consciousness have more positive attitudes towards organic products when the consumer is exposed to a loss-framed message. H4b: Consumers that are high in health consciousness have more positive attitudes towards organic products when the consumer is exposed to a gain-framed message.
2.3. Construal level
Whereas message framing focusses on describing consequences of actions in gain and losses (Tversky & Kahneman, 1981), construal level theory focuses on several types of psychological distance that is created in the consumer’s mind. (Trope & Liberman, 2010). This study focuses on one of the most common dimensions of construal level, namely temporal distance. Temporal distance describes the consumers’ mental interpretation of distance in time (Trope & Liberman, 2010). This means that as temporal distance decreases in the consumer mindset, construals become more concrete, and as the level of concreteness increases, so too would the time frame in which the action is envisaged (Trope & Liberman, 2010).
When consumers are exposed to high construal, advertisements focus on the future consequences of the consumers’ actions (Trope & Liberman, 2010). Future-focused messages encourage abstract thinking which contains less concrete details of the action performed and its immediate context than present-focused messages (Förster, Friedman, & Liberman, 2004).
Thus, when consumers are exposed to a present focus, more detailed and concrete consequences of the consumers’ actions are demonstrated (Förster et al, 2004). To illustrate: a person thinking about performing a healthy behavior a year from now might think about in terms of more abstract goals, such as thinking about reasons why it is important to do so. Whereas a person thinking about performing healthy behavior today might construe it in terms of more concrete goals, such as exercising or purchasing organic products. White, MacDonnell, and Dahl (2011) have confirmed these findings, in the context of recycling behavior. It might be that the findings of White et al. (2011) can be generalized to health behavior, as recycling behavior can make differences in protecting health (Wright, 2011). 2.3.1. The moderating effect on purchase intention
Forster, Friedman, and Liberman (2004) argue that the level of construal influences consumers’ information processing and decision-making. Therefore, construal level can act as an important variable when predicting purchase intention through advertisements. Previous research shows that the present focus makes health outcomes seem more concrete and evokes a greater sense of behavior intention (Chandran & Menon, 2004). In turn, the future focus makes health outcomes seem more abstract and therefore seem less likely to change immediate behavior and intention (Chandran & Menon, 2004).
Walsh (1995) suggests that behavioral change due to a present- or future- framed message depends on the predominant characteristics of the consumer itself. In other words, consumers who are present-oriented are more prone to purchases to satisfy immediate needs (Walsh, 1995) In turn, future-oriented consumers take into account future consequences (Walsh, 1995). This can be explained with the concept ‘delay discounting’.
Delay discounting is defined as the ‘willingness to postpone receiving an immediate reward to gain additional benefits in the future’ (Daugherty & Brase, 2010, p.203). Daugherty
and Brase (2010) studied the effects of delay discounting, construal level and health-related behavior on purchase intention. Specifically, they provide evidence that when exposed to a future framed message, consumers who express their concerns about their health are more likely engage in healthy behaviors compared to a present frames message (Daugherty & Brase, 2010). This is an important application of construal level in the context of health-related behavior, as the consequences of this behavior often only become apparent in the future (Daugherty & Brase, 2010). To illustrate: it may be that consumers with a low-level of health consciousness do not take into account the long-term effects such as environmental pollution and poor living conditions of livestock when exposed to an advertisement that focuses on the future consequences. However, it may be that when the advertisement focuses on the consequences in the present, the consumer has a higher purchase intention than when exposed to a future-focused message.
The underlying reasoning of this is that proximal consequences are more concrete than future-framed messages, which increases the intention to reduce the risk (Chandran & Menon, 2004). In turn, highly health-conscious consumers may forego immediate gratification behaviors (such as purchasing non-organic products) in order reap rewards in the future (enhanced environmental conditions and quality livestock), as they are more motivated to focus on the future benefits of their behaviors. To confirm this line of reasoning, this study hypothesizes the following:
H5: The relationship between health consciousness and consumers’ purchase intention of organic products is moderated by construal level. Specifically,
H5a: Consumers that are low in health consciousness have higher purchase intentions of organic products when the consumer is exposed to a present-focused message.
H5b: Consumers that are high in health consciousness have higher purchase intentions of organic products when the consumer is exposed to a future-focused message.
2.3.2. The moderating effect on attitude towards advertisements and organic products
Limited studies have explored construal level and consumer individual differences in predicting attitudes. One of these studies that focus on this area proposes that ‘people have individual differences in how they perceive the world vis-à-vis the past, present, or future’ (Martin, Gnoth & Strong, 2009, p. 7). Here, consumers are categorized in different levels of temporal orientation, such as future and present orientations (Martin et al., 2009). Martin, et al. (2009) find evidence that consumers who are future oriented have more positive attitudes towards advertisements, and in turn more positive attitudes towards the product, when the message focuses on the future (Martin et al., 2009). In contrast, consumers who are future oriented have more negative attitudes towards advertisements and products when the message focuses on the present (Martin et al., 2009).
2.3.2.1. The moderating effect on attitude towards advertisements
This study extends on the findings of the study of Martin et al. (2009) by investigating whether health-conscious consumers, who generally take into account the future consequences of their behavior, respond more favorably to advertisements that focus on future consequences. In turn, when health consumers are exposed to messages with a present focus, it is expected that the consumer has a more negative attitude towards the advertisement. Therefore, this study hypothesizes the following:
H6: The relationship between health consciousness and consumers’ attitude towards the advertisement is moderated by construal level. Specifically,
towards the advertisement when the consumer is exposed to a present-focused message.
H6b: Consumers that are high in health consciousness have more positive attitudes towards the advertisement when the consumer is exposed to a future-focused message. 2.3.2.2. The moderating effect on attitude towards organic products
As mentioned, health-conscious consumers generally have more positive attitudes towards organic products. However, the study of Martin et al. (2009) implies that whether the advertisement focuses on the present or the future consequences can influence consumers' attitude towards organic products. Health-conscious consumers generally take into account the future consequences of their behavior. Therefore it is expected that they have more positive attitudes towards organic products when exposed to a future-focussed message. In addition, as the proximal consequences are more concrete, consumers who are little health consciousness may adapt their attitudes towards organic products, as they can be categorized as consumers with present orientations (Martin et al., 2009). To confirm this line of reasoning, this study hypothesizes:
H7: The relationship between health consciousness and consumers’ attitude towards organic products is moderated by construal level. Specifically,
H7a: Consumers that are low in health consciousness have more positive attitudes towards organic products when the consumer is exposed to a present-focused message.
H7b: Consumers that are high in health consciousness have more positive attitudes towards organic products when the consumer is exposed to a future-focused message.
2.4. Conceptual framework
derived from the literature review is shown in Appendix A.
3. Data and method
This section gives an overview of the sample, manipulation of the variables and the statistical procedure.
3.1. Sample
A cross-sectional experimental Internet-based survey was used to collect respondents. From the 258 respondents that started the survey, 25 careless responses are deleted based on an instructed response item (Meade & Craig, 2012). This leaves 233 cases. From these 233 cases, 168 cases are fully completed by the respondent (completion rate of 72%). The average age of the respondent is 26 years old (M=25.78, SD=.69, range: 16-78), with the largest group being 23 years old (25%). Of all respondents, 27% was male, 72% female and the remaining 1% preferred not to tell their gender. 24% follow or have completed their bachelor’s degree and 35% follow or have completed their master’s degree and the majority has a monthly net income of below 1.000 euros (52%). Approximately 53% answered the questionnaire in English and 47% in Dutch. The respondents were almost equally divided across the conditions: control group n=71, gain/future condition n=42, gain/present n=37,
3.2. Manipulation
Four versions of an ad for organic tea brand is created (see appendix B and C) and one group will not see an ad (control group). In the advertisements organic tea is presented. The tagline of all advertisements is ‘Think about the consequences of the choices you make’ to prime the importance of consumers’ decisions, which is adopted from a similar study from White et al. (2011).
In the four advertisements, message frame and construal level manipulations are included. The message frame is manipulated in the advertisement by varying the focus on the advantages of organic products (i.e. gain) or the disadvantages of non-organic products (i.e. loss). This research refers to low message framing when mentioning a 'loss-focused message' and high message framing when mentioning a 'gain-focused message'. This increases the relevancy of the advertisement as the advertisement explains how organic products lead to gains or why non-organic products lead to losses. This is in line with the research of Zanoli and Naspetti (2002), who found that the consumers primarily want to understand how organic products can be distinguished from non-organic products. This manipulation was checked by asking participants to what extend the advertisement focuses on advantages or disadvantages of the targeted product.
The message construal level was manipulated in the ad by varying the temporal focus (present- or future-frame) of the message. This research refers to low construal level when mentioning 'present focus' and high construal level when mentioning a 'future focus'. Again, this manipulation was checked by asking participants to what extend the advertisement focuses on the present or the future. In line with Chang et al. (2015), the specific consequences were emphasized. For example, in the proximal (vs. distal) temporal focus condition, participants read ‘Think about the consequences of the choices you make today
(vs. in the future)’ and ‘the benefits of organic tea (vs. the disadvantages of non-organic tea). Furthermore, an unknown, new brand name (Yermana) is used to avoid any potential confounding effect due to prior brand attitudes.
3.2.1. Pre-test
A pretest with another 72 respondents helped check whether the ads were successfully manipulated in terms of the gain-loss framing and temporal distance. To check the construal (present/future) manipulations, participants were asked the following one question on a 7 point bipolar scale (1 being ‘present’ and 7 ‘future’): To what extent does the ad focus on the present/future? The results of the independent samples t-test indicate that the participants in the future condition reported that the message focused on more on the future than the present (n=24, M=5.83, SD=1.167, SE=.238). Those exposed to the present condition reported that the message focused more on the present than the future (n=36, M=3.92, SD=1.713, SE=.286). The difference between the two conditions, 1.917 (95% CI [-2.718, -1.115]) was significant t(58)=-4.784, p<.001. Therefore, the null hypothesis stating that there is no difference in present/future framing between groups is rejected.
To check the message framing (gain/loss) manipulations, participants were asked the following one question on a 7 point bipolar scale (1 being ‘gain’ and 7 ‘loss’): To what extent does the ad focus on the gain/loss? The results of the independent samples t-test show that the participants in the loss condition reported that the message focused more on the disadvantages than the advantages (n=31, M=2.387, SD=1.667, SE=.299). Those exposed to the gain condition reported that the message focused more on the advantages than the disadvantages (n=29, M=5.757, SD=1.527, SE=.284). The difference between the two conditions, 3.372 (95% CI [-4.188, -2.544]) was significant t(58)=-8.152, p<.001. Therefore, the null hypothesis stating that there is no difference in gain/loss framing between groups is
rejected. The result of the pretest provide sufficient evidence that both manipulations worked effectively, therefore, they are adopted in the main study.
3.3. Measurement of the variables
This section states the measurements of the variables and their reliability. For the survey questions, see appendix D and E.
3.3.1. Translation, back-translation procedure
All items used are originally tested in English. Since the native language of the main population in The Netherlands is Dutch, the survey had two versions: English and Dutch. The original items have been translated into Dutch and have been translated back to English to check discrepancies. This has been done according to the translation-back-translation procedure (Brislin, 1980). For all survey questions, please refer to appendix D and E.
3.3.2. Attitude toward the ad
Attitude toward the advertisement (Aad) comprises three items and is assessed using Mackenzie and Lutz (1989) 7-point Likert scale items where lower values indicate positive attitudes, i.e. ‘good/bad, pleasant/unpleasant, favourable/unfavourable’ (α=.866). The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). None of the items would substantially affect reliability if they were deleted (all Cronbach’s α if deleted Δ<.10), so no variables were deleted.
3.3.3. Consumers’ attitude toward organic products
Consumers’ attitude toward organic products (Aorg) consists of eight items and is assessed using Gil, Gracia & Sanchez’s (2000) 7-point Likert scale items of attitude towards organic products where higher values indicate positive attitudes, i.e. ‘Organic food products are healthier’ (α=.65). The corrected item-total correlations indicate that all the items have a good
correlation with the total score of the scale (all above .30), except for question eight (‘Organic products are in fashion’). However, none of the items would substantially affect reliability if they were deleted (all Cronbach’s α if deleted Δ<.10), so no variables were deleted.
3.3.4. Purchase intention of organic products
Purchase intention of organic products consists of three items and is assessed using Michaelidou and Hassan (2008)’s 7-point semantic differential scale items where higher values indicate stronger intention, i.e. ‘I intend to purchase organic food produce within the next fortnight’ (α=.93). The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). None of the items would substantially affect reliability if they were deleted (all Cronbach’s α if deleted Δ<.10).
3.3.5. Health consciousness
Health consciousness is measured by using the 11 item Likert Health Consciousness Scale adopted from Oude Ophuis (1989) where higher values indicate higher health consciousness, i.e. ‘I have the impression that I sacrifice a lot for my health’ (α=.78). The corrected item-total correlations indicate that all the items have a good correlation with the item-total score of the scale (all above .30), except for the first question (‘I have the impression that I sacrifice a lot for my health’). However, none of the items would substantially affect reliability if they were deleted (all Cronbach’s α if deleted Δ<.10).
3.3.6. Control variables
As recommended by Briz and Ward (2009), age, gender, education, and income are used as control variables to identify characteristics of the meaningful groups of respondents. Briz and Ward (2009) argue that people aged 25-44, people with higher income, and people with higher education tend to be more aware of organic products. Also, it is expected that people
aged 25-44 are more able to adapt to changes (Briz & Ward, 2009. Moreover, consumers with higher incomes have greater purchasing power and experience fewer barriers to overcome the challenge that the organic sector faces regarding accessible pricing of organic products (Briz & Ward, 2009. In addition, people with higher education could be more concerned about their health and what they are eating.
3.4. Statistical procedure
Participants responded to items using self-reported measures in an online survey that is administered in Dutch and English between the period of 21 November 2016 and 11 December 2016. Statistical analyses are performed using the Statistical software Package for Social Sciences (SPSS). During the analyses, several abbreviations are used to describe the variables, see table 1. First participants read a cover story describing the survey as a university project and to gain insights for the start-up Yermana. After that, participants answered a series of questions about their level of health consciousness. Then the participants were exposed randomly to one of the four different advertisements. The participants in the treatment group (exposed to the advertisement) answered questions regarding the message framing, construal level, and attitude towards the advertisement. These questions are skipped for the control group, as they did not see an advertisement on which they can base their answers. All participants answered a series of questions regarding their attitude towards organic products and purchase intention of organic products. The participants also answered questions about basic demographics namely age, gender, education, and income. No formal definition of organic products was provided to the respondents, given the broad scope and emerging nature of organic products.
in data entry. As mentioned, an instructed response item is adopted in the survey from Meade and Craig (2012). By including this item, 25 careless responses are revealed and left out of the analysis to ensure data quality (i.e. ‘to monitor quality, please respond with ‘disagree’ for this item’). This was dealt with the missing data by excluding cases on a pairwise basis. Data is excluded only for analyses for which they have missing data so that the data is used to measure relevant variables and not the manipulated variables (Field, 2013). Some variables are re-coded as they had counter-indicative items. (rAad_1, rAad_2, rAad_3, rMF_Mani_Check, rHC_7, rHC_8, rHC_9, rHC_10, rAorg_3, rAorg_5).
A normality check is done by checking skewness, kurtosis, histograms, Q-Q plots and Kolmogorov-Smirnov (See appendix F). All variables indicate normality. No more than .5% of outliers were found for all scales, which is expected in a normal distribution.
The first analysis checks whether the manipulations in the advertisement are successfully manipulated. This is done in the same way as done for the pre-test, namely by comparing the means of all conditions per variable in an independent-samples t-test. The second analysis tests the multiple mediations of both consumers’ attitude towards organic products and attitude towards the advertisement via the PROCESS extension model 6 developed by Hayes (2012). The other six analyses test the moderation effect of message framing and construal level separately via the PROCESS extension model 1 developed by Hayes (2012).
Table 1
Abbreviations of the variables
Variable name Abbreviation
Attitude towards the advertisement Aad Attitude towards organic products Aorg Purchase intention of organic products PIorg
Health consciousness HC
4. Results
This section discusses the results of this research. First, it discusses the manipulation, then a correlation analysis of the data and after that, the mediation and moderation analyses. In line with Hayes (2012) all samples are resampled based on bootstrap of 5000 times. Confidence intervals are set on a 95% interval. Health consciousness is mean centered prior to the analysis. Message framing is automatically coded into dummy variables by PROCESS. Also, each analysis controls for age, gender, education, and income.
4.1. Manipulation checks
To check the construal (present/future) manipulations, participants were asked the following one question on a 7 point bipolar scale (1 being ‘present’ and 7 ‘future’): To what extent does the ad focus on the present/future? The results of the independent samples t-test indicate that the participants in the present condition reported that the message focused more on the present than the future (n=74, M=4.30, SD=1.87, SE=.22). Those exposed to the future condition reported that the message focused more on the future than the present (n=54,
M=5.02, SD=1.536, SE=.21). This difference, .721 (95% CI [-1.337, -.106]) was significant t(126)=-2.319, p=.022. Therefore, the null hypothesis stating that there is no difference in present/future framing between groups is rejected.
To check the message framing (loss/gain) manipulations, participants were asked the following one question on a 7 point bipolar scale (1 being ‘loss’ and 7 ‘gain’): To what extent does the ad focus on the advantages/disadvantages? The results of the independent samples t-test indicate that the participants in the loss condition reported that the message focused more on the disadvantages than the advantages (n=68, M=2.81, SD=1.77, SE=.215. Those exposed to the gain condition, who reported that the message focused more on the advantages than the disadvantages (n=58, M=5.84, SD=1.36, SE=.179). The difference between the two conditions, 3.04 (95% CI [-3.601, -2.471]) was significant t(124)=-10.64, p<.001. Therefore, the null hypothesis stating that there is no difference in gain/loss framing between groups is rejected.
4.2. Correlation analysis
Before testing the hypotheses, scale means and correlations have been computed for all variables (See table 2). The moderately significant correlation between age and purchase intention could be explained due to the highly significant correlation between income and age. This makes sense in the way that younger people earn less and more experienced people (thus older) earn more and thus have more purchasing power. Also, a negative correlation is found between income and attitude towards the advertisement. Another moderately significant correlation is found between attitude towards organic products and attitude towards the advertisement. This implies that consumers who have a positive attitude towards the advertisement also have a positive attitude towards organic products and vice versa. The
attitude towards organic products is moderately correlated with health consciousness and even highly correlated with purchase intention. Also, purchase intention is highly correlated with health consciousness. Subsequent analyses are done while controlling for the control variables age, gender, education and income.
Table 2
Means, Standard Deviations, Correlations and Reliabilities
Variables Items Mean SD 1. 2. 3. 4. 5. 6. 7. 8.
1. Age 1 25.79 8.95 - 2. Gender 1 1.73 .46 -.03 - 3. Education 1 6.24 1.43 .06 -.015 - 4. Income 1 2.31 1.95 .55** -.14 .03 - 5. Aad 3 4.49 1.35 -.13 -.05 -.02 -.19* (.87) 6. Aorg 8 4.90 .72 -.02 .15 -.03 -.06 .20* (.65) 7. PIOrg 3 4.13 1.74 .18* .11 -.13 .08 .04 .52** (.93) 8. HC 11 4.40 .80 -.08 -.003 -.10 -.04 .14 .19* .25** (.78) Note: N=175. Reliabilities are reported along the diagonal.
**Correlation is significant at the 0.01 level (2-tailed) *Correlation is significant at the 0.05 level (2-tailed)
4.3. Serial mediation
To test the first hypothesis, namely the serial mediation of Aad and Aorg on the relationship between HC and PIorg, PROCESS model 6 is used (n=157). For the statistics of the serial
mediation, see appendix G. For the statistical diagram with corresponding correlations see figure 2. The direct effect of health consciousness on purchase intentions is positive and significant, c’=.484, t(149)=3.371, p=.001. In other words, among two consumers with the same levels of Aad and Aorg, the one that has one unit higher in health consciousness is estimated to score .48 points higher in purchase intention of organic products. Additionally, three indirect effects further illuminate the underlying process.
The first indirect effect of HC on PIorg through Aad is found insignificant. This means that whether or not consumers have higher levels of HC do not necessarily have a more positive Aad (a1=.185, p=.146), which is also not associated with PIorg. (b1=-.100,
p=.091), independently of Aorg. This effect is not significant (indirect effect=-.019, SE=.026, CI: -.103 to .011).
The second indirect effect indicates the specific effect of HC on PIorg through Aorg. Consumers with higher levels of health consciousness have significantly more positive Aorg (a2=.193, p=.004), which also does lead to higher PIorg b2=1.052, p<.001), regardless of
consumers’ Aad. This effect is significant (indirect effect=.203, SE=.077, CI: .072 to .384). The third indirect effect is the effect of HC on PIorg through Aad and Aorg in serial. An examination of the coefficients revealed that, even though Aad is positively related to Aorg (a3=.107, p=.013), HC was not significantly related to Aad (a1=.185, p=.146), but only
to positively related to Aorg (a2=.193, p=.004). In turn, Aorg is positively related to PIorg
(b2=1.05, p<.001). This serial effect is found not significant and does therefore not support
the serial mediation hypothesis (indirect effect=.021, SE=.019, CI: -.005 to .073). To conclude, the direct relationship between HC and PIorg can not be explained through Aad and Aorg serial. Therefore, H1 is rejected as only Aorg mediated the relationship between HC and PIorg.
4.4. Moderation effects of message framing
In this section, three moderation effects of message framing are investigated. Specifically, the moderation effect on purchase intention, attitude towards the advertisement and attitude towards organic products is analyzed.
4.4.1. The moderating effect on purchase intention
To investigate the second hypothesis, namely the moderation effect of message framing on the relationship between health consciousness and purchase intention of organic products, PROCESS model 1 is used (n=168) (See output in appendix H). The interaction terms both show that moderation is not taking place. For the loss condition on HC, the regression coefficient is int_1=-.328 and is not statistically different from zero, t(158)=-1.024, p=.308. For the gain condition on HC, the regression coefficient is int_2=.01 and is not statistically different from zero, t(158)=-.241, p=.981. Thus, the effect of health consciousness on
purchase intention does not depend on whether the consumer is exposed to a message focussing on gains or losses. Therefore, the rest of the output is not analyzed any further. 4.4.2. The moderating effect on attitude towards advertisements
To investigate the third hypothesis, namely the moderation effect of message framing on the relationship between HC and Aad, PROCESS model 1 is used (n=157) (See output in appendix I). The interaction terms both show that moderation is not taking place. For the loss condition, the regression coefficient is int_1=.116 and is not statistically different from zero, t(147)=.403, p=.687. For the gain condition, the regression coefficient is int_2=-.074 and is not statistically different from zero, t(147)=-.256, p=.798. Thus, the effect of HC on Aad does not depend on whether the consumer is exposed to a message focussing on gains or losses. Therefore, the rest of the output is not analyzed any further.
4.4.3. The moderating effect on attitude towards organic products
To investigate the fourth hypothesis, namely the moderation effect of message framing on the relationship between HC and Aorg, PROCESS model 1 is used (n=168) (See output in appendix J). One of the interaction terms shows that moderation is marginally taking place. For the loss condition, the regression coefficient is int_1=.288 and is at the margin of statistical significance, t(158)=1.836, p=.068 (with a confidence of 90%). Thus, the effect of HC on Aorg seems to depend on message framing. Moreover, this model accounts for 10% of the variance in Aorg (Rsq=.102). For the gain condition, the regression coefficient is int_2=.037 and is not statistically different from zero, t(158)=.260, p=.795. Thus, the effect of health consciousness on purchase intention does not depend on whether the consumer is exposed to a message focussing on gains or losses.
A closer inspection of the conditional effects indicates a significant relationship between HC and Aorg only when the consumers are exposed to a message focusing on a
loss-framed message (effect=.334, SE=.122, p=.007, CI: .131 to .536) compared to no (effect=.0456, SE=.096, p=.635, CI:-.113 to .204) or gain-framed messages (effect=.082, SE=.104, p=.431, CI: -.090 to .255) (see appendix K for the output of the conditional effect). As can be seen from probing the interactions in figure 3, the slope linking HC and Aorg shows that when consumers are low in health consciousness, their Aorg is higher when exposed to a loss-framed message than to a gain-framed message. This interaction is found significant (effect=.334, SE=.122, p=.007, CI: .131 to .536). For consumers with high levels of health consciousness no significant interaction is found (effect=.082, SE=.104, p=.431, CI: -.090 to .255). Therefore, the slope indicating that consumers with high levels of health consciousness have a higher Aorg when exposed to a gain-framed message than to loss-framed message is not supported by the data in this study. Therefore, this study finds partial support for the hypothesis stating the moderation of message framing on the relationship between HC and Aorg only for loss-framed messages and not for gain-framed messages.
4.5. Moderation effects of construal level
In this section, three moderation effects of construal level are investigated. Specifically, the moderation effect on purchase intention, attitude towards the advertisement and attitude
towards organic products are analyzed. 4.5.1. The moderating effect on purchase intention
To investigate the fifth hypothesis, namely the moderation effect of construal level on the relationship between HC on PIorg, PROCESS model 1 is used (n=168) (See output in appendix L). The interaction terms both show that moderation is not taking place. For the present condition, the regression coefficient is int_1=-.401 and is not statistically different from zero, t(158)=-1.175, p=.242. For the future condition, the regression coefficient is int_2=.206 and is not statistically different from zero, t(158)=.586, p=.559. Thus, the effect of health consciousness on purchase intention does not depend on whether the consumer is exposed to a message focussing on the present or the future. Therefore, the rest of the output is not analyzed any further.
4.5.2. The moderating effect on attitude towards advertisements
To investigate the sixth hypothesis, namely the moderation effect of construal level on the relationship between HC on Aad, PROCESS model 1 is used (n=157) (See output in appendix M). The interaction terms both show that moderation is not taking place. For the present condition, the regression coefficient is int_1=-.087 and is not statistically different from zero, t(147)=-.305, p=.761. For the future condition, the regression coefficient is int_2=.208 and is not statistically different from zero, t(147)=.766, p=.445. Thus, the effect of HC on Aad does not depend on whether the consumer is exposed to a message focussing on the present or the future. Therefore, the rest of the output is not analyzed any further.
4.5.3. The moderating effect on attitude towards organic products
To investigate the seventh hypothesis, namely the moderation effect of construal level on the relationship between HC on Aorg, PROCESS model 1 is used (n=168) (See output in appendix N). The interaction terms both show that moderation is not taking place. For the present condition, the regression coefficient is int_1=.231 and is not statistically different from zero, t(158)=1.53, p=.128. For the future condition, the regression coefficient is int_2=.083 and is not statistically different from zero, t(158)=.083, p=.600. Thus, the effect of HC on Aad does not depend on whether the consumer is exposed to a message focussing on the present or the future. Therefore, the rest of the output is not analyzed any further.
5. Discussion
This study investigates how message framing and construal level influence the relationship between health consciousness and purchase intention of organic products. This study tests seven hypotheses: two are partially supported and five are rejected. Specifically, H1 is partially confirmed (only mediation for attitude towards organic products and not for attitude towards the advertisement), H4 is partially confirmed (only moderation for loss-framed messages), H3, H4, H5, H6, and H7 are rejected (see appendix O for an overview of the hypothesis outcomes). In the next section, each hypothesis is discussed. Specifically, it reviews the hypotheses stated in the introduction and discusses which hypotheses are supported by the data. The second section states the main theoretical and practical implications of the findings of this study by comparing it to previous studies which have been discussed in the introduction. The last section discusses the limitations and directions for future research.
5.1. Serial mediation
The investigation of several mechanisms by which health consciousness causally influences purchase intention of organic products is one key contribution to current literature. Specifically, this study investigates whether the effect of health consciousness on purchase intention can be explained by consumers' attitude towards the advertisement and attitude towards organic products in serial.
First, the findings confirm the direct effect of health consciousness on purchase intention of organic products. In other words, more health-conscious consumers have higher purchase intentions of organic products. This supports the belief that health consciousness increases purchase intention of organic products (Schifferstein & Oude Ophuis, 1998; Williamson, 2007; Vukasovič, 2016).
Second, this study does not find support for the serial mediation of consumers’ attitude towards the advertisement on the relationship between health consciousness and purchase intention of organic products. This means that consumers who have higher levels of health consciousness and higher levels of purchase intention are not influenced by their attitudes towards the advertisement. In other words, when consumers evaluate an advertisement, they do not adapt their attitudes towards organic products nor do they influence their purchase intentions in return. Moreover, there is also no relationship found between consumers’ attitude towards the advertisement and purchase intention. These findings are in contrast to the findings of previous studies who find that purchase intention is largely determined by consumers’ attitude towards the advertisement (Mackenzie & Lutz, 1989; Mitchell & Olson, 1981; Shimp, 1981). It might be that consumers' attitude towards the advertisement mediates another relationship that leads to purchase intentions, but that this relationship does not include health consciousness.
Third, the findings do confirm the indirect effect of health consciousness on purchase intention through consumers' attitude towards organic products. In line with Cheung et al. (2015), this study finds that health-conscious consumers have more positive attitudes towards organic products, which in turn leads to higher purchase intentions. This insight can be used by companies to encourage organic product consumption among health-conscious consumers. However, this effect does seem to depend on the organic product being advertised. For example, previous research such as Tarkiainen and Sundqvist (2005) did not show a mediation effect of attitude towards organic bread and flour, which indicates that an indirect effect is not always apparent for each organic product. In addition, consumers with positive attitudes towards organic products that indicate a high purchase intention, do not always actually purchase the product as indicated (Pearson et al., 2011). These limitations are further discussed in the section ‘Limitations and further research directions’.
5.2. The moderating role of message framing
To understand the conditions under which the effect of health consciousness on purchase intention of organic products operates, the moderation effect of message framing is investigated on several relationships. First, the moderation analysis of message framing on the relationship between health consciousness and purchase intention reveals that no moderation is taking place. Thus, whether consumers are exposed to loss- or gain-framed messages does not influence the relationship of health consciousness and purchase intention. This is in contrast to several studies who argue that gain- and loss-framed messages can increase the persuasiveness of the message (O’Keefe & Jensen, 2006 and Levin et al., 1998).
Second, the moderation analysis of message framing on the relationship between health consciousness and consumers' attitude towards the advertisement reveals that no
