The influence of convenience food, health motivations and quality perception on consumer food waste

(1)

Convenience in the kitchen.

Problem or solution for food

waste?

The influence of convenience food, health motivations

and quality perception on consumer food waste

(2)

1

Convenience in the kitchen. Problem or solution for food waste?

The influence of convenience food, health motivations and quality perception on consumer food waste

Completion date: 9th January, 2016

Master Thesis

Msc Marketing, specialization Marketing Management University of Groningen, Faculty of Economics and Business

By JUSTIN J. BAKKER Student number: 2022206 Herestraat 111a 9711LG Groningen Phone: +31 (0)641159059 E-mail: j.j.bakker.13@student.rug.nl

(3)

2

Abstract

By using an experimental field study, this research investigates the influence of convenient food products as well as the quality perception of these products on food waste generated by consumers in comparison to non-convenient food products. In addition the influence of different types of health motivation on the perceived quality of these two food types is examined. With a field experiment in which participants received free groceries, the amount of food waste for convenient and non-convenient food products was measured and analyzed. Results show that when people used convenient food products, they wasted less food than when they used their non-convenient counterparts. People with autonomous health motivation were found to perceive convenient food products as lower quality than non-convenient food products, while no difference in quality perception was found for people without this type of motivation. However, the perceived quality of the food products did not influence the amount of food waste generated by consumers.

Keywords: food waste, convenience food, quality perception, health, motivation,

(4)

3

Management summary

The purpose of this paper was to investigate factors that affect food waste behavior of consumers. Food waste is a very large problem on a global scale. Focusing on specific countries, the magnitude of the problem becomes clear. In the Netherlands, people annually waste approximately 105 kg of food per household which is around 50 kg per person. This amount is even higher in UK where households waste 160 kg of food each year (Quested, March, Stunnel & Parry, 2013). These quantities of food waste lead to negative economic, natural, social and environmental consequences. Despite all these negative consequences, food waste has not been investigated on a large scale by academic literature. This study tries to gain more insights into the causes of food waste by consumers and thereby contributing to the academic stream of food waste literature. But most importantly, the results of this study can be used to reduce the amount of food waste generated by consumers.

The focus of this study lies on the influence of convenience food on food waste compared to its non-convenience counterparts. Convenient food products are pre-cut and pre-sliced products that minimize preparation, cooking and cleaning up time and are therefore perceived and used differently by consumers than their non-convenient counterparts. This study also draws attention to the perceived quality of these two food types and if this quality perception affects food waste. Finally it is also investigated if different types of motivation to live healthy influence the quality perception of these food types. To determine if the expected relationships exist, a field experiment was conducted in which participants received a package of free groceries which was used to measure their food waste behavior.

The results of the study show that people waste less food when they use convenient food products compared to non-convenient food products. These food types are not perceived differently in terms of quality in general, but when people have autonomous motivation to live healthy they do perceive convenient food products to be of lower quality than their non-convenient counterparts. In turn, quality perception did not influence the food waste behavior of consumers.

(5)

4

(6)

5

1. Introduction

It is clear that food is of essential importance to humans and without it, live would not be possible. In his famous hierarchy of needs theory, Maslow already stated in 1943 that food is one of the most important requirements for human survival. Food has a very central role is the lives of humans. It not only provides fuel, nutrients and sustenance, it is also central to many social interactions and feeds into the sense of identity (Darnton et al., 2009). But when food is so important to us, why is it wasted on such a large scale?

It is estimated that on a global scale around 1,3 billion tons of food is lost or wasted each year in production, manufacturing and distribution, and in homes (Gustavsson, Cederberg, Sonesson, Otterdijk & Meybeck, 2011). This is approximately one third of all the edible parts of food which are appropriate for human consumption. A large part of this global food waste is generated on a consumer level. Households are according to research the single largest contributor to the total amount of wasted food (Griffen, Sobal & Lyson, 2009). In the Netherlands for instance a total of 800 million kg of food is wasted annually by households, which is around 105 kg per household and 50 kg per resident. In the UK this amount is even higher with 160 kg per household, generating a total of 4,4 million tons of food waste which is avoidable (Quested, March, Stunnel & Parry, 2013). This large amount of food waste has many different consequences. Firstly, food represents economic value and when this is wasted it leads to a monetary loss for households. Secondly, natural resources, which are used for the production of food, are wasted when food is unnecessarily disposed. Thirdly, it has a social impact since food waste contributes towards increasing global food prices and making food less accessible to developing countries. Lastly, food waste also has a large environmental impact since it contributes to the emission of greenhouse gases (Stuart, 2009).

(9)

8

(Doron, 2012a; Koivupuro et al., 2011) and how people feel about wasting food (Cox & Downing 2007; Doron, 2012b).

The reasons why people waste food have also been investigated albeit on a small scale. Several reasons for food waste have been found in the literature and it can be concluded that it is a rather complex phenomena. The waste or disposal of food is the final phase in the process of provisioning food (Munro, 1995). It entails multiple behaviors related to food, from purchasing to preparing and eating it (Jensen et al., 2012) and in this process people make multiple interrelated decisions influencing how much food is wasted (Stefan et al., 2012). Evans (2011) concludes that the reasons for these kinds of decisions are very much related to the dynamics of everyday life, like the household structure, participation in the work force and long working hours and social contacts. Because of these dynamics, some food that is provisioned will not be eaten within the timeframe in which it is appropriate to eat and therefore get wasted.

(10)

9

Does usage of convenience food have an effect on the amount of food that is wasted by consumers?

Convenience food products and their non-convenient counterparts are two different types of products and can be perceived differently by consumers, especially when it comes to quality. The quality perception of food products is a very important aspect of the products because it influences the perceived value of the products and thereby the purchase intention (Zeithaml, 1988). Superior quality is an important tool used by many companies to create superior customer value and a sustainable competitive advantage (Day, 1990). Quality can be seen as a multifaceted concept which is based on different dimensions which cannot be all evaluated by consumers (Oude Ophuis & Van Trijp, 1995). To make a judgment about the perceived quality of products, consumers use surrogate or indirect indicators of quality. The convenience of food can be one of those indicators. Although the presence of convenience food is growing in the market, there are some negative beliefs surrounding this type of food. Convenience food should be less healthy than non-convenience food and is not perceived to be a proper meal (Costa, Schoolmeester, Dekker & Jongen, 2007). These researchers also state that convenience food products are perceived to be of lower quality than their non-convenient twins. Although this attitude is changing, probably due to the increased presence of these products in the market, convenience food products are still perceived as inferior products in terms of quality. Another study shows that especially people who find healthy living important, have negative perceptions of convenience food (Buckley, Cowan, McCarthy & O’Sullivan, 2005). A reason for this could be that because of processing, some food types lose different vitamin content (Barret, Beaulieu & Schewfelt, 2010). To find out if convenience food is still perceived as an inferior product compared to non-convenience food, the following research question will be investigated:

Are convenience food products differently perceived in terms of quality compared to non-convenience food products?

(11)

10

Johnson, 2009). According to this result, if higher perceived quality products are wasted, consumers will perceive this as an even bigger waste of money or good food. Quality perception of food products can therefore influence food waste. This effect is to this date not investigated on a large scale in food waste literature. To investigate this effect, the following research question will be researched:

Does the quality perception of food products have an effect on the amount of food that is wasted by consumers?

Negative quality perceptions can influence if people waste food more easily or leave convenience food and choose non-convenience instead. Especially the health value of food can affect the types of food people choose. This is acknowledged by multiple outcomes of research showing that health is an important motivator of food choice (Contento, Michela, & Goldberg, 1988; Roininen, Lahteenmaki, & Tuorila, 1999; Roininen, Tuorila, Zandstra, de Graaf, Vehkalahti, Stubenitsky, & Mela, 2001; Schifferstein & Oude Ophuis, 1998). In a study by Rortveit and Olsen (2009) it was found that food choice is a trade-off between convenience aspects on one side and health related aspects on the other. Also knowledge about nutrition, of which healthiness is an aspect, is assumed to influence food attitudes and beliefs about quality. These in turn influence behavior, for example the types of food that people choose (Axelson & Brinberg, 1992). Brunner and colleagues (2010) confirmed that nutrition knowledge was a strong predictor of convenience food consumption, demonstrating that the more consolidated the knowledge of nutrition, the lower the consumption of convenience food.

(12)

11

Is the quality perception of convenience food products different for people who have different types of motivations to live healthy?

Based on these goals, this research extends current academic literature in consumer behavior examining more reasons why food waste occurs and exploring the unexplored relationship between convenience food usage and consumer food waste. This is relevant because with better insights, communications with consumers about wasting food and how to reduce it can be better targeted, more personal and more effective.

This paper will continue as follows: In the next chapter, the existing literature about the current research topic will be investigated. In the third chapter the conceptual model will be explained combined with the hypothesis development. After that, the methodology of the research is presented. In the fifth chapter the results will be discussed and the final chapter an overview of the results is presented, together with the discussion, implications of the results and the limitations of the study.

2. Literature review

2.1 Household food waste

(13)

12

possibly avoidable food waste which is food that is eaten is certain situations but not in others, like potato skins.

It can be concluded from literature that households are the most important source of food waste. Research by Griffin and colleagues indicated that consumers are the single biggest contributor to the total amount of generated food waste. The same result was found for developed countries where households were the major contributor to food waste (Parfitt, Barthel & MacNaughton, 2010).

(14)

13

Another important reason for food waste mentioned in literature is that food is not used in time and is passed its sell by, use by or best before date. (Cox & Downing, 2007; Exodus, 2007; Evans, 2011). Especially food like fruit, vegetables and meat which are susceptible for rapid decay are thrown away easily when not used in time. Households share a characteristic which applies to all types of food, which is the awareness that food harbors the potential to make people ill (Evans, 2011). This awareness accelerates the evaluations of food as past its best. Although many people report wasting food as “being wrong”, it is stated by Evans (2011) that food safety is more important to people and a well-founded justification than the feeling of guilt or anxiety for wasting food. In a behavioral study by Cox & Downing (2007) several other reasons for food waste like “do not want to eat leftovers” or “do not know what to do with them” were mentioned by participants. Although these factors can be controlled by people, many factors influencing food waste are outside people’s control for example work patterns, time constraints, family structure, household traditions around meals and social contacts (Evans, 2011, 2012; Quested et al., 2013). Regarding this issue, Evans (2011) describes that it is almost impossible to consider food waste as a consequence of individuals making negative decisions to engage in behaviors leading to food waste. His analysis suggests that food waste arises as a consequence of household negotiating the contingencies of everyday life.

An interesting insight into these factors was mentioned by Quested and Johnson (2009) who found that the amount of food wasted per capita decreased when the household size increased. People in four-person households wasted approximately half the amount of food than single-person households did. In another analysis on food waste behavior, Quested and colleagues also found that people over the age of 65 waste around 25% less food than do people who were younger than 65 (Quested et al., 2013).

(15)

14

intention to waste food, it is interesting to note that the intention not to waste food does not have a significant effect on reported food waste (Stefan et al., 2013).

When shopping at a supermarket, food products are available to consumers in many different forms. Consumers can buy for instance unprocessed vegetables which need to be washed, cut and cooked at home, but they can also select food which can be consumed directly or only needs minor heating. These totally different products can be divided into two categories: non-convenient and non-convenient food products.

Besides the different predictors and outcomes of food waste already found in literature, the link between convenience food usage and food waste is lacking in academic literature. In their research examining the drivers of convenience food consumption, Brunner and colleagues (2010) mentioned avoiding waste as one of the possible drivers for convenience food consumption which turned out to be a significant driver for the consumption of some types of convenience food. Still the effect of convenience food usage on food waste, compared to the non-convenience counterparts of these food products, is lacking insight. To generate more insight into this relationship, it is first useful to look at convenience food in more detail.

2.2. Convenience food

(16)

15

and colleagues (2004) there is still no clear definition of convenience foods. But Buckley and colleagues (2007) mention that convenience is defined in terms of time, physical energy and mental effort savings offered to the consumer in food related activities.

Most of the literature about convenience food investigated convenience foods in the most general form, ranging from complete ready meals which did not need any heating to minimally prepared food which required full cooking. A clear distinction between different types of convenience food is made in the article by Costa, Dekker, Beumer, Rombouts and Jongen (2001). They make the distinction between (1) Ready-to-eat, like take-away main courses and chilled sandwiches. (2) Ready to heat, like chilled and frozen pizzas and canned food which require less than 15 minutes of heating before consumption. (3) Ready to end-cook like chilled and frozen lasagna and dehydrated pasta which require more than 15 minutes of heating. And finally (4) Ready to cook which is food with minimal preparation and still needs full cooking for some or all of its components.

It is suggested by Gofton (1995) that in dual-income households, convenience food is often served when parents are too tired or have too little time to prepare a meal cooked at home. High stress levels along working women have also resulted in a higher degree of convenience food usage (McKenzie, 1986; Traill & Harmsen, 1997). Describing the more general working population, Candel (2001) identified that meal preparers who worked over 30 hours a week were more convenience oriented than people working less than 9 hours per week. This result was contradicted by Brunner and colleagues (2010) who found that working full time decreased people’s convenience food consumption.

The increasing amount of single-person households is another factor leading to an increased consumption of convenience foods. An important demand of people living alone is that there is minimal cleaning up time for foods (Kahn, 1976) which caused the increasing usage of foods requiring quick preparation and cleaning up times which symbolizes convenience food. Having children and household size were both found to be negatively related to convenience food usage meaning that the bigger the household, the less convenience food was used (Belizzi & Hite, 1986; Bonke, 1996). This result was confirmed by Brunner and colleagues (2010) for the usage of general convenience food products.

(17)

16

more recent outcomes (Brunner et al., 2010). Previous research relating to age has also shown that young people are heavy users of convenience food as stated in the article by Olsen, Menichelli, Sørheim and Naes (2011). Other literature describes convenience food users as less price sensitive (Swoboda & Morschett, 2001) and find that males consume more convenience food than women as do higher educated compared to lower educated people (Brunner et al., 2010).

For specific convenience categories, Brunner and colleagues (2010) generated some insights in the predictors for consumption. For a convenience category which they call single components (e.g. pre-cut, bagged, canned or frozen vegetables), age, nutrition knowledge and naturalness of food were all found to have a negative effect on the consumption. Whereas the decreased physical effort needed for meal preparation had a positive effect on the consumption of single components by consumers. Other insights show that the bigger the household the higher the consumption of single components. An interesting finding from this study, closely related to the current research, is that the more the participants wanted to avoid waste, the lower their consumption of single components and moderately processed food items was. This suggests a positive effect of single components and moderately processed food usage on food waste, where usage of single components and moderately processed foods increase food waste. This relationship will be elaborated in the hypothesis section.

Besides the predictors mentioned by Brunner and colleagues (2010), there are other factors which could affect the consumption of convenience food. One of those factors is how people view health and how motivated they are to pursue a healthy lifestyle.

2.3.1 Health motivation

(18)

17

exists between food consumption and health but the concept of healthy eating differs per person. Several studies tried to investigate these concepts of healthy eating. A healthy diet was many times defined by respondents mentioning a certain food or food group like fruits or vegetables (Keane & Willetts, 1996), or using other expressions like ‘balanced’ diet or eating ‘proper meals’ (Charles & Kerr, 1988). In a qualitative research, Winter Falk and colleagues (2001) identified seven issues which were defined as healthy eating in their sample. Healthy eating is: (1) eating low fat, (2) eating natural/unprocessed foods, (3) balanced eating, (4) eating to prevent disease, (5) maintaining nutrient balance, (6) eating to manage an existing disease and (7) eating to control weight.

The growing presence and importance of convenience food in the food market caused researchers to explore the link between health and convenience food. Several studies have found that consumers’ needs for convenience are related to food choice (Anderson & Shugan, 1991; Verbeke, 2001; Verlegh & Candel, 1999). Costa & colleagues (2007) state that ready meal consumption is a trade-off between, among other things, convenience, health and sensory appeal. Suggesting that the more convenient the meal, the less healthy. Examining the connection between health beliefs and convenience food, it was found that people’s beliefs about being physically healthy had a negative effect on the purchase of convenience food (De Boer et al., 2004), indicating that convenience food is perceived as being worse for your physical health than non-convenient food. Measuring the importance of product attributes of minimally processed vegetables and packaged fruits, Ragaert and colleagues (2004) found that health was not in the top 5 of most important attributes in the purchase and consumption stage. From their research it is also clear that for both processed fruits and vegetables health motivation was not an important motivation for buying it. Instead, motivations like: convenient, quick and delicious were given by participants as purchasing motivations. Even though people are aware of the food-health association, health is not a major motivator for purchasing when it comes to minimally processed vegetables and fruits (Ragaert, Verbeke, Devlieghere & Debevere, 2004).

(19)

18

2.3.2 Self-determination theory

Many theories of motivation, both historical and contemporary, have treated human motivation mainly as a unitary concept. These theories focus on the overall amount of motivation which people have for certain behaviors or activities. Self-determination theory (SDT) diverges from this view by making a distinction between different types of motivation. The idea behind the SDT is that the type or quality of someone’s motivation is more important than the total amount of motivation in predicting certain outcomes like, psychological health and well-being, effective performance and creative problem solving (Deci & Ryan, 2008a). According to the SDT, there are different types of motivation based on their level of self-determination (i.e. the extent to which a behavior is freely endorsed by people) which represents the quality of the motivation (Ratelle, Guay, Vallerand, Larose & Senécal, 2007). The SDT differentiates between three types of motivation: Autonomous motivation, controlled motivation and amotivation. Autonomous motivation consists of intrinsic motivation (performing a behavior for reasons connected to it, like pleasure) and types of extrinsic motivation (for reasons external to the behavior itself). These extrinsic types of motivation, often called identified regulation, are motivations in which people have identified with an activity’s value and find it personally important. Controlled motivation comprises two forms of extrinsic motivation: external regulation and introjected regulation. External regulations are motivations to behave in a certain manner based on external contingencies like obtaining a reward or avoiding punishment. Introjected regulation refers to behavior that is performed on account of internal pressures and is energized by issues like an approval motive, shame avoidance, contingent self-esteem and ego-involvements. People with a controlled motivation experience pressure, often from external sources, to think, feel or behave in specific ways. This stands in contrast with amotivation, which refers to the lack of intention and motivation, which is observed when people do not perceive the contingencies between their actions and consequences (Deci & Ryan 2008a; Ratelle et al, 2007).

(20)

19

and controlled motivation are not opposite dimensions but are rather related to each other (Ryan & Connel, 1989). So in SDT, individuals can have both autonomous and controlled motivations in a specific domain (Ryan, Plant & O’Malley, 1995). But a higher degree of autonomous motivation compared to controlled leads to more positive outcomes. Research has also gained some insights in the domain of health and the self-determination theory. People with a higher autonomous motivation were found to be more persistent in maintaining change towards healthier behavior (Ryan, Patrick, Deci & Williams, 2008). Where most theories state that self-determination depletes the energy available to the self, Ryan and Deci (2008b) found that autonomous regulation can be vitalizing resulting in an increase in energy. Discussing physical health, Miquelon and Vallerand (2008) identified that autonomous motivations indirectly caused less physical symptoms via self-realization. Controlled motivation increased the physical symptoms via self-realization. This shows that people with high autonomous health motivations actually live and are more inclined to live healthier.

It was already mentioned that convenience food is perceived as being worse for your physical health than non-convenience food. Since different forms of health motivations lead to different behavioral outcomes, it is likely that convenience food is perceived differently by people with autonomous or controlled health motivation. One of product attributes which these consumers can perceive differently is the product quality.

2.4 Quality perception

Consumers say that convenience food is generally of low quality, even though they perceive convenience as a desirable property of food products (Zeithaml, 1988; Olsen, 2002). This already shows that consumers can perceive product quality in different ways. A definition of perceived quality was given by Aaker (1991) who stated that perceived quality was: ‘the customer’s perception of the overall quality or superiority of a product or service with respect to its intended purpose, relative to alternatives.’ In this research the quality perception of products will be discussed, especially that of food products.

(21)

20

There are many quality attributes on which consumers can base their quality judgment. Steenkamp (1990) distinguishes two groups: quality cues and quality attributes. Quality cues can be determined by the senses prior to consumption. They are categorized as either intrinsic or extrinsic. Intrinsic cues are part of the physical product. Examples for food products are appearance, color and shape. Extrinsic cues are connected to the product but are not physically part of it. Examples are price, brand name or country of origin. Quality attributes are benefit-generating product attributes and cannot be observed preceding consumption. They can be distinguished in experience quality attributes and credence quality attributes. Experience attributes can be ascertained based on actual experience of the product (taste, freshness, convenience) whereas credence attributes cannot be determined, even after normal use for a long time (healthfulness, naturalness)

Some literature describes the quality perception of convenience food by consumers. Gofton and Marshall (1995), state that convenience food has been associated with inferior products, lowering the quality perception of these products. According to other research, convenience food is still perceived as lower quality than non-convenience food is not perceived to be a proper meal (Costa et al., 2007; De Boer et al., 2004; Mäkelä, 2000). This shows that although convenience is an important experience quality attribute of food products, other cues and attributes are perceived as lower quality by consumers than non-convenience food, decreasing the general quality perception of these products.

The effect of quality perception on food waste has to this date not been researched on a large scale. Hayes and Kendrick (1995) have tried to identify if quality perception affected the amount of plate waste in a nutrition program for the elderly in the U.S. They found differences in quality perception of food, but this did not influence the amount of food wasted.

3. Conceptual model & hypothesis

3.1 Conceptual model

When constructing the conceptual model, it is important to mention some perspectives in this research. First, the convenience food mentioned in the model are not all different types of convenience food like the ones mentioned by Costa and colleagues (2001) or Brunner and colleagues (2010). The convenience foods selected in this research setup will be convenience vegetables and risotto rice with a short preparation time.

(22)

non-21

convenience food (Costa et al., 2007; De Boer et al., 2004; Mäkelä, 2000). It is therefore expected that the type of food influences the quality perception of consumers. Autonomous motivation leads to different behavioral outcomes than controlled and amotivation, also in the health domain (Deci & Ryan, 2008a; Ryan, Patrick, Deci & Williams, 2008). Also a relationship between health and food consumption can be found. Since these connections exist, it is expected that different types of health motivation lead to a different perceived quality of convenience food products compared to non-convenience food products. Academic literature is lacking around this relationship, therefore it is interesting to explore how types of health motivation influence the effect of convenience food usage on quality perception. By means of this, new insights will be generated on food waste behavior. These different quality perceptions can also lead to a change in food waste behavior. Since people see food waste as a waste of money and a waste of good food (Quested & Johnson, 2009), it is likely that a higher perceived quality of food products would influence their food waste behavior.

Time constraints are one important reason for food waste. Since convenience food is a solution to overcome these time constraints (De Boer et al., 2004) it is also expected that the usage of convenience food affects food waste behavior differently than non-convenience food. Therefore the conceptual model as shown in figure 1 was selected.

A detailed explanation why each effect is expected and the formulation of the hypotheses will be presented in the next section.

(23)

22

3.2 Hypotheses

3.2.1 The quality perception of convenience food

As can be seen in figure 1, it is expected that the effect of convenience food on food waste is partially explained by quality perception. This means that consumers waste more or less convenience food because they perceive these products differently in terms of quality compared to non-convenience food.

As described by Steenkamp (1990) and Oude Ophuis and Van Trijp (1995), there are four main elements which influence the quality perception of food products. These are: intrinsic quality cues, extrinsic quality cues, experience quality attributes and credence quality attributes. The perception of these four elements of a product, determines how good or how bad a product is perceived.

The connection between convenience food and quality perception has been investigated by many researchers and almost every time with the same results. Gofton and Marshall (1995) show that convenience food is associated with inferior ingredients and other researchers demonstrate that convenience food is perceived as lower quality than non-convenience food and that convenience food is not seen as a proper meal (Costa et al., 2007; De Boer et al., 2004; Mäkelä, 2000). Only Wansink and colleagues (2013) found that fruit was consumed more when it was offered as a convenience product, suggesting a higher quality perception of the convenience alternative of fruit. Since this was only found for one product category and the lower quality perception was found multiple times in other literature, this is also expected in the current study. This leads to the formulation of hypothesis 1:

H1: Convenient food products are perceived to be of lower quality than their non-convenient counterparts

3.2.2 The role of health motivation

(24)

23

vegetables that have been trimmed and/or peeled and/or cut into 100% usable product that is bagged or pre-packaged to offer consumers high nutrition, convenience, and flavor while still maintaining its freshness (Lamikanra, 2002). According to this definition, these products are very nutrient and are not less healthy or of lower quality than their non-convenient twins. However, according to Rico and colleagues (2007), it is well-known that processing of vegetables leads to a faster physiological deterioration. Even when only slight processing procedures are used, these lead to biochemical changes and microbial degradation of the product (O’Beirne & Francis, 2003). This may result in the degradation of the color, texture and flavor (Kabir, 1994; Varoquaux & Wiley, 1994). Antioxidant status, an element which make fruits and vegetables healthy, is also affected by processing (Lindley, 1998). Decreases in antioxidant levels were reported after the processing of spinach (Gil, Ferreres & Tomas Barberan, 1999) and mandarin (Piga, Agabbio, Gambella & Nicoli, 2002). According to Barrett and colleagues (2010) cutting can also lead to the loss of different vitamins in fruits and vegetables. It is clear that processing food can have negative influence on the healthiness of these types of food. These negative influences are very likely to have a negative impact on the quality perception of convenience food products and especially vegetables. This negative impact would be even stronger for people who know that processing has these negative effects on food.

(25)

24

People with an autonomous health motivation perform healthy behavior because they really want it themselves, for reasons connected to healthy behavior like feeling well or energized and because the behavior is personally important to the individual. High levels of autonomous motivation were found to lead to more positive outcomes for the individual like achievement (Guay & Vallerand, 1997) and long term persistence like maintained change towards healthier behavior (Deci & Ryan, 2008). It can be concluded that people with high levels of autonomous health motivation are more likely to reach their health related goals and be more persistent performing healthy behavior, implying that people who are more autonomously motivated to eat healthy will perform more healthy eating behaviors over time. Combining this with the fact that convenience food is perceived as unhealthy, especially by people with high health motivations, it is likely that people with high levels of autonomous health motivation are perceiving convenience food products as lower quality compared to non-convenience food products. This leads to the following hypothesis:

H2a: The higher the autonomous health motivation of people, the lower their quality perceptions of convenient food products.

Controlled motivated people perform a behavior to gain reward or avoid punishment, or for ego-involvements, avoidance of shame or to be approved by others (Deci & Ryan, 2008). It is clear that people with this type of motivation perform a behavior because of how they are viewed by others. Research shows that controlled motivation leads to totally different, more negative outcomes than autonomous motivation (Ratelle et al., 2007). In the case of health, this implies that controlled motivated people are less persistent and find it much more difficult to achieve their health related goals. It is very well possible that people with controlled health motivation use healthy products because they want to be seen as healthy by others. Because of less persistence they are more likely to yield to unhealthy food products. For people with controlled health motivation the advantages of convenience food like time savings and physical and mental effort savings could be more important than the health characteristics since they find it more difficult to achieve their health goals. Therefore controlled motivated people are more likely to perceive convenience food products to be of higher quality compared to non-convenience food products. This leads to the formulation of hypothesis 2b

(26)

25

Amotivated people have no intention or motivation to perform a certain behavior (Deci & Ryan, 2008). Because these people do not really care about performing healthy behavior, it is very likely that they do not distinguish between convenience and non-convenience food concerning healthiness. Since convenience food saves them time and effort in food related activities, they will perceive convenience food products to be of higher quality compared to non-convenience food products. For amotivated people, the following hypothesis is formulated.

H2c: The higher the amotivation of people concerning health, the higher their quality perceptions of convenient food products.

The effect of quality perception of food was is not a well-researched subject in the food waste literature. But food waste is described by consumers as a waste of money and a waste of good food (Quested & Johnson, 2009). Therefore it might be concluded that products that are perceived as higher quality products will be wasted less. When the perceived quality is high, consumers may be motivated to consume the food in due time or are not willing to act careless with the benefits of these products.

Tversky and Kahneman (1991) state that losses will be weighed more heavily than the

equivalent gains, which is known as loss aversion. When people waste food that they perceive to be of high quality, they will perceive this as a bigger loss than when the perceived quality is low. Combining the loss aversion theory with the findings of Quested and Johnson (2009), one may conclude that wasting food of higher quality will be perceived as a bigger loss which consumers want to avoid. When the product quality is low, consumers may act more careless with the products and perceive wastage of these products as a smaller loss which lead to less aversion of this waste behavior. Therefore the following hypothesis is formulated:

H3: The higher the perceived quality of food products, the lower the amount of food waste generated by consumers.

3.2.3 The influence of convenience food usage on food waste

(27)

26

everyday life (Evans, 2011, 2012). One of those dynamics perceived by people is increasing time constraints. Considering fruits and vegetables, Ragaert and colleagues (2004) mention that even among consumers who are highly motivated to eat these food types, barriers exist preventing them from consuming it. These barriers can be economic, but within higher income groups these barriers relate to for instance, more women working outside the home, less time for cooking and leisure instead of cooking (Marshall, Duxbury & Heslop, 1995; Ragaert et al., 2004). Vegetables and fruits are food types which decay rapidly, and since people know very well that food can make people ill (Evans, 2011) these foods will surely not be consumed when not used in time and will therefore get wasted. Combining this with the finding that less time invested in preparation of food created more food waste (Buchli & Lucas, 2001), it is clear that time constraints can increase food waste. One of the most important aspects of convenience food are time savings, mostly in stages like: preparation, cooking and cleaning up (De Boer et al., 2004). That convenience is an outcome for people with little time, is for instance shown by the fact that people who are responsible for meal preparation who had a paid job used more convenience food than people without a job (Verlegh & Candel, 1999). Convenience consumption was also positively correlated to the number of working hours. These results show that time constraints can be a driver of food waste. Since convenience food can help to overcome these time constraints it is likely that convenience food usage leads to less food waste.

(28)

27

H4: Usage of convenient food products, compared to its non-convenient counterparts, decreases food waste.

4. Methodology

4.1 Research design

To test the hypotheses stated in the previous section, an experimental field research was conducted. One major advantage of a field experiment is that the causal influence of one variable on another can be determined with more certainty than in a laboratory setting. (Aronson, Wilson & Brewer, 1998). It also leads to high external validity and since people’s food waste behaviors occur in their natural environment like their homes, a field experiment taking place inside the homes op participants is well suited in this research for gaining insights into food waste. There were only two different conditions in this experiment. Participants were randomly assigned to the convenience condition in which convenience food products were used, or to the non-convenience condition in which non-convenience food products were used.

4.2 Procedure

The experiment was conducted in the town of Eelde-Paterswolde in the north of The Netherlands and had a duration of ten consecutive days. Before the experiment, participants were recruited on a consecutive Thursday, Friday and Saturday at a large supermarket in Eelde-Paterswolde. These three days were selected because these were most of the time the busiest days with customers and therefore participants could be more easily recruited. The Saturday was also selected since on this day most dual income households do their grocery shopping and this group needed to be present in the sample.

(29)

28

story told to the participants described that the purpose of the research was to investigate how consumers perceive different products available in grocery packages. After this they were told what participating would mean for them, as described later on, and finally they were asked if they wanted to participate in the research. When the approached consumers agreed to participate in the research, they were told that they needed to read and sign the informed consent of the research. By signing the informed consent, participants agreed that they were properly informed about the purpose of the research and that their participation was fully voluntary. This document also stated that their data would remain anonymous and that participants could stop with participating at every moment without a reason. A copy of the informed consent document can be found in the appendix. After the participants signed the informed consent they were asked for an email address or phone number, in case of announcements about the research.

When consumers agreed to participate in the study, they were first asked how many people were currently living in their household. By means of this, the appropriate package size could be selected. The free grocery packages were distributed to the participants from shopping carts located next to the stand containing small batches of grocery bags. In total 200 packages were distributed. Because this amount was too large to place next to the stand, smaller batches of around eight packages for 1-2 person households and eight packages for households with more than two persons were used. Participants were randomly assigned to the convenience condition or the non-convenience condition. In demographic data about Eelde-Paterswolde (www.weetmeer.nl) it was found that around 60% of the inhabitants lived in two person households and approximately 40% lived in households with more than 2 persons. Because of this data, 60% of all the packages were suited for 1-2 person households, the remaining 40% were packages for more than two person households. A clear overview is given in table 1.

Convenience food packages Non-convenience food packages

Total 100 100

1-2 person households 60 60

2+ person households 40 40

Table 1: overview of packages per condition

(30)

29

distributed, 90 packages on Friday and 87 packages were distributed on Saturday.

In the experiment that was conducted participants were offered a package with free groceries which they could use at home. They could use these products for a period of ten consecutive days. Two different sizes of packages were available, one for 1-2 person households and one for households of more than two persons. It was important to assign the appropriate packages to the households since otherwise they would have too much food for the persons in a household which could be a major reason to waste more food which could lead to unreliable results. The value of the groceries in the package was approximately €10,- depending on the household size. One half of all the packages contained convenience products and the other half contained their non-convenient counterparts. The exact contents of the grocery packages per condition and per household size can be found in the appendix.

In the period of ten days participants needed to fill in a questionnaire which be explained in more detail later in this section. Participants could fill in this questionnaire on paper or online. Besides this, they needed fill in a food diary which will also be described in more detail later on. The questionnaire and the food diary were both available in the package together with the groceries. Finally participants needed to collect their waste of the provided products and return it after the ten day period. Special freezer bags were provided to the participants in which they could collect the waste. Participants were asked to collect all kinds of waste like peels, unused products and leftovers of complete meals. Participants were asked come to the same supermarket in Eelde-Paterswolde after the ten day period and hand in their filled in diaries, questionnaires and their bags with waste. The researchers were present at the supermarket to collect all the material which participants returned. To motivate people to bring everything back, a little present was awarded to participants if they would bring back the diary and the questionnaire. If people would also hand in their bags of waste, they would participate in a lottery in which five people could win a grocery coupon of 50 euros. When 75% of the participants would hand in their waste, the researchers would make a financial donation to local charity, in this case the food bank.

4.3 Measures

(31)

30

importance to this research, they will not be elaborated on. The complete questionnaire can be found in the appendix.

Quality perception

Because the amount of food wasted was measured by means of the food diary, the most important construct measured by the questionnaire was the quality perception of products in the grocery package. To measure how the participants perceived the products in terms of quality, participants needed to answer fourteen statements about quality attributes of the products (question 20 in the questionnaire). The statements could be answered on a 7-point Likert scale ranging from totally disagree to totally agree. Quality attributes that were considered were for instance taste, appearance and freshness of the products. This scale was based on Oude Ophuis and Van Trijp (1995)

Health motivation

Another important construct measured by the questionnaire is the health motivation of the participants, since it is expected that this influences the quality perception of convenience food. To measure the levels of health related autonomous, controlled and amotivation of the respondents, participants needed to answer fifteen statements about the reasons why they eat healthy or why not (question 25 in questionnaire). They could answer these statements on a 7-point Likert scale ranging from totally disagree to totally agree. These statements took into consideration if participants were motivated by themselves to eat healthy or if this motivation was due to how they were perceived by others. Some statements considered if there was any motivation to eat healthy at all. This scale is based on Ryan and Deci (2000) and Levesque and colleagues (2007).

Food waste

(32)

31

which of the ten days they did their grocery shopping. The complete food waste diary can be found in the appendix.

To check if the participants filled in the diary correctly and to investigate what people waste, the bags of waste which the participants brought back after the ten day period were all weighted. The weight of the waste per product was listed per participant. This weight data was used to create other waste variables, besides the waste that participants filled in in the diary, which were used in the analysis.

Control variables

In the questionnaire, multiple constructs were measured which are used as control variables in this study. Socio-demographic variables like age, gender, household size, education level and income of the participants are used as control variables. It is important to control for these variables in the analysis since these variables can affect the main effects in this study. In the experiment, half of the participants received three recipes together with the grocery products. The effect of recipes on food waste is not investigated in this research. But the usage of recipes can influence the results of this study, therefore if people received the recipes or not was selected as control variable. Two package sizes were used in the experiment (1-2 person households and more than 2-person households). To investigate if the results are the same for the participants using the 1-2 person package or the more than 2-person package, package size was selected as control variable.

5. Results

The result section is divided into different parts. First the sample will be described in general. Second, it will be explained how the different constructs are prepared for the analysis process, which steps were performed before the analysis and multiple pre-analysis tests will be described. Then the full model will be tested. First different parts of the complete interaction model will be tested and finally the complete model as can be seen in figure 1 will be investigated.

5.1 The sample

(33)

32

cases because both the diary and the questionnaire were needed for the analysis. When the incomplete cases were deleted, 150 cases remained which from which the data was suitable for analysis. For the variable age, the amount of missing values was above 10%. Therefore the missing values of the variable age were replaced by the mean. For all the other descriptive variables discussed here, the amount of missing values was under 10% and these missing values were therefore ignored.

Sample characteristics

An overview of the most important sample characteristics can be found in table 2. From the 150 participants, 128 (85,3%) were female and 22 (14,7%) were male. 45% of the participants lived in 2-person households, 41,6% lived in either 3 or 4-person households. This shows that the major part of the sample were 2-4 person households. 62% of the participants in the sample were highly educated, with 44% who had a HBO degree and 18% having a University degree. 18,7% of the participants only finished high school. 62,4% of the sample was employed and 6,7% were self-employed. This shows that 69,1% of the participants had a paid job, while 12,1% was unemployed and 15,4% was retired. The income level was measured on a household level and shows that 54,7% of the sample earned more than €3000 per month per household. The distribution of the different characteristics is quite similar over the two different conditions.

Table 2: Sample characteristics

5.2 Creation of variables

(34)

33

items. To create separate variables for the different types of health motivation, a factor analysis with Varimax rotation was performed on the 15 items. To investigate if factor analysis was appropriate, a KMO and Bartlett’s test was performed. The KMO value of .830 was above .50 and Bartlett’s test of sphericity was significant (.000) showing that factor analysis could be performed. The factor analysis showed three components with an eigenvalue above 1 and a total variance explained of 57,93%, therefore three factors were selected. The two items which should measure amotivation (item 5 and 15) according to literature, did not load high enough (>.50) on any of the three factors. Because a new variable for amotivation could not be created, a new factor analysis was performed in which a fixed number of 2 factors to extract was selected.

The KMO before this factor analysis was .830 and the Bartlett’s test of sphericity was significant (.000). Since the items which should measure amotivation (item 5 & 15) did not load high enough on any factor, a fixed number of 2 factors to extract was selected, because only autonomous and controlled should be measured. The items which should measure autonomous motivation according to literature, loaded all on one factor. This was the same for all but one item which should measure controlled motivation. Item 2 scored higher on the factor which resembles autonomous motivation instead of controlled motivation, and was therefore included in the new computed autonomous motivation variable. To check if the new variables “autonomous motivation” and “controlled motivation” were reliable to use in further analysis, a reliability analysis was performed. The analysis for “autonomous motivation” showed a Cronbach’s Alpha of .810 and for “controlled motivation” showed a Cronbach’s Alpha of .772. Since these values are above .60, these variables are reliable for further analysis and will therefore be used when testing the hypothesis. Because the items which should measure amotivation did not load high enough on any factor, no reliability analysis was performed with these items.

(35)

34 Item Wording: The reason why I eat healthy is

because… Factor 1 Factor 2 Alpha if item deleted Cronbach’s Alpha

1 I feel responsible for my own health .665 -.434 .789

.810

2 I would feel bad or guilty if I would not eat healthy

.657 .277 .790

3 I belief that healthy food is the best for my health

.737 -.241 .781

6 I think that a healthy diet is good and important in my life

.818 -.246 .772

7 it I would feel bad if I would eat unhealthy

.649 .070 .788

8 It is an important decision that I want to make

.620 .082 .793

11 It is a part of my life goals .572 .274 .816

13 I find it important to live as healthy as possible

.723 -.226 .783

4 People in my surroundings would be annoyed when I eat unhealthy

.204 .558 .801

.772

9 I feel pressurized by other people -.194 .815 .685

10 It is easier to do what other people want me to do than to think about it for myself

-.228 .737 .725

12 I want get appreciation from other people .002 .762 .703 14 I want to show others that I can do it .057 .692 .728

5 I do not really think about -.398 .357 Item was excluded from

further analysis

15 I do not really know why -.308 .269 Item was excluded from

further analysis Table 5: Rotated factor loadings and Cronbach’s Alpha health motivation

(36)

35

A variable for food waste was created before the analysis out of the weight of the waste that participants returned. Because non-convenience food products were different from convenience food products in terms of weight, some adaptions had to be made to make the amounts of waste comparable between the convenience and non-convenience condition. For the calculation of this variable, the amounts of waste that were returned by the participants and weighted by the researchers were used. Because of this, this data was more reliable than the amounts of waste mentioned by participants in the diary.

Because the non-convenient food products have peels which the convenience food products do not have, it is impossible to compare the waste of these food types. To make the waste comparable, the weight of the unavoidable waste like peels and broccoli stump needed to be calculated for all the participants in the non-convenience condition. Not every participant returned their waste and some participants delivered all the waste in one bag. This made it impossible to calculate the right amount of unavoidable waste per participant in the non-convenience condition. To overcome this problem an average amount of unavoidable waste was calculated. From all the participants who returned their waste and for whom a reliable amount of unavoidable waste per product was weighted, the average unavoidable waste per product was calculated. This average was in total 607 grams for the 1-2 person households and 927 grams for the more than 2-person households. These amounts of unavoidable waste were subtracted from the total weighted waste per participant in the non-convenience condition. Since convenience food products are processed and do not have peels, the unavoidable waste was not subtracted from the total weighted waste per participant in the convenience condition. By means of this the amounts of waste in the convenience condition and non-convenience condition are comparable. Participants in the non-convenience condition received 150 grams more risotto rice than in the convenience condition. The difference of 150 grams was subtracted from the total weighted amount of waste per participant in the non-convenience condition to make it fully comparable with the waste in the non-convenience condition.

(37)

36

autonomous health motivation is much higher than the average level of controlled health motivation. This means that overall the participants in this study were more autonomously motivated when it comes to health than controlled.

Table 6: Descriptive statistics of independent and dependent variables

As a robustness check to find out if the results found for the food waste variable mentioned above are also found for other food waste variables, a second food waste variable was included in the analysis. For this food waste variable the data that participants filled in in the food waste diary was used. In the diary, participants described how much of each product they consumed in the ten day period. This second food waste variable was calculated by subtracting the consumption from the total original weight of the grocery package. Since there is a difference between convenience and non-convenience food products, this variable was also corrected for the peels and risotto weight difference as described earlier.

5.3 Normality check

To check if the data of the different variables used in the analysis had a normal distribution, a Kolmogorov-Smirnov test and a Shapiro-Wilk test were performed. For all the variables except autonomous motivation, these tests generated significant results. This shows that every variable used in the analysis except autonomous motivation has a non-normal distribution.

5.4 Correlations

(38)

37

Table 7: Correlation matrix

5.5 Hypothesis testing

5.5.1 The effect of convenience and non-convenience food on quality perception

To test the part of the conceptual model in which the effect of convenience food on quality perception is investigated, a multiple regression was performed. In the analysis, age, gender, household size, education, income, recipe and package size were included as control variables. This was performed to determine the effect of quality perception on food waste while statistically controlling for these variables. The significance level in this regression was 5%. The model was significant (p-value = .000) showing that the results from this analysis are reliable.

(39)

38

5.5.2 The influence of health motivation

In the same analysis as described above, it was investigated if the type of health motivation of participants influenced their quality perception of the products in the grocery package. The results of the multiple regression are presented in table 8. The analysis yielded a significant positive effect (p-value = .000) of autonomous motivation on quality perception. Because of the positive Beta, the result shows that the higher the autonomous health motivation of people was, the higher their quality perception of the products. But his is not the case for the participants who received the grocery package with convenience food products. A significant negative interaction effect of convenience food and autonomous motivation on quality was found (p-value = .002). This means that the higher the level of autonomous health motivation of people is, the lower they perceive the quality of convenience food products. Since the usage of convenience food did not have an direct effect on the quality perception of these products as described above, this result means that only when people had autonomous health motivation it influenced their quality perception of the products in the grocery package. Because of this result H2a can be confirmed. A visualization of this effect is presented in figure 2. From this figure it is clear that people with high levels of autonomous health motivation perceive non-convenient food products as higher quality than people with low levels of autonomous health motivation. Because of the negative interaction effect, the quality perception of convenient food products decreases more for people with high levels of autonomous health motivation compared to people with low levels of this type of motivation. This leads to a slightly lower quality perception of convenient food products for people with high levels of autonomous health motivation compared to people with low levels of autonomous health motivation

In the analysis, both controlled motivation and the interaction of controlled motivation with convenience food did not deliver significant effects (p-value = .457 and .197). This means that the level of controlled health motivation does not influence the quality perception of products. Participants with controlled health motivation did not perceive the grocery package with convenience food differently in terms of quality than the grocery package with non-convenience food. Therefore H2b will be rejected.

(40)

39

Model 1

Variable _Beta _t _{p-value VIF}

Convenience food .230 -1.640 .103 1.112

Autonomous motivation .451 3.575 .000 2.690

Controlled motivation -.085 -.745 .457 2.842

Autonomous* convenience food -.461 -3.163 .002 2.840

Controlled* convenience food .184 1.297 .197 2.705

Age -.004 -.573 .568 1.409 Gender .112 .566 .572 1.109 Household size .122 1.407 .162 2.079 Education -.195 -3.993 .000 1.240 Income .031 .543 .588 1.178 Recipe -.105 -.720 .473 1.201 Package size -.095 -.523 .602 1.848 R2 .230

Dependent variable = quality N = 150

Table 8: Results regression

(41)

40

5.5.3 The effect of quality perception on food waste

To test the part of the conceptual model in which the effect of quality perception on food waste is examined, a multiple regression was performed. In the regression, age, gender, household size, education, income, recipe and package size were included as control variables. The significance level in this regression was 5% and the model was significant (p-value = .000)

The results show that quality perception has no significant effect on food waste (p-value = .158). This result shows that the quality perception of the products in the grocery package did not affect how much food people wasted. It is interesting to see that in the model where the data from the food waste diary was used as food waste variable, quality perception did have a significant effect on food waste (p-value = .030). This shows that for this food waste variable, the higher people perceived the quality of the products in the grocery package, the less food they wasted. It can be concluded that quality perception does influence the amount of food waste, but this depends on which waste variable is selected. Since the effect cannot be found for the main food waste variable in this study, H3 is rejected.

An overview of the results is presented in table 9. The R-squared shows that there is not a strong fit for these models. The VIF scores which are used to test for muliticollinearity, all show a value lower than 4. Therefore it can be concluded that the muliticollinearity is low in these analyses.

Model 2 Variable

DV = Food waste DV = Package weight - consumption

Beta t p-value VIF Beta t p-value VIF

Quality -92.835 -1.422 .158 1.162 -189.462 -2.187 .030 1.162 Convenience food -483.685 -4.350 .000 1.096 -63.684 .432 .667 1.096 Age -14.034 -2.679 .008 1.271 -14.026 -2.018 .046 1.271 Gender 78.490 .509 .611 1.058 114.356 .599 .577 1.058 Household size -95.447 -1.368 .174 2.098 -162.357 -1.754 .082 2.098 Education -9.555 -.236 .813 1.325 9.291 .173 .863 1.325 Income -22.197 -.499 .619 1.131 -46.226 -.783 .435 1.131 Recipe -157.882 -1,421 .158 1.096 -138.040 -.936 .351 1.096 Package size 206,792 1.434 .154 1.811 712.210 3.723 .000 1.811 R2 .235 .174

Note: N = 125 for DV = food waste

N = 150 for DV = package weight - consumption

The influence of convenience food, health motivations and quality perception on consumer food waste

Convenience in the kitchen.

Problem or solution for food

waste?

The influence of convenience food, health motivations

and quality perception on consumer food waste

Convenience in the kitchen. Problem or solution for food waste?

Abstract

Management summary

Table of contents

1. Introduction

2. Literature review

3. Conceptual model & hypothesis

4. Methodology

5. Results