Reducing food waste:
The role of consumer values on the effectiveness of a
food waste intervention in the preparation phase
Reducing food waste: The role of consumer values on the effectiveness of a
food waste intervention in the preparation phase
University of Groningen
Faculty of Economics and Business
MSc Marketing Management
First supervisor: dr. J. van Doorn
Second supervisor: dr Y. Joye
Merinde Elianne van de Ven
Food waste is becoming of increasing concern due to its economic, social and environmental impact. Regarding this food waste problem, consumers are found to be the largest contributor, accounting for 53% of the waste in Europe (Gustavsson et al. 2011). However, little is known about interventions aimed at reducing household food waste and factors influencing the use of these interventions. Therefore, this study investigates the effectiveness of an intervention in the preparation phase on food waste reduction by consumers. In addition, possible moderating effects of biospheric, altruistic, egoistic and hedonic values on the effectiveness of the intervention on food waste reduction, are included. The participants in this field research either received the intervention aimed at portion size control or no intervention, both in combination with a food waste diary and questionnaire. The results show that the intervention did not lead to a significant reduction of food waste. However, hedonic values were found to weaken the effect of the intervention on food waste reduction. In addition, biospheric, altruistic and hedonic values had a direct decreasing effect on the amount of food waste. This research tries to fill the gap in the existing literature and is first to study the effect of an intervention in the preparation phase on food waste reduction and the influence of consumers values. It shows the importance of taking consumer values into account when designing food waste interventions.
Keywords: household food waste, preparation phase, food waste interventions, consumer
According to the European Commission (2016) around 88 million tonnes of food are wasted in the EU every year, with the corresponding costs estimated at 143 billion euros. Besides the economic impact, food waste also has a direct social and environmental impact. The largest share of waste and derived from consumers, who are responsible for 53% of the overall food waste in Europe (Gustavsson et al. 2011). In The Netherlands this comes down to 32 kilos of household waste (The Netherlands Nutrition Centre, 2014). Moreover, food waste in the preparation phase is found to be the most significant contributor to household waste (Lyndhurst et al.,2007; Parfitt et al.,2010). Although food waste is recognized as an important problem and is related to consumer behaviour (Parfitt et al., 2010), in reality consumers still have difficulties with preventing food waste.
Within the field of consumer behaviour, household food waste has been a neglected topic (Porpino, 2016). The existing literature on food waste mainly focusses on indicating consumer food waste behaviour and the quantity of food that is lost (Griffin et al., 2009), instead of paying more attention to the factors that drive food losses. According to De Groot and Steg (2009), interventions can be useful to raise awareness and focus on the cost and benefits of pro-environmental behaviour, like food waste reduction. Regarding consumer behaviour, consumer values have been found to influence behavioural beliefs and have been widely accepted as important determinants of consumer behaviour (De Groot and Steg, 2009; Homer and Kahle, 1988; Schwartz, 1992). The following consumer values have been identified as important: biospheric values, altruistic values, egoistic values and hedonic values (De Groot and Steg, 2007;2009; Steg et al., 2014).
The participants in this research received either a portion size control intervention or nothing. With the use of a food waste diary and questionnaire the food waste of 40 participants was analysed. The findings suggest that in contrast to the theory, there was no significant difference between the amount of food waste of consumers who did and who did not use the portion size control intervention. Regarding the hypothesized moderators, only a significant moderating effect was found for hedonic values. In addition, a significant negative direct effect was found for biospheric, altruistic and hedonic values on food waste.
This research is my master thesis and was written to finalize my master Marketing Management at the faculty of economics and business at the Rijksuniversiteit Groningen. The decision to join the ‘food waste’ thesis group was not a difficult one, because the subject has caught my attention due to its relation with sustainability. Sustainability is already an important topic for businesses and consumers, and will become even more important in the future. However, in my daily life I see a lot of food thrown away by supermarkets, restaurants, people around me and even by myself. And the extent to which this happens became even more clear during my part-time job at a restaurant. Therefore, I wanted to know more about food waste and what causes this behaviour.
Table of contents
1.Introduction ... 8
2. Literature review ... 12
2.1 Food waste and consumer behaviour... 12
2.2 Drivers of food waste ... 13
2.3 Consumer values... 15
2.3.1 Definition consumer values ... 15
2.3.2 Impact of consumer values on pro-environmental behaviour ... 15
3.Hypotheses ... 17
3.1 Food waste intervention in preparation phase ... 17
3.2.1 Biospheric values ... 18
3.2.2 Altruistic values and food waste reduction... 20
3.2.3 Egoistic values and food waste reduction... 22
3.2.4 Hedonic values ... 23
3.3 Conceptual model ... 25
4. Methodology ... 26
4.1 Data collection ... 26
4.2 Food waste diary ... 27
4.3 The questionnaire ... 27
5. Descriptive statistics and reliability analysis ... 29
5.1 Descriptive statistics ... 29 5.2 Demographics ... 29 5.3 Food waste ... 30 5.4 Consumer values... 30 5.5 Control variables... 31 5.6 Correlation ... 32 6. Results ... 33
6.1 Hypothesis 1 – Use of the intervention ... 33
6.2 Hypothesis 2 – Biospheric values ... 34
6.3 Hypothesis 3 – Altruistic values ... 35
6.4 Hypothesis 4 – Egoistic values ... 36
6.5 Hypothesis 5 – Hedonic values ... 37
7. Discussion ... 40
7.2 The moderating effect of consumer values... 41
7.3 Control variables... 43
8. Conclusion ... 43
8.1 Academic contribution ... 43
8.2 Managerial contribution ... 44
8.3 Limitations and further research ... 45
References ... 48
Appendix ... 53
Appendix A: Food waste diary ... 53
Appendix B: Questionnaire ‘Eetmaatje’... 55
Worldwide around one-third of all the food is thrown away or lost in the supply chain. In the EU alone around 88 million tonnes of food are wasted every year with the corresponding costs estimated at 143 billion euros (European Commission, 2016). Food waste has a direct economic, social and environmental impact. It threatens the environment due to greenhouse gasses and wasted resources throughout the production, processing, marketing, transporting and refrigerating of food (Rosenzweig and Parry, 1994; Parry et al., 2004; Godfray et al., 2010; Parfitt et al., 2010). Food waste is also an important societal problem, because of the negative consequences for food security and consumer well-being (Van Doorn, 2016). According to the World Food Program (2016) there are almost 800 million people in the world who do not have enough food to pursue an healthy and active life, while on the other side a large quantity of food is wasted.
Food waste takes place in the entire food supply chain. In fact, there are five different stages where this appears: agricultural production, post-harvest handling and storage, processing, distribution and consumption (Gustavsson et al. 2011). The largest share of waste is derived from consumers. They are responsible for 53% of the overall food waste in Europe (Gustavsson et al. 2011). In order to reduce food waste in the consumption stage, previous research focused on food spoilage reduction by examining the behaviour of consumers and providing these insights to food policies and industry standards (Tsiros and Heilman, 2005). Stuart (2009) highlights the need to raise awareness among consumers focusing on non-financial costs of food waste, like environmental impacts and world hunger.
Within the field of consumer behaviour, household food waste has been a neglected topic (Porpino, 2016). Although some researches have focused on food waste and consumer behaviour, academic literature concerning this topic is still insufficient. The existing literature on food waste mainly focusses on indicating consumer food waste behaviour and the quantity of food that is lost (Griffin et al., 2009), instead of paying more attention to the factors that drive food losses. Consumer behaviours that have been identified as causes of food waste are: buying and/or cooking too much, not being able to shop according to the shopping list, not being able to store food properly, failing to plan meals in advance and cooking too much (Brook, 2007; Parfitt et al., 2010). Further, food waste can occur in one of the following subsequent stages of the domestic food cycle, which include planning, shopping, storage, preparation and consumption (Flower and Collett, 2014).
According to Lyndhurst et al. (2007), food waste in the preparation phase is the most significant contributor to household waste. This is also in line with research conducted in the UK by Parfitt et al. (2010), who indicate that the two main reasons for avoidable food waste are “food is not used in time” and “too much food is cooked, prepared or served”. In The Netherlands 32 kilos of food waste is defined as household waste, of which around 9 kilos of solid food has already been prepared (The Netherlands Nutrition Centre, 2014). In the end this will result in post-preparation waste, which can be defined as food that is left on the plate (Lyndhurst et al., 2007). Obviously, preparing the right amount of food will make a significant difference in reducing the amount of food waste in households. Preparing the right amount can be reached by using interventions aimed at better portion control (Parfitt et al., 2010; The Netherlands Nutrition Centre, 2014). Therefore, this paper will focus on the effect of an intervention in the preparation phase on reducing food waste.
Schwartz (1992) states that consumer values are ordered in a system of value priorities. When multiple values are activated and competing with each other, behaviour will be based on the values that are most important (De Groot and Steg, 2009). Different consumer values will therefore lead to different consumer behaviours. The consumer behaviour, which is influenced by consumer values, can either be beneficial for the environment (pro-environmental) or harmful for the environment. In addition, it is assumed that different consumer values have different effects on consumer behaviour, therefore resulting in different applications of the interventions aimed at food waste reduction.
Biospheric values, one of the consumer values identified by De Groot and Steg (2007), are related to ‘ecocentric’ behaviour. Consumers with strong biospheric values will therefore primarily act in a way which is beneficial for ecosystems and the biosphere. In other words, people with biospheric values will act more pro-environmentally. Concerning food waste interventions, De Groot and Steg (2009) found that interventions promoting pro-environmental behaviour are more effective in the long run when consumers hold strong biospheric values or biospheric values are made more salient.
Next to biospheric values, the literature also identifies altruistic values (De Groot and Steg, 2007). Likewise, consumers with altruistic values are more willing to act pro-environmentally and use pro-environmental interventions. However this is due to different reasons, since altruistic values are concerned with pro-social and self-transcendence (De Groot and Steg, 2009). According to Steg et al. (2014) consumers with altruistic values “consider this behaviour to be worth the effort when it benefits the welfare of others and the environment, rather than themselves exclusively.”
On the other hand, egoistic values are concerned with self-enhancement (De Groot and Steg, 2009). Acting on egoistic values will often not result in pro-environmental behaviour, because it is at the expense of personal benefits like pleasure and comfort and personal resources like money and power (Dogan et al. 2014). Still, there is a possibility that interventions aimed at food waste reduction can be used as a result of egoistic values if for example the increase in comfort or personal resources, like money, is sufficient enough.
Because of these reasons, the intervention aimed at food waste reduction in the preparation phase will be less effective for consumers with strong hedonic values or they will completely refrain from using it.
It has been clear that different consumer values can result in different consumer behaviour. The important consumer values that have been identified in the literature are biospheric, altruistic, egoistic and hedonic values. Despite the importance of consumer values, the effects of consumer values have been mainly studied in the purchase and shopping phase (Kim et al., 2012; Graham-Rowe et al. 2013). Most importantly, the effect of consumer values on the use of interventions aimed at reducing food waste in the preparation phase has not been investigated before. This study will include the previous stated consumer values, since they play an important role in determining consumer behaviour. Therefore it is expected that consumer values also influence the consumer behaviour concerning the effect of the intervention aimed at food waste reduction in the preparation phase.
The overall purpose of this paper is to find out how consumer values influence the effect of the intervention in the preparation phase on food waste reduction. The intervention that will be used to reduce food waste is aimed at controlling portion sizes. The intervention is a measuring cup designed to help estimating the right amounts of rice and pasta. The main purpose of this study is to research the moderating role of consumer values on the effect of the intervention on food waste reduction, leading to the following research question:
How do consumer values influence the effectiveness of the intervention in the preparation phase on food waste reduction by consumers?
As a result, this study will contribute to the current literature because it increases the knowledge of the relation between consumer behaviour and food waste. This is necessary, because the research field of consumer behaviour and food waste is a relatively new and therefore not fully explored. Dogan et al. (2014) found that it is important to include both hedonic and egoistic values (types of enhancement) and altruistic and biospheric values (types of self-transcendence) when researching consumer behaviour. Until now, these values have not been studied in combination with an intervention in the preparation phase aimed at reducing food waste.
these findings for designing effective interventions. Since food waste has a wide environmental, social and economic impact and is a growing problem, it requires a better understanding. Therefore, in order to secure a sustainable future, people need to consume less resources, use their resources more efficiently and produce less waste (McKenzie-Mohr, 2000).
The structure of this paper is the following: it starts with a literature review which includes the conceptual model and hypotheses. Next, the methodologies that have been applied will be explained and will be followed by the results. Thereafter, it will end with conclusions, recommendations and possibilities for further research.
2. Literature review
2.1 Food waste and consumer behaviour
Nowadays, “as much as half of all food grown is lost or wasted before and after it reaches the consumer (Lundqvist et al., 2008)”. Especially in developed countries the main source of food losses and food waste comes from consumers and is related to consumer behaviour (Gustavsson et al., 2011; Parfitt et al., 2010). However, in low-income countries, food losses will normally occur early in the supply chain due to low technological support, lack of structure for storing and producing and underdeveloped infrastructure for distribution (Porpino et al., 2015). Food waste caused by consumers has the largest impact on the environment, because by the time the food reaches the consumer it has moved along the complete supply chain. Besides the waste of food itself, also resources and energy will go to waste that have been used during processing, transport, packaging and preparation (The Netherlands Nutrition Centre, 2014). In line with this, Williams et al. (2012) state that “food waste by consumers has a higher environmental impact than food wasted in the distribution chain, and is therefore even more important to reduce”.
In addition, some scholars make a distinction between food waste based on the edibility of the food (Schneider and Obersteiner, 2007). This results in a distinction between avoidable and unavoidable food waste. Avoidable food waste refers to food that could have been consumed if stored or prepared properly (Koivupuro et al. 2012). And unavoidable food waste refers to inedible food like vegetable peels (Koivupuro et al., 2012). In this paper, the term food waste will be described as food wasted by consumers that could have been avoided. According to Parfitt et al. (2010) avoidable food waste occurring at the end of the supply chain relates to the behaviour of consumers.
Previous research of food waste interventions and consumer behaviour was aimed at grouping food together in the refrigerator through colour coding. Colour coding increased the awareness of available food in the fridge, especially for those not directly involved in shopping and storage. Consequently, the increased awareness reduced the amount of expired food (Farr-Wharton et al., 2012). In addition, Quest et al. (2013) found that using a shopping list as an intervention reduces food waste. However previous research on food waste interventions targeted at portion sizes by Vermeer et al. (2014) found that portion size labelling, offering a larger variety of portion sizes, and proportional pricing only had a modest effect on consumption. In general, Matson and Meah (2012) state that “reducing domestic food waste ultimately depends on intervention into these moments, for which we need to better understanding of what relations and processes are significant in making food into waste.”
2.2 Drivers of food waste
Porpino et al. (2015) argue that there is still a lack of empirical studies who focus on identifying the antecedents of food waste in households. Literature regarding the driving factors behind food waste caused by consumer behaviours therefore remains divided and unstructured. A study by Koivupuro et al. (2012) shows that household size, type of household and appreciation of low food prices are all correlated with food waste. Regarding the household size, food waste increases with every additional person in a household. This means that large households waste more avoidable food per person than small households (Koivupuro et al. 2012).
research, parents (mainly mothers), indicated the importance of buying a mixture of healthy food, even if it resulted in food going to waste.
Further, Lyndhurst et al. (2007) indicate that people of lower social class, living in private or social rented accommodation, having a full time job or being a parent at home and young professionals (aged 16-34), are more likely to throw away uneaten food. This is in line with findings of Wassermann and Schneider (2005), who found that younger people waste more than older people. Particularly when people are older than 55-60 years, avoidable food waste is significantly less. In addition to age, also the level of education, type of employment and area of living (urban, suburban and rural) influence the amount of avoidable food waste produced by consumers (Wassermann and Schneider, 2005).
At large, the previous antecedents can be classified as socio-demographic factors influencing consumer behaviour leading to food waste. Research on other types of factors by Lyndhurst et al. (2007) found a strong relationship between ‘poor’ home economic skills and the tendency to waste food. Attributes of home economic skills are pre-shopping planning, discipline in store, looking at labels in store, meal planning and cooking skills. Other factors that can contribute are impulse buying (Parfitt et al., 2010), retail offers and promotions (Schneider, 2008), lack of knowledge about storing and handling food (Parfitt et al., 2010; Koivupuro et al., 2012). Previous research by WRAP demonstrates that food waste by households can occur at five different stages: planning, shopping, storage, preparation and consumption (Flower and Collett, 2014). Food waste by consumer behaviour in the planning phase is caused by failing to make a shopping list and check what is in cupboards, fridges and freezers (Parfitt et al. 2010). In the shopping phase, food waste can be a result of buying too much food. This is especially driven by discounts and special offers, which makes people buy more than necessary (Lyndhurst et al., 2007). Also in the storage phase, a lot of food is wasted due to a lack in understanding the ‘use by’ and ‘best before’ date labels, high sensitivity of food hygiene and not storing it at the right temperature (Parfitt et al., 2010). Last, food waste in the preparation phase is caused by preparing too much and using the wrong preparation method (The Netherlands Nutrition Centre, 2014)
et al. (2015) one of the reasons for preparing an extensive amount of food was related to the aversion of being identified as poor. Moreover, other characteristics related to over-preparation are hospitality, being a good provider and saving time (Porpino et al., 2015).
2.3 Consumer values
2.3.1 Definition consumer values
A major contribution to the consumer value literature has been done by Schwartz (1992), who is the founder of the value theory. Schwartz (1992) classified 56 values based on data from 44 countries. These 56 values were further narrowed down into 10 motivational types of values. In addition, Schwartz (1992) developed a two-dimensional scale with one the one hand self-enhancement versus self-transcendence and on the other hand self-direction versus tradition. In his research, Schwartz (1992) defines values as “a desirable trans-situational goal varying in importance, which serves as a guiding principle in the life of a person or other social entity”. Besides the value theory, Schwartz and Bilsky (1987) found five formal features that form the basis of values in the existing literature. They state that values are “ (1) concepts or believes, (2) about desirable end states or behaviour, (3) that transcend specific situations, (4) guide selection or evaluation of behaviour and (5) events and are ordered by relative importance (Schwartz and Bilsky, 1987).”
Another definition of values by Schwartz and Bilsky (1987) states that “the primary content aspect of a value is the type of goal or motivational concern it expresses.” Linking this to food waste, the most important motives to reduce food waste found by Koens (2006) are ‘waste of money’ and ‘feelings of guilt’. Besides these motives, Bos-Brouwers et al. (2013) also found that environmental impact, saving money, worldwide food scarcity and being an example for one’s children are important motives to reduce food waste.
Existing literature argues that values of consumers are relatively stable over time. This is also confirmed by Vermeir and Verbeke (2006), who define values as “relatively stable beliefs about the personal or social desirability of certain behaviours and modes of existence.” Since, values are stable over time the effects of values on consumer behaviour are typically the same. Because of this, it is possible to make predictions about consumer behaviour based on consumer values.
2.3.2 Impact of consumer values on pro-environmental behaviour
by focussing on consumer values. Since food waste is related to a consumer’s environmental beliefs and behaviour, values can be very useful to get a broader understanding of consumer behaviour leading to food waste.
Confirming this, the psychologist Rokeach (1973) states that consumer values should play a central role in environmental research and is able to merge all the different interests from diverse sciences concerned with human behaviour. Environmental research in the shopping stage by Burgess (1992) indicates that consumer values play an important role in the decision-making process of buying sustainable products and brands. The relationship between consumer values and consumer behaviour has also been identified by Homer and Kahle (1988). They conducted research on the consumption of natural foods and confirmed the existence of a causal relationship between consumer behaviour and consumer values. Moreover, according to Schwartz (1994) values are able to motivate action, give direction and provide emotional intensity. Also Rokeach (1973) states that the main consequence of personal values is behaviour. Consumer values are therefore a good predictor of consumer behaviour.
In order to extend the current knowledge, the value theory developed by Schwartz (1992) is a very important starting point. Based on his theory, values can be divided into two types of self-transcendence values, and two types of self-enhancement values (Schwartz, 1992). When combining the values found by De Groot and Steg (2007; 2009) with the value theory of Schwartz (1992), biospheric and altruistic values can be categorized as self-transcendence values, while egoistic and hedonic values are part of self-enhancement values. Swartz (1992) and De Groot and Steg (2007; 2009) argue that it is important to make a clear distinction between biospheric, altruistic and egoistic values.
In addition, Steg et al. (2014) state that: “hedonic values are important for understanding environmentally relevant beliefs, preferences, and actions, next to egoistic, altruistic, and biospheric values.” For example, hedonic values will often not lead to food waste reduction, since this behaviour does not have positive hedonic consequences, like fun or pleasure. Therefore, it is expected that for consumers with hedonic values the portion size control intervention is not very effective or not effective at all, because using the intervention does not have (many) positive hedonic consequences.
value will drive actions (Schwartz, 1992). In other words, when multiple values are activated simultaneously, the value with the highest priority will be the basis for the corresponding behaviour. Nordlund and Garvill (2002) discovered in their research that people who prioritize self-transcendent values are more aware of the threats to the environment and perceive a stronger moral obligation to act in order to protect the environment than individuals prioritizing self-enhancement values. In other words, altruistic and biospheric values (self-transcendent values) will probably result in more pro-environmental behaviour than hedonic and egoistic values (self-enhancement values).
3.1 Food waste intervention in preparation phase
In this article, the dependent variable is defined as food waste reduction in the preparation phase caused by consumers. The independent variable is the intervention used in the preparation stage aimed at portion control for rice and pasta. So far, existing interventions aimed at reducing food waste by addressing the gap between knowledge, attitude and behaviour have difficulties with recognizing the complexity and dynamics of everyday life (Evans, 2011). In addition, Stern (2000) argues that an intervention must be tailored at the situation of an individual and should be able to remove key barriers that are preventing change.
A possible reason for over-preparation can be to prevent food scarcity. Stuart (2009) states that: “Having surplus, even in excess of what is ever likely to be needed, can be reassuring”. In line with this, Porpino et al. (2015) found that in low-income countries the over-preparation and overstock of food are perceived as a security necessity. Another reason for over-preparation according to Evans (2011) is the availability of food in different quantities, because it encourages consumers to prepare and waste more. In addition, The Netherlands Nutrition Centre (2014) found that one in five consumers purchases large portions if smaller quantities are not available. Since food is often packed in large quantities, estimation of the appropriate portion is more challenging. Moreover, only one in five consumers measure their food on a daily basis during preparation (The Netherlands Nutrition Centre, 2014).
Van Ittersum (2007) state that “ for many individuals, determining how many ounces of pasta to serve themselves for dinner is a relatively low-involvement behaviour that is a difficult nuisance to repeatedly and accurately monitor.” Therefore people often base their portion size on what they normally consume, trust their instincts or are not aware of the right amount (The Netherlands Nutrition Centre, 2014). The portion size control intervention makes it easier to estimate the right portion (amount an average person consumes), resulting in better estimated portions and therefore less waste. In principle, the intervention lowers the costs of determining the right portion size.
Estimation of the right amount of food can also be unknowingly influenced by other norms and cues from the environment (Wansink and Van Ittersum, 2007). For example, larger sized packages in grocery stores, larger kitchenware and larger portions in restaurants can all influence people’s believes about the appropriate portion size (Wansink and Van Ittersum, 2007). Regarding to kitchenware, Wansink and Van Ittersum (2013) found that using larger plates leads to serving more, eating more and wasting more food. Therefore, using a smaller plate can reduce food waste for those people who base portion sizes on the size of the plate instead of how much they eat. Regarding package size, Wansink (1996) performed five studies to examine the effect of package size on usage volume. He found that individuals are encouraged to use more of a product with larger package sizes than with smaller package sizes. Based on the previous findings, it can be concluded that rice and pasta are packed in large quantities and determining the right portion size is difficult and often results in over-preparation (Wansink and Van Ittersum, 2007). In order to solve this problem, interventions can be useful to raise awareness and focus on the cost and benefits of pro-environmental behaviour (De Groot and Steg, 2009). The intervention in the preparation stage will therefore have a positive effect on food waste reduction. Further, due to the intervention, portion sizes will be better estimated and therefore less food is thrown away. This is translated into the following hypothesis:
H1: The intervention in the preparation phase has a positive effect on food waste
3.2.1 Biospheric values
(2014) indicated that 31% of the respondents wants to avoid food waste, because it is better for the environment. However, this number is based on the intention to reduce food waste and thus will not guarantee an actual reduction of food waste.
According to Ganglbauer et al. (2013) any intervention aimed at reducing food waste must take into account an individual’s values and perceptions. Biospheric values will lead to pro-environmental behaviour, when the perceived costs of the behaviour are less than the perceived benefits for the environment. In contrast, Steg et al. (2014) argue that this relationship can also be the other way around. They state that “people evaluate behavioural options in light of how these options will affect the values that are most important to them (Steg et al., 2014).” Independently of what started this relationship, it has been clear that biospheric values and pro-environmental behaviours are connected.
Research on the effects of consumer values on consumer behaviour by De Groot and Steg (2007) highlight that individuals with a biospheric orientation have a stronger intention to donate to environmental organizations. In addition, research by Nguyen et al. (2016) found that biospheric values strengthen active commitment to pro-environmental purchase behaviour. They state that this is done by “enhancing consumers' attitudes towards environmental protection, their subjective norms and environmental self-identity, and by mitigating their perceived inconvenience associated with eco-friendly products (Nguyen et al., 2016).”
Moreover, Perkins and Brown (2012) investigated the relation between pro-environmental behaviour, consumer values and tourism. They found that people with biospheric values are greatly interested in eco-tourism, tourism-specific pro-environmental attitudes and committed to protect the environment. This research also indicates that biospheric values lead to stronger intentions and a higher commitment to act pro-environmentally (Perkins and Brown, 2012). In addition, Steg et al. (2005) found that biospheric values increase environmental awareness and responsibility and strengthen personal norms for taking corrective actions.
more committed to act pro-environmental behaviour (Perkins and Brown, 2012). Therefore they perceive the benefits of using the intervention as more important than the costs, resulting in a higher effectiveness of the portion size control intervention aimed at reducing food waste. Moreover, since consumers with biospheric values are greatly concerned with the environment and quality of nature (Steg et al., 2014), the costs like inconveniences and effort needed when using the intervention, are seen as relatively small in comparison the benefits it provides for the environment. Accordingly, it is expected that biospheric values have a positive moderating influence on the effect of the intervention in food waste reduction. This results in the following hypothesis:
H2: Biospheric values strengthen the effect of the intervention in reducing food
3.2.2 Altruistic values and food waste reduction
Altruistic values are concerned with welfare of other human beings (Steg et al. 2014). In literature, altruistic and biospheric values are often grouped together as both being predictors of pro-environmental behaviour (De Groot and Steg, 2007; 2009; Nilsson et al., 2004; Nordlund and Garvill, 2002). However, there is an important distinction between biospheric and altruistic values. Biospheric values are concerned with the well-being of nature for the welfare of nature itself. Whereas, altruistic values are more related to the welfare of humankind, where the well-being of nature is mainly understood with regard to providing benefits for humans (Perkins and Brown, 2012). This distinction is also supported by De Groot and Steg (2007) who found a significant distinction between altruistic and biospheric values based on three empirical studies. Strong altruistic values will lead to pro-environmental behaviour when the perceived costs of the behaviour are less than the perceived benefits for welfare of other human beings. The Netherlands Nutrition Centre (2014) found that 41% of the respondents state that “there is a lot of hunger in the world” as one of the main reasons to avoid food waste. This percentage is even higher than the 31% who pointed out that food waste should be reduced to save the environment. However, in contrast to these percentages researchers found that biospheric values are a better predictor of pro-environmental behaviour than altruistic values (De Groot and Steg, 2007; Nilsson et al., 2004).
who found that people who do voluntarily work, participate based on the desire to do ‘good’. Therefore it does not matter if the work of a volunteer improves the lives of 10 or that of 100 humans (Dogan et al., 2014). Extending this line of thinking to food waste, it is expected that consumers with altruistic values do not care if using the intervention aimed at portion control reduces a small or large amount of food waste. This is because the use of the intervention is based on the desire to do ‘good’ for the benefit of human well-being and not solely resulting from the quantity of food that is saved.
In addition, Cecere et al. (2014) argue that behaviour leading to the reduction of waste is rarely socially oriented or related to peer pressure but is very susceptible to pure altruistic attitudes. For example, people who reduce waste often have altruistic motivations, which are not automatically related to financial incentives or social norm pressure (Cecere et al., 2014). Altruistic behaviour can therefore be seen as an intrinsic motivation caused by an individual’s altruistic values. In line with this, one of the reasons why consumers care about the well-being of others is because people’s own utility function is directly and positively influenced by this, for example with donating to a good cause (Cecere et al., 2014). As with donating, interventions aimed at reducing food waste can contribute to a good cause, namely reducing world hunger, increasing food security and the overall well-being of others. Altruistic values are concerned with these outcomes as well as with their own utility function and will therefore positively influence the effectiveness of interventions.
However, as in case with biospheric values, the scale of the effect of the intervention on reducing food waste can be less in comparison to the effect of egoistic and hedonic values. This can be due to the possibility that consumers with altruistic values already waste less, since their awareness of the negative consequences for the human well-being is higher (Perkins and Brown, 2012). Despite this possibility, altruistic values still have a positive influence on the effect of the intervention on food waste reduction.
H3:Altruistic values strengthen the effect of the intervention on food waste
3.2.3 Egoistic values and food waste reduction
In contrast to biospheric and altruistic values, Schultz et al. (2005) found that egoistic values have a weaker correlation with environmental beliefs and attitudes and result in less pro-environmental behaviour. Moreover, people with strong egoistic values care less about the environment and are more concerned with personal gains. Egoistic values are together with hedonic values, part of self-enhancement values. Self-enhancement values are concerned with the personal costs and benefits of decisions and behaviour (Steg et al., 2014). Individuals with strong self-enhancement values will therefore behave pro-environmentally when the perceived personal benefits of the behaviour are greater than the perceived costs (Steg et al., 2014). In line with this, De Groot and Steg (2009) argue that when the egoistic costs of acting pro-environmental are perceived as too high, individuals with strong egoistic values will refrain from acting pro-environmental.
Research on ‘eco-tourists’ by Perkins and Brown (2012) indicated that consumers with strong egoistic values were less likely to acknowledge their personal impact when making decisions about traveling. Also in other environmental research, consumers with egoistic values focus more on the costs and benefits when making choices that impact an individual’s resources like wealth, power, and achievement (De Groot and Steg, 2007; Nordlund and Garvill, 2002). These personal resources have also been defined as wealth, social power, influence and authority (Stern et al., 1995; De Groot and Steg, 2007). In addition, egoistic values are correlated with egocentric ethics. According to De Groot and Steg (2007) egocentric ethics imply “that individuals are entitled to extract and use natural resources to enhance their own lives and those of other members of society.” The intervention in the preparation stage will therefore be less effective, since a reduction in the amount of wasted food, is of less importance for consumers with egoistic values.
the amount of money and effort saved by using the intervention, may not be substantial enough to facilitate this kind of pro-environmental behaviour. In addition, Miller (1999) argues that people are primarily motivated by (short-term) self-interest. Therefore, consumers would be more likely to adjust their consumption and preparation behaviour when the behaviour is related to some direct, concrete and tangible individual benefits like monetary benefits (Dogan et al., 2014). However, the money for the food is already spent and benefits of using the intervention may therefore be seen by the consumer as indirect and relatively small.
To sum up, egoistic consumer values negatively influence the effect of the intervention on food waste. Based on the literature, this effect is negative because many researchers argue that egoistic values are negatively related to pro-environmental behaviour (Schultz et al., 2005; Steg et al., 2014) and mainly focus on doing what is best for oneself (De Groot and Steg, 2007; 2009; Nordlund and Garvill, 2002). Further, the personal costs of pro-environmental behaviour are often perceived as higher than the benefits (De Groot and Steg. 2009; Steg et al., 2014). Regarding the use of the intervention, consumers will therefore probably not perceive the increase in personal resources like money sufficient enough and worth the effort it takes to use the intervention. This is because the increase in resources is rather small and in the consumers mind the money is already spent. Therefore egoistic values negatively moderate the effect of the intervention on food waste reduction. This will lead to the following hypothesis:
H4: Egoistic values weaken the effect of the intervention in reducing food waste.
3.2.4 Hedonic values
So far, biospheric, altruistic and egoistic values have been discussed. In the literature, biospheric, altruistic and egoistic values have often been discussed together representing different aspects of consumer values (De Groot and Steg, 2007; 2009). However, it is important to include hedonic values, since they contribute to a more complete understanding of the influence of consumer values on behaviour. This is also argued by Steg et al. (2014), who found that hedonic consequences of behaviour should be taken into account when interventions aimed at promoting pro-environmental behaviour are used, because these hedonic consequences can act as barriers for behaviour change.
Lindenberg and Steg (2007) state that a hedonic goal frame is concerned with: “avoiding effort, avoiding negative thoughts and events, avoiding direct uncertainty, seeking direct pleasure, seeking direct improvement in self-esteem and seeking excitement.” In addition, hedonic values are focussed on behaviour that provides pleasure, joy, fun (Yim et al. 2014).
Focussing more on the consumption of food, hedonic consumption is concerned with pleasure-oriented consumption. Hirschman and Holbrook (1982) state that hedonic consumption is “primarily motivated by the desire for sensual pleasure, fantasy and fun”. Therefore, from a hedonic consumption point of view, food does not only serve as a basic physiological need like utilitarian consumption, but can also fulfil needs like that of fun and pleasure. Moreover, Babin et al. (1994) defined hedonic values as being “more subjective and personal than its utilitarian counterpart and resulting more from fun and playfulness than from task completion.” Hedonic values and food consumption are therefore in the first place more concerned with feelings and emotions relating to food. And second, food waste reduction as a task does not satisfy consumers with hedonic values if it is not combined with playfulness and fun.
Another issue when making environmental choices is the social dilemma consumers face. This dilemma consist of the conflict between individual interests (hedonic and egoistic values) in the short term and collective interests (pro-environmental behaviour) in the long term (Steg et al., 2014). For consumers with hedonic values this conflict is even further complicated by their mood. Because, according to Lindenberg and Steg (2007), the behaviour of consumer with a hedonic goal frame depends on their mood. Therefore, if consumers do not feel like it, they will not act pro-environmentally.
Contrastingly, some literature argues that hedonic values can provide a basis for pro-environmental behaviour. De Young (2000) states that pro-pro-environmental behaviour can bring personal and internal contentment. Based on empirical research, Steg et al. (2014) states that: “people can derive more pleasure and satisfaction from acting pro-environmentally and are more likely to comply with environmental appeals when pro-environmental behaviour is advertised as morally good rather than economical.” This provides evidence that food waste reduction can make consumers ‘feel good’. This positive feeling can improve consumers self-esteem when having strong hedonic values. However, the question remains if ‘feeling good’ is enough to actually strengthen the effect of the intervention on food waste reduction.
strong hedonic values is small. In addition, using the intervention can be perceived as inconvenient and requiring effort, because an additional food preparation action (measuring rice and paste) is needed. Effort is one of the strongest influencers of food consumptions. The amount of effort needed to obtain and prepare food repeatedly explains which type of food and how much food is consumed (Wing and Jeffrey, 2001). Furthermore, consumers with strong hedonic values also want to avoid negative thoughts and events. Using the intervention can remind them of food waste and the negative consequences, which can lead to refraining from using the intervention in the first place.
The intervention can reduce the amount of food prepared, however consumers can then feel like there is not enough food available. For some people having more food available than needed, can feel reassuring (Stuart, 2009). For consumers with strong hedonic values not having enough food can therefore lead to negative feelings like uncertainty. These feelings are the opposite of what hedonic values stand for. Consumers with strong hedonic values don’t want to worry about having enough food. They want just want to enjoy food and experience pleasure, joy and fun when preparing and consuming the food (Yim et al. 2014). The effect of the intervention on food waste reduction will therefore be less effective for consumers with strong hedonic values. Although acting pro-environmental behaviour can make consumers feel good (De Young, 2000; Steg et al. 2014), this feeling is expected not outweigh the negative implications stated above. To sum up, for consumers with strong hedonic values the use of the intervention in the preparation phase will have a negative moderating effect, because using it takes effort, does not provide pleasure and fun and can trigger negative feelings and thoughts, which they like to avoid (Lindenberg and Steg, 2007). Therefore hedonic values weaken the effect of the intervention of food waste reduction. This results in the following hypothesis:
H5: Hedonic values weaken the effect of the intervention in reducing food waste.
3.3 Conceptual model
waste. Moreover, egoistic and hedonic values weaken the effect of the intervention on food waste.
Figure 1: Conceptual model
First the data collection methods and participants will be explained. The data collection methods include a food waste diary and a questionnaire.
4.1 Data collection
For the collection of the data two supplementary approaches are used. These approaches are (1) a food waste diary and (2) a questionnaire. The research was conducted on Friday 11 November and Friday 18 November 2016 in the ‘Albert Heijn’ supermarket located in Paterswolde in the Netherlands. Consumers that visit this supermarket are expected to give a good representation of the households as defined in the literature. Further, the 11th and 18th of November have been
chosen, because at these days there are no holidays or other special days that can influence the usual stream of customers. All in all, these decisions have been taking into account to get a representative and substantial sample for the research.
After agreeing with the informed consent, customers will receive the questionnaire and the link to the online questionnaire. In addition, they will receive the food waste diary. Regarding the intervention, customers will either receive the intervention in the preparation phase (the Eetmaatje) or receive nothing. The customers receiving no intervention will serve as the control group. For the data collection 70 participants of the intervention group and of the control group will be recruited. This will lead to a sample with a maximum of 140 participants.
4.2 Food waste diary
The food waste diary consists of three different parts and will be filled in for seven consecutive days. The food waste diary can be found in appendix A. The first part measures how often a person has ordered a meal, for example take away or dining out in a restaurant. The second part measures what kind of groceries have been done. Participants can choose between: small groceries, large groceries or extra groceries in case they forgot something or did not buy enough. The third and final part of the food waste diary focusses on the amount of food and drinks wasted within the household. The amount of wasted drinks are measured in millilitres, with one glass consisting of 150 millilitres. The amount of food is measured in grams, with portion sizes in line with the type of good. When food is thrown away, participants have to indicate the wasted amount of food and drinks themselves. The portion sizes have been included to simplify the measuring process and facilitate a more precise estimation of the amount of wasted food. However, the food waste diary has some limitations since participants are measuring the food without supervision. According to Koivupuro et al. (2012) participants may forget or choose not to record some of the wasted food within the seven days. Moreover, being aware of the research and reporting all the wasted food can also result in a lower amount of reported food waste in comparison to the normal situation (Koivupuro et al., 2012). Both limitations can cause a decrease in the reported amount of food waste.
Furthermore, the focus of this food diary is on avoidable food waste, therefore unavoidable food waste has been left out. Unavoidable food waste has been defined as waste from food or drink preparation that is under normal not edible (Quested, 2009). Examples of unavoidable food waste are meat bones, egg shells, pineapple skin, tea bags.
4.3 The questionnaire
is Dutch. The first eight questions focus on the socio-demographics of the household (e.g. age, gender, education, profession, pets and household size). In addition, questions are asked regarding an individual’s economic background (e.g. income). Further, this is followed by questions regarding individual’s shopping behaviour of food (e.g. amount of time spent on doing groceries, size of grocery shopping, usage of shopping list, sticking to the shopping list). The questions regarding consumer values are based on research by De Groot and Steg (2007). For, biospheric, altruistic and egoistic values a 9-point Likert scale is used. Each value is measured on the basis of four statements. Individuals have to fill in to which extent these statements are in line with their principles. To measure the hedonic values a scale of Steg and colleagues (2014) is used. The hedonic values are also measured on a 9-point Likert scale. However, instead of four statements now three statements will be used. Both researches provide support for the validity and reliability of these value instruments (De Groot and Steg, 2007; Steg et al., 2014). All questions regarding customer values are presented below (Table 1).
Self-transcendence values Self-enhancement values
Biospheric values Egoistic values
Preventing pollution: protecting natural resources
Social power: control over others, dominance
Respecting the earth: harmony with other species
Wealth: material possessions, money Unity with nature: fitting into nature Authority: the right to lead or command Protecting the environment: preserving nature Influential: having an impact on people
Altruistic values Hedonic values
Equality: equal opportunity for all Pleasure A world at peace: free of war and conflict Enjoying life Social justice: correcting injustice, care for the
Gratification for oneself Helpful: working for the welfare of others
5. Descriptive statistics and reliability analysis
5.1 Descriptive statistics
The original dataset started with a total of 70 respondents. From this original dataset three respondents were excluded due to a missing respondent number. Because of the missing respondent number the food waste diary could not be linked to the corresponding respondent. In addition, 27 respondents were excluded since the food waste diary was not filled in correctly. In these cases the consumption instead of waste was reported in the diary. In total, 30 respondents have been excluded to improve the reliability and quality of the analyses. This resulted in a complete dataset with 40 respondents. Of these respondents 18 did not receive an intervention and served as the control group, representing 45% of the sample. The 22 remaining respondents did receive an intervention aimed at portion size control, representing 55% of the sample. However, of these 22 respondents in the intervention group, only 15 stated that they actually used the intervention. Reasons for not using the intervention will be addressed in the discussion section.
The sample consisted of 9 men and 31 women. In percentages 22.5% of the respondents were male and 77.5% was female. The respondents had a mean age of 55 years (M = 54.55, SD = 13.08). However, 7 out of the 40 respondents did not fill in their age. The youngest participant has an age of 25 years and the oldest respondent an age of 84 years. The majority of the respondents has a household size of 2 people (47.5%). The other respondents have a household size of respectively 3 people (22.5%), 4 people (20%) or one person (5%). Of the 21 households with children, the average age of the children is 13 years. In addition, 45% of the respondents was an owner of one or more pets.
more (32.5%). This indicates that on average the respondents have a high income. In addition, the remaining 27.5% has an income between €1000 - €2999.
5.3 Food waste
The dependent variable food waste can be analysed from two different perspectives. The first possibility is to take into account drinks, diary, soup and other liquids and second it is possible to solely focus on solid food. Table 2 provides an overview of both types of food waste. In addition, the average, maximum, minimum and standard deviation are displayed. Food waste has been measured in grams.
The total average amount of food waste including liquids is 873.90 gram and the average amount of food waste excluding liquids is 699.40 gram. For both, food waste with and without liquids, a minimum of 0.00 and a maximum of 5254.00 is stated. In addition, the average amount of waste is described of both the control and intervention group. Regarding food waste including liquids the control groups wastes on average 765.00 gram and the group who received the intervention 601.47 gram. For food waste excluding liquids the control groups wastes on average 594.17 gram and the group who received the intervention 490.13 gram.
Sample Minimum Maximum Mean Std. Deviation
Food waste including liquids 40 .00 5254.00 873.90 1089.73
Food waste excluding liquids 40 .00 5254.00 699.40 980.83
Table 2: Descriptive statistics of total amount of food waste
5.4 Consumer values
value resulted in an α=.512. This is below the minimum of .6. Therefore, item 4 was deleted regarding the question “Helpful: working for the welfare of others”. This resulted in an α =.603 and therefore aggregation and averaging of the remaining three items into one scale is allowed. For the egoistic value an alpha of .691 was found. Deleting item 1 regarding “Social power: control over others, dominance” would have resulted in an alpha of .755. However, since the scale came from previous literature and is above the minimum, no items were deleted. Therefore, all four items were aggregated and averaged into one scale for altruistic value. Last, reliability analysis for the hedonic value resulted in α= .907. Deleting item 3 regarding “gratification for oneself” would have resulted in an alpha of .926. However, the alpha is already very close to 1 and therefore optimization is not necessary and would only result in a loss of data. Cronbach’s alpha New Cronbach’s alpha Mean * Std. Deviation * Biospheric values .862 - 7.544 0.866 Altruistic values .512 .603 7.925 1.198 Egoistic values .691 - 3.225 1.285 Hedonic values .907 - 8.017 0.978
* Mean and Std. Deviation are calculated for the final variable
Table 3: Reliability analysis of consumer values
5.5 Control variables
It is important to take into account control variables, because they can influence the results of the relationships between the dependent and independent variables. The influence of the control variables is checked by using a One-way Anova, regression and an independent sample t-test (depending on the type of control variable), in order to find out if there is significant relationship with food waste. For further analyses only the significant control variables are taken into account. Age, level of education, type of employment, income, working hours per month and household size have therefore not been included as control variables.
In addition, an independent samples t-test was used to see if the average amount of food waste is different between households with or without a pet. For food waste including liquids the independent samples t-test was significant, t(38) = -1.66, p = .087 (10% confidence interval). For food waste excluding liquids the independent samples ttest was also significant, t(38) = -1.37, p = .095 (Table 4). Therefore, the average food waste of households with a pet is significantly different from the average food waste of households without a pet.
Control variable DV df t p-value
Gender Food waste including liquids
Food waste excluding liquids
38 38 1.75 1.78 .006 .007 Ownership of pet Food waste including liquids
Food waste excluding liquids
38 38 -1.66 -1.37 .087 .095 Table 4: Overview of significant control variables for both types of food waste
Next, to gain knowledge about the relationships between all the independent and dependent variables in the model a Pearson correlation was used (Appendix C). From here on, it has been chosen to only present the results from food waste excluding liquids, also called solid food waste. This decision has been made because the intervention primarily focusses on solid food, namely rice, pasta and couscous, and more significant relationships are found. Based on the correlation matrix the dependent variable solid food waste has a negative relationship with the altruistic value variable (p = .091, r= -0.299). Another negative significant relationship was found between the control variable gender and egoistic values (p= .013. r= -0.428). Also the control variable ownership of pets was positively correlating with biospheric values (p=.001, r=-0.551).
In this chapter the results will be presented from the analyses of the hypotheses stated in the literature review. For these analyses SPSS 23 was used. First, the relationship between food waste (solid food) and the use of the intervention aimed at portion size control, was tested. Secondly, it was tested if this relationship is moderated by the different consumer values. During these analyses the control variables have been taken into account.
6.1 Hypothesis 1 – Use of the intervention
First, it is important to state that the respondents who received the intervention but did not use it, are not included in this analysis. This was done, because is resulted in more significant relationships. Therefore the sample consist of 33 respondents. In addition, solid food waste is used as the dependent variable, because is also lead to more significant relationships.
To test the relationship between the use of the intervention and food waste a One-way Anova was used (Table 5). A One-Way Anova shows whether there is a statistically significant difference between the means of two groups on the variable food waste (dependent continuous variable). The two groups consist of one group who did use the intervention and the other group who did not use the intervention. The significance value is .220 (p=.220), which is above the confidence interval of 10% (p > .10). Therefore, it can be concluded that there is no significant difference in the average food waste of respondents who did or did not use the intervention. This relationship is also graphically presented in figure 2.
Df Mean Square F-stat. Significance
Between groups 1 1659519.282 1.564 0.220
Within groups 31 1060862.588
34 Figure 2: The relationship between solid food waste and use of the intervention
The Levene test is significant with a p-value of .017 and a Levene statistic of 6.425. Therefore it can be concluded that there is a significant difference in the variances. This means one of the assumptions regarding Anova is disregarded. In this case a corrected test is necessary, therefore a Welch Anova will be performed. This is an appropriate test when sample sizes are unequal and there is a significant difference in variances. The results of the Welch test report a significance of p = .191. This value can be used to replace the p-value of the One-Way Anova. However, this effect still remains insignificant. Therefore, hypothesis 1 is not confirmed.
6.2 Hypothesis 2 – Biospheric values
In the conceptual model it is predicted that biospheric values have a moderating influence and strengthen the effect of the intervention on food waste reduction. Due to the small sample size all the moderators will be tested separately. Therefore, the type of moderation is a first-order single moderation and will be tested with a linear regression model. The model has a R square of .214, which means that the independent variables explain 21.4% of the variance in the amount of food waste. In addition, the adjusted R square is estimated at .173 and the complete model is significant with a p-value of .069 (p< .10). Therefore, a regression is an appropriate model to use for these data.
The results from the model show a significant negative direct effect of biospheric values on food waste (p = .035, β= -.512). Meaning that biospheric values decrease the amount of food waste. And a significant negative main effect regarding the use of the intervention was found (p =.092, β= -1.774). On the other side, there is no significant effect of the moderator biospheric
940,5 490,1 0 100 200 300 400 500 600 700 800 900 1000 0 1
Use of the intervention (no = 0, yes =1)
values (p = .123, β= 1.619). It is interesting to see that the moderating effect is not significant and has a positive beta, since both the direct effect of biospheric values and the main effect of the intervention have a significant negative effect. The results can be found in table 6 and are graphically presented in figure 3. Moreover, the VIF-scores of the intervention (39.948) and the moderator (39.905) are not below the critical threshold and should therefore be interpreted with caution. The other VIF-scores are below the critical threshold, with a range between 1.165 and 2.048.
Variable Beta T-stat. p-value
(Contstant) 2.639 .014*
Intervention -1.774 -1.746 .092*
Biospheric value - .512 -2.227 .035*
Intervention * Biospheric value 1.619 1.594 .123
Ownership of pets .182 .951 .088*
Gender - .307 -1.771 .350
* Significant at level .10
Table 6: Moderating effect of biospheric values on the effect of the intervention on food waste (N=33)
Figure 3: Moderation of biospheric values on the effect of the intervention on food waste
6.3 Hypothesis 3 – Altruistic values
For altruistic values it is predicted that it also has a moderating influence and strengthens the effect of the intervention on food waste reduction. The model has a R square of .244, which means that the independent variables explain 24.4 % the variance in the amount of food waste. In addition, the adjusted R square is estimated at .104. However, the complete model is not
0 500 1000 1500 2000
Low biospheric values High biospheric values
significant with a p-value of .160 (p > .10). Therefore, a regression is not an appropriate model to use for this data. One explanation for this finding can be the small sample.
Since a linear regression model is not appropriate, a Two-way Anova was used. First it is important to perform a median split on the independent variable altruistic value, because the independent variables should consist of 2 or more categorical groups. The Levene’s test is significant with a p-value of .001. Therefore, there is a significant difference in variances. Second, the Test of Between-Subjects Effects is presented in table 7 and shows that there is no significant moderating effect (p =.564, F = .341). However, a direct effect of the moderator altruistic values was found significant (p = .079, F =3.321). This means that the difference in the amount of food waste between the control and intervention group depends on the level of altruistic values. In order to find out if this relationship is positive or negative the results have been graphically presented in figure 4. This figure indicates that altruistic values have a negative direct effect, therefore resulting in less food waste.
Variable F-stat. p-value
(Contstant) 1.981 .171
Intervention .677 .418
Altruistic value 3.321 .079*
Intervention * Altruistic value .341 .564
Ownership of pets 2.851 .103
Gender 3.566 .070*
* Significant at level .10 ** Significant at level .05
Table 7: Direct effect of altruistic values on food waste (N=33)
0 200 400 600 800 1000 1200 1400 1600
Low altruistic values High altruistic values
6.4 Hypothesis 4 – Egoistic values
Regarding egoistic values the following prediction has been made in the conceptual model: egoistic values weaken the effect of the intervention in reducing food waste. To test the moderating effect of the egoistic values a first a linear regression was used. The model has a R square of .179, meaning that independent variables explains 17.9% of the amount of variance in the dependent variable. The adjusted R square has a value of .027. Here we can already see that the independent variables explain very little variance in the dependent variable. Further, the model is insignificant with a p-value of .346 (p>.10, F = 1.178). Therefore the model is not appropriate to use and to draw any conclusions on. Biospheric values therefore do not strengthen or weaken the effect of the intervention on food waste reduction.
Also in this case a Two-way Anova test was used to gather further information about the relationships. First, a median split was used on the egoistic value variable to divide it into low and high egoistic value. However, in this model there were no significant effects found, which is displayed in table 8.
Variable F-stat p-value
(Contstant) 1.249 .274
Intervention .659 .424
Egoistic value .059 .809
Intervention * Egoistic value .032 .860
Ownership of pets 2.289 .142
Gender 2.042 .164
* Significant at level .10 ** Significant at level .05
Table 8: Results two-way Anova egoistic values, food waste and intervention (N=33)
6.5 Hypothesis 5 – Hedonic values
For hedonic values it is predicted that they have a moderating influence and weaken the effect of the intervention on food waste reduction. The model has a R square of .331, which means that the independent variables explain 33.1% the variance in the amount of food waste. In addition, the adjusted R square is estimated at .207 and the complete model is significant with a p-value of .044 (p< .10). Therefore, a regression is an appropriate model to use for these kind of data.