Understanding the role of consumer
behaviors associated with food waste
Analyzing the effect of a shopping list, cooking enjoyment and sale proneness
1
Understanding the role of consumer behaviors associated with food waste Analyzing the effect of a shopping list, cooking enjoyment and sale proneness
Completion date: 16th January, 2016
Master Thesis
Msc Marketing, specialization Marketing Management University of Groningen, Faculty of Economics and Busisness
By L.V. MULDER Student number: 2031124 Breuningslaan 62 8471 ZT Wolvega Phone: +31 (0)610554118 E-mail: laura_mulder@live.nl
1st supervisor: Dr. Jenny van Doorn
2
Abstract
With a field experiment, this study aims to examine the effectiveness of using an enhanced shopping list when grocery shopping in the supermarket on the amount of food waste. In addition, the influence of cooking enjoyment and sale proneness is included in the model. By recruiting participants in a local supermarket, the data was gathered to measure and analyze the hypothesized relationships. However, the data analysis showed no significant results for the influence of a shopping on food waste generation. Moreover, the results for the effect of cooking enjoyment and sale proneness also came out not significant. Therefore, no reliable conclusion can be drawn from the results and future research is needed to dive deeper into the topic of food waste and its drivers.
3
Table of Contents
1. INTRODUCTION 4 2. THEORETICAL FRAMEWORK 7 2.1 Food waste 7 2.2 Shopping list 102.2.1 Shopping lists and food waste generation 14
2.3 Cooking enjoyment 14
2.3.1. The role of cooking enjoyment 17
2.4 Sale proneness 17
2.4.1. The influence of sale proneness 19
2.5 Conceptual model 20
3. METHODOLOGY 22
3.1 Research design 22
3.2 Procedure 22
3.3 Data collection 24
3.3.1 Food waste diary 24
3.3.2 Questionnaire 24 3.4 Measurements 25 3.4.1 Food waste 25 3.4.2 Cooking enjoyment 25 3.4.3 Sale proneness 25 3.4.4. Control variables 26 4. RESULTS 26 4.1 Data 26 4.1.1 Data preparation 26 4.1.2 Sample descriptives 27 4.1.3 Random assignment 29 4.2 Pre-analysis 29 4.2.1. Reliability analysis 29 4.2.2 Normality test 31
4.2.3 Homogeneity of variances test 32
4.3 Hypotheses testing 33
4.3.1. The effect of using an enhanced shopping list on food waste 33
4.3.2 The interaction effect of cooking enjoyment 35
4.3.3 The interaction effect of sale proneness 37
5. DISCUSSION 39
5.1 Limitations 41
REFERENCES 43
APPENDIX 47
Appendix I – Informed consent form control group 47
Appendix II – “Enhanced” shopping list 48
Appendix III – Food waste diary 49
Appendix IVa – Questionnaire control group 51
4
1. INTRODUCTION
In recent years, increasing attention has been given to the economic and environmental impact of food waste and the positive outcomes of tackling this waste at the source. In the European Union, food waste is estimated around 88 million tons annually, representing costs around 143 billion euros (FUSIONS, 2016). Besides the fact that wasting food is an ethical and economic issue, it depletes the earth’s natural resources.
In high income countries, food waste generated by households makes up roughly half of the total amount of food being wasted (Stancu et al., 2016). This makes the household level of the food chain, and thus consumers, the biggest contributor to food waste (Parfitt et al., 2010). It illustrates the necessity to develop an effective intervention for individual consumers that will bring down these numbers. Focusing food waste campaigns about the negative consequences resulting from their current behavior on this specific target group has the potential to have a substantial positive environmental effect. Preventing food from being wasted in the first place reduces the energy, water and other limited resources that are necessary to fill the shelfs in supermarkets with food has the largest environmental benefit (Quested et al., 2013). Thus, in order to successfully reduce consumer-related food waste, it is useful to generate a deeper understanding of the underlying factors contributing to food waste-related consumer behaviors. Yet, there is a lack of knowledge in the current literature about individual consumer
characteristics related to food waste generation.
5
Shopping lists help with purchasing only what is necessary and avoid making unnecessary impulse purchases (Block et al., 1999). Taking note of inventories, has the potential to have a positive effect on food waste reduction (Chandon & Wansink, 2006) by preventing consumers from underestimating their food inventory at home and purchasing products they already bought in a previous shopping trip. Moreover, making a structured shopping list could also help
consumers to limit unplanned purchases and decrease food waste (Bell et al., 2011). To investigate this aspect of food waste generation, the following research question will be addressed:
Research question 1: How can the use of a shopping list during grocery shopping contribute to food waste reduction?
A central question that arises in the food waste discussion is: Why do consumers buy groceries they never us and finally throw away? By creating an understanding of this problem, consumers can become more aware of their purchase and usage habits. This will lead to a reduction of product abandonment that allows consumers to save money and reduce their food waste generation. A reasonable part of an explanation for this phenomenon stems from motivation: When someone purchases a product, he or she has presumably the motivation to consume it (Prince, 1991). Somewhere down the line, there is a discrepancy between this motivation when purchasing the product(s) and the actual behavior that occurs returning home with the
product(s).
Adding to this line of reasoning, Wansink et al. (2000) argue that when the ability to use a product in the situation that is was intended for passes, a narrow window of further usage
6
This consumer characteristic has not yet been linked to food waste generation in the literature as a potential moderating variable. However, because of the predominant role it plays when it comes to unused product (and thus food waste), it deserves more attention and will be
investigated deeper in the current study. To investigate this aspect of food waste generation, the following research question will be addressed:
Research question 2: How is the relationship between an enhanced shopping list and food waste reduction affected by cooking enjoyment?
In previous literature, an important consumer trait that is often discussed when it comes to marketing-related topics is the concept of sale proneness. In this article, it is suggested that sale proneness could have a moderating effect on the relationship between using an enhance
shopping list and food waste. Because sale-prone consumers derive pleasure out of searching for sales to get products at the lowest price (Jin & Sternquist, 2003), they are more likely to respond to price promotions. In 2011, WRAP conducted a study investigating the possible impact of promotions on food waste generation. This study found a link between some types of promotions (multi-buy and y for $x) and the purchase of larger amounts of food. Because discounts are often offered on products in large(r) quantities, these packages are prone to becoming a source of food waste due to too large quantities and consumers not being able to consume these products before the expiration date (Halloran et al., 2014).
7
To investigate this aspect of food waste generation, the following research question will be addressed:
Research question 3: In what way does sale proneness influence the effect of using an enhanced shopping list?
Based on the research questions, this paper will contribute to the existing literature by providing insights into how better pre-shop planning could help with minimizing food waste generation. It will be analyzed whether encouraging the use of a shopping list will be a prosperous
intervention to reduce food waste. Moreover, it will be investigated how cooking enjoyment and consumer sale proneness may influence the effects of using the shopping list. The relevance comes from being able to communicate more effectively and personally target consumers about wasting food and how to reduce it with the insights gained from this study.
This paper will continue with the following structure: In the next chapter, the literature to date will be reviewed to form the foundation for the current study. Next, the conceptual model and the accompanying hypotheses that were developed will be discussed. After this, the
methodology that is used to gather data for the analysis is presented. Then, the results of the data analysis are described. Finally, the results are discussed, together with their (managerial)
implications and the limitations of the study.
2. THEORETICAL FRAMEWORK
2.1 Food wasteFood losses could occur at any point in the food supply chain resulting from inappropriate behavior of producers, retailers or households (Leal Filho & Kovaleva, 2015). Before
8
In the Netherlands, food being wasted by consumers has a value of approximately 2.5 billion Euros1 pet year. This means that each household throws away 340 Euros worth of food each
year, equivalent to 150 Euros per individual consumer. This demonstrates the hedonic beneficial effect for consumers (e.g. money savings) when keeping food waste at a minimum. Because household food waste accounts for approximately 50 percent of the total amount of food waste (Stancu et al., 2016), addressing this type of waste and finding a way to minimize it is both necessary and urgent. Therefore, this study will focus on the underlying consumer behaviors and characteristics that are responsible for food waste. These aspects will be investigated into further detail in the sections below.
In their study, Gustavsson et al. (2011) found that approximately 33 percent of the total amount of food that is purchased ends up as food waste. Animal nutrition and non-edible product parts are excluded from food waste and thereby only those products that are meant for human
consumption are included. In other words, the study will only include food waste measurements of food that was edible at the point of purchase, but ended up being wasted for any reason whatsoever. Following their approach, food waste will be measured as the discrepancy between the amount of food purchased and available for consumption and the amount that eventually ends up in the trash bin (Griffin et al., 2009).
Continuing, Quested and Johnson (2009) make an additional categorization within the household food waste concept based on whether the waste could have been prevented or not: Avoidable, possibly avoidable and unavoidable food waste. The first category, which accounts for the major part of the total amount of food waste, refers to products that were still appropriate for human consumption when thrown away (e.g. a slice of bread) and thus avoidable. The second category refers to products that are not consumed by everyone (e.g. a bread crust) and products that can be consumed when prepared a certain way (e.g. the skin of a fried salmon) and thus possibly avoidable. The third category refers to products that are not suitable for human consumption (e.g. fish bones) and thus unavoidable. For the purpose of this study, only
avoidable food waste will be included in the analysis. Also measuring possibly avoidable food waste is too difficult, because the method to gather food waste data should then be personalized and adjusted to individual consumers. Unavoidable food has no relevance, because it can’t be prevented with any kind of intervention and thus doesn’t deserve any attention in this study.
1Numbers are taken from a study of the Ministry of Economic Affairs, retrieved from
9
Food waste can occur at different phases of the food process: the acquisition, the preparation and the consumption stage (Griffin et al., 2009). According to Evans (2012), this food waste is a consequence of the way household practices are socially and materially organized. It is
necessary to mention that that food waste generation should not be addressed as a single behavior, but as an outcome of multiple behaviors concerning planning, shopping, storage, preparation and consumption of food (Quested et al., 2011). This makes the moment when food is being wasted the final step of the chain and is thus the opportunity to prevent this food from being wasted already passed. The actions and decisions that have led to this waste have been made earlier (Stefan et al., 2013). Therefore, the suggestions and interventions to minimize food waste should begin with addressing consumer behaviors in the early stages of the process.
Previous research has identified several reasons for household food waste to occur, including purchase behaviors of consumers and the failure to follow a shopping list (Evans, 2012;
Lyndhurst et al., 2007; Graham-Rowe et al., 2014). In particular, Evans (2012) found that over-purchasing contributes largely to the total amount of household food waste, because most household have developed shopping routines that make them structurally purchase more products than necessary and hence waste the majority of this surplus. An intervention to reduce food waste generation should therefore be strong enough to overcome these routines.
Explanations for over-purchasing can be found in various consumer behaviors. Especially, the desire to be a “good provider” for the household forms a strong predictor for over-purchasing (Graham-Rowe et al., 2014). However, explanations for over-purchasing can also be found within the retail setting. In-store marketing tactics also play a crucial role in the food waste problem. For instance, certain products are only available in large(r) quantities and therefore push consumers to buy more than they need (Evans, 2012; Graham-Rowe et al., 2014). Also, volume discounts encourage consumers to over-purchase in order to save money (Lyndhurst, 2008).
10
They over-estimate the proper amounts of ingredients and this results into households wasting more than necessary. However, leftovers from previous meals can be re-used, but either consumers lack the knowledge to prepare these leftovers or they are unwilling to do so (Cox & Downing, 2007). This demonstrates the necessity of better planning and increasing knowledge about the topic when designing the intervention.
Besides the different antecedents and outcomes of food waste that are already discussed in previous literature, the relationship between a shopping list and food waste generation has not received enough attention in the existing literature, especially by means of an experiment. Therefore, this relationship will be investigated into deep in this study and more insights into the food waste problem will be provided.
2.2 Shopping list
Unplanned purchases account for roughly 60 percent of all purchases that are made (Inman & Winter, 1998). Moreover, a lack of planning shopping trips in advance and buying grocery products on impulse result into excessive purchasing and eventually lead to an increase in the amount of food waste (Porpino et al., 2015). Also, strategies that are used to save money actually can have the opposite effect. For instance, buying groceries in bulk (e.g. buy one, get one free discounts) has been demonstrated to increase food waste generation (Porpino et al., 2015). Continuing, a consumer that is walking through a supermarket has to make multiple purchase decisions and is faced with numerous distractions when doing so. The issues of unplanned and impulse purchases, a lack of planning and difficult decision-making processes need to be addressed with the intervention to be able to reduce food waste.
Quested et al. (2013) suggest two ways to attempt to reduce the amount of household food waste: 1) By influencing individual actions of consumers, or 2) By changing the way food is sold. Using a shopping list while grocery shopping could be an answer based on the first
11
The shopping list that will be analyzed is tangible and written on a piece of paper. A written shopping list is made up of a situation-specific external memory aid (Block & Morwitz, 1999) that encodes the purchase intentions of consumers. The content of a shopping list suggests the degree to which consumers make their purchase intentions concrete while constructing their lists (Spiggle, 1987). Including inventories when making the shopping list can have a positive effect on food waste generation (Chanson & Wansink, 2006). It prevents consumers from
underestimating their food inventory at home and purchasing products they already bought in a previous shopping trip.
A model that can be found in the literature concerning shopping list is the
“motivational-volitional model of action phases” (Schmidt, 2012). This model is based on the assumption that consumers go on a shopping trip because they are motivated to achieve a goal. It tries to outline explicit temporal tasks that turn intentions into behavior by four sequential steps:
1. Pre-decision: Establishing preferences based on wishes. 2. Pre-action: Planning goal-directed actions.
3. Action: Effectively executing the previously established goal-directed actions. 4. Post-action: Evaluating outcomes.
The shopping list intervention has to ensure that someone’s willingness to reduce the amount of food waste (Evans, 2011) becomes a motivation. For instance, this can be accomplished by emphasizing the hedonic or environmental beneficial effect of using the shopping list. The goal of the shopping list will be to reduce food waste generation.
The intentions of consumers are translated into the items on the shopping list in the pre-decision stage and the shopping list will function as an external memory aid (Block & Morwitz, 1999) in the action step.
Arnaud et al. (2015) have identified three stages that are connected to shopping lists, respectively a development stage, a fulfilment (or usage) stage and an outcome (or post-purchase) stage. This model can be compared with Schmidt’s (2012) “motivational-volitional model of action phases” that is discussed above.
12
Pre-planning a shopping trip by making a shopping list increases the likelihood that only products that are listed will be purchased (Kelly et al., 2000). Because the intention (or goal) is to purchase the items on the list, efficiency is increased and average expenditure is reduced (Thomas & Garland, 1993). In other words, this means that consumers using a shopping list spend less on groceries than consumers without a shopping list. Continuing, households who spent more on groceries turn out to throw away more than households with a smaller food budget (Parizeau et al., 2015). Therefore, paying less is connected to also throwing away less and thus minimizing food waste. The shopping list intervention will realize the above mentioned effects.
• Phase 2 represents the different purposes of shopping lists. This phase is
interchangeable with Schmidt’s (20212) action step. The most commons reasons found for using shopping lists include 1) having a memory aid 2) having control over the shopping process and 3) controlling expenditures (Thomas & Garland, 2004). The shopping list can function as a memory aid when a consumer has to make multiple decisions and is exposed to different distractions during the shopping trip. It can also help simplify the decision-making process, which can reduce impulse buying and help consumers resist in-store bargains (Baumeister, 2002). Moreover, a shopping list can contribute to having control by means of assisting a consumer in the process of only purchasing items that are on the list and thus prevent over-purchasing, which was previously indicated as a major driver of food waste. Shopping lists can also function as a planning tool in respect that the purchase activities are planned beforehand, as well as the preferred brands. This can limit the effect of in-store stimuli (Inman et al., 2009).
• Phase 3 represents the outcomes or consequences from the use of a shopping list. This phase is interchangeable with Schmidt’s (20212) post-action step. Previous research indicates three outcomes related to using a shopping list, namely reducing unplanned purchases, reducing the time spend in-store and reducing expenditure (Arnaud et al., 2015).
13
1. Pre-planned brand-specific items (e.g. Coca Cola light) 2. Pre-planned product specifications (e.g. Cola)
3. Pre-planned product class items (e.g. diary or meat)
4. “Need recognition” (e.g. “Something to eat for tomorrow’s breakfast”) 5. “Need not recognized” (e.g. Some unaware need to be activated in-store)
The shopping list intervention will aim to stimulate consumers to be specific when planning their purchases. Making a structured shopping list can help consumers to limit unplanned purchases (Bell et al., 2011) and decrease food waste.
In summary, shopping lists can have a number of purposes. They can serve as a mean to make the time that a consumers spends in a supermarket as efficient as possible (Thomas & Garland, 1996). Also, they can serve as a memory aid to remind consumers which items they need to purchase (Block & Morwitz, 1999). Using a shopping list is also an indication of pre-shopping planning (Polegato & Zaichkowsky, 1994). Moreover, it has been suggested that consumers use shopping lists to manage their expenditures and make sure they don’t purchase excessive
products (Block & Morwitz, 1999). Especially for families, a shopping list can function as a tool to make the coordination between the family members more effective (Polegato &
Zaichkowsky, 1994). It simplifies the grocery shopping trip by giving the activities an order (Arnaud et al., 2000). Finally, a shopping list helps consumers to stick to their plan and not be distracted by different stimuli, to achieve the goal of the shopping trip (Inman et al., 2009).
Thomas and Garland (1993) already found a relationship between using a supermarket shopping list and its effect on consumer expenditure. In their study, the authors analyzed two important factors for supermarket retailers, namely time spent in store and grocery expenditure. The behavior of two matched samples was compared, one with a shopping list and one without. Their findings demonstrated that shopping lists significantly reduce the average amount of money that is spend on groceries. Moreover, they found increased expenditures for consumers that were accompanied by children and that these shoppers spend more time in-store.
14
This study will go beyond the ‘ordinary’ shopping list by grouping grocery items in different product categories, including inventories, planning meals in advance and suggesting recipes for using left-overs. From this point on, it will be referred to as an ‘enhanced shopping list’.
2.2.1 Shopping lists and food waste generation
Based on the reviewed literature about shopping lists, it is hypothesized that the use of a shopping list during grocery shopping has a positive effect on the reduction of the amount of food waste generation. This means that consumers will be able to plan their shopping trips and grocery product purchases better with the help of a shopping list and thus throw away less in the end. Moreover, it is argued that shopping lists reduce impulse purchases (Baumeister, 2002; Block et al., 1999; Bell et al., 2011) and prevent over-purchasing, which is identified as an important cause of food waste generation (Evans, 2012). This leads to the formulation of the first hypothesis:
H1: Using an enhanced shopping list during grocery shopping positively contributes to food waste reduction.
By adding the elements of the enhanced shopping list (i.e. product categories, including inventories, planning meals in advance and suggesting recipes for using left-overs), the
shopping list intervention is expected to contribute to a reduction of food waste. Considering the different purposes a shopping list can have that are discussed previously (e.g. memory aid, manage expenditures), these factors will also allow consumers to reduce their food waste. Moreover, by motivating consumers to better plan their purchases in the pre-action step of the motivational-volitional model of action phases (Schmidt, 2012) or in the similar phase 1 of Arnaud et al. (2015), food waste is expected to be reduced. Finally, by reinforcing consumers to stick to the shopping list in the action step of the model of Schmidt (2012) or in the similar phase 2 of Arnaud et al. (2015), the hypothesized effect will be positive.
2.3 Cooking enjoyment
15
Hartmann et al. (2013) have identified the importance of cooking skills and cooking enjoyment in relationship to making balanced food choices. Their articles concludes that cooking
enjoyment can help consumers to meet nutrition guidelines in their daily consumption and that it allows them to make healthier food choices. Candel (2001) has studied cooking enjoyment in the context of a consumers’ convenience orientation. The author found that convenience orientation is negatively related to cooking enjoyment, involvement with food products and variety-seeking, and to be positively related to role overload. Ternier (2010) also measured cooking enjoyment as a factor influencing convenience food purchases and consumption and found similar results. Moreover, cooking enjoyment has been studied in the field of children’s eating behaviors and eating enjoyment (van der Horst, 2012). Her study examined if increasing cooking enjoyment can help children to overcome their picky eating habits. The results showed significant direct effects between cooking enjoyment and picky eating. Woods (2004) has included differences in cooking enjoyment in the emotional territory for brands and came up with ‘Brandcepts’ to actively help consumers to make distinctions in the world of brands. In summary, cooking enjoyment has not yet been linked to food waste generation.
16
The specific opportunity to use the product passes and respondents are not making an effort in trying to find alternative recipes containing the same ingredient (Wansink et al. 2000).
A concept that is very much related to cooking enjoyment is ‘convenience orientation’. The term convenience includes the minimization of physical and mental effort in the preparation stage of food (Man & Fullerton, 1990). Convenience orientation towards meal preparation is considered as a relevant construct for understanding consumer behaviors towards food. This understanding is crucial when designing an intervention that aims to change these behaviors and reduce food waste generation. Convenience orientation can be Consumers who enjoy the
activity of meal preparation (or cooking) thus will be less convenience oriented towards this activity. In other words, because cooking enjoyment and convenience orientation are negatively related (Candel, 2001), consumers that experience high cooking enjoyment are less convenience oriented and vice versa.
Nowadays, cooking is more and more perceived as a recreational activity or a hobby and not solely as a necessary daily duty (Lang & Caraher, 2001). To come back to the concept of cooking enjoyment, consumers that experience high levels of cooking enjoyment are likely to cook for enjoyment purposes, they love cooking, and eating and food has an important role in the sense of entertainment (Buckley et al., 2007). In their article, de Boer et al. (2004) examined the influence of lifestyle characteristics on the demand for convenience food in Ireland. The authors compared consumers that often eat convenience foods with consumers that rarely eat convenience food based on their general food-related lifestyles, convenience food-related lifestyles and belief about convenience foods. They segmented consumers into tertiles based on this frequency of consuming convenience foods. The tertile that represents consumers who rarely eat convenience foods (i.e. the lowest tertile) is characterized as making more use of shopping lists, (…) plan more in advance, (…) and like to spend more time in the kitchen. Moreover, cooking enjoyment is identified by measuring someone’s enjoyment of trying new recipes (Brunso & Grunert, 1995; Labrecque et al., 2006). Here, the link is made between cooking enjoyment and the previously discussed concept of product abandonment and
17
Based on Wansink et al. (2000), consumers that experience high cooking enjoyment will buy more specific-use products compared to consumers with lower cooking enjoyment.
2.3.1. The role of cooking enjoyment
Based on previous literature, the proposition is made that consumers who experience high cooking enjoyment are more willing to try out new recipes compared to consumers who experience low cooking enjoyment (Brunso & Grunert, 1995; Labrecque et al., 2006). Trying out new recipes automatically entails purchasing grocery items for a specific use (i.e. the recipe) and based on previous literature it is demonstrated that these products have a high probability of ending up as a castaway (Wansink et al., 2000). It is hypothesized that, although a shopping list reduces the impulse buying tendency in general (Block & Morwitz, 1999), consumers with high cooking enjoyment will buy certain ingredients anyway, and thus the use of shopping list will have less effect. This leads to the formulation of the second hypothesis:
H2: Cooking enjoyment decreases the beneficial effect of using an enhanced shopping list on the reduction of food waste.
The interaction effect of cooking enjoyment on the relationship between the use of an enhanced shopping list and food waste reduction is based on the fact that for consumers with high cooking enjoyment a shopping list has less effect compared to consumers with low cooking enjoyment, because they will buy these ingredients anyway. In other words, these consumers are less convenience oriented and will purchase products for specific purposes (e.g. recipes) that are proven to be contributors to food waste (Wansink et al., 2000). Based on de Boer et al. (2004), these consumers are characterized as already planning more in advance and using shopping lists. Therefore, the direct positive effect of using a shopping list on the reduction of food waste generation doesn’t account for consumers with a high level of cooking enjoyment. This will be tested with the experiment that will be described in further detail in the methodology section.
2.4 Sale proneness
18
Together with price consciousness, sale proneness is likely to influence a consumers’ perception of the advertised reference prices and their behavioral intentions (Alford & Biswas, 2000).
Alford and Biswas (2002) examined sale proneness and price consciousness together with discount level and their connection with consumers’ outcome evaluations of offer value, search intention and purchase intention. The results of their study showed a relationship between someone’s level of sale proneness and their evaluation of all three outcome variables. The authors conclude that highly sale-prone consumers reported higher perceptions of value and higher buying intention. Garretson and Burton (2003) studied the differences between
consumers highly coupon- and sale prone and consumers less prone to sales promotions. The results demonstrated that highly prone consumers are drawn to reduced prices, but also like to spend time shopping and experience a sense of achievement when purchasing a product on special. The latter finding is also confirmed by Lichtenstein et al. (1993), who found sale prone consumers report that they feel like they are getting a good deal when they buy a brand that’s on sale. Another study by Burton et al. (1999) indicated that being exposed to an advertising sale flyer for retail supermarkets leads to a significant increase in the total number of advertised products that is purchased, the amount that is spent on these products, the number of coupons redeemed and the total number of products purchased.
In his study, Evans (2012) states that participant of his study reported to routinely buy more food than they needed. One of the reasons for buying too much food mentioned in this article were in-store marketing techniques. Promotions are an important element of the marketing mix in the field of retail management. They represent approximately 34 percent of the money that is spend on groceries and are widely used to attract consumers and increase market share (WRAP, 2014). They are designed to encourage consumers to take direct and unreflective decisions, e.g. impulse purchases, and seduce consumers to buy a product more quickly, more frequently and/or in greater quantities (Hawkes, 2009). There are different types of price promotions, i.e. multi-buy, BOGOF (buy one get one free), extra free, y for $x and temporary price reduction. In a previous study, one of the most frequently mentioned explanations for waste given by
19
Moreover, promotions stimulate increased quantity purchases (Blattberg & Neslin, 1989) and increase the consumption rate of different product categories (Ailawadi, 1998). In 2011, WRAP conducted a study investigating the possible impact of promotions on food waste generation. This study found a link between some types of promotions (multi-buy and y for $x) and the purchase of larger amounts of food. In the above sections it has already been demonstrated that over-purchasing has negative consequences for food waste reduction (Block & Morwitz, 1999).
Due to the above mentioned price promotions consumers feel they are put in a position where they have to choose between bulk (i.e. more value) that increases the likelihood of food being wasted, or smaller quantities (i.e. more expensive per quantity) that reduce the likelihood of food being wasted (Graham-Rowe et al., 2014). Because of the fact that discounts are often offered on products in large(r) quantities, these packages are prone to becoming a source of food waste due to too large quantities and consumers not being able to consume these products before the expiration date (Halloran et al., 2014). In line with these findings, Ashemann-Witzel et al. (2015) found what they call the “overarching consumer price orientation” as another factor that plays an important role during the shopping trip. This illustrates a consumers’ search for the optimal price/quality ratio and is thus related to price promotions. It explains why some consumers tend to over-purchase in reaction to pricing strategies of a store (e.g. volume discounts and price gradients).
2.4.1. The influence of sale proneness
20
Also, because these discounts are often offered on products in large(r) quantities, these packages are more likely to become a source of food waste (Halloran et al., 2014). This leads to the formulation of the third hypothesis:
H3: Sale proneness negatively affects the positive relationship between using an enhanced shopping list and food waste reduction.
It is argued here that sale-prone consumers, although they have the support of an enhanced shopping list to suppress impulsive buys, can’t resist the temptation caused by in-store
promotion materials and thus will be seduced to purchase certain products they do not need. In other words, the use of an enhanced shopping list reduces impulse buying, but an in-store bargain eliminates this decreased tendency to buy on impulse. Moreover, it is hypothesized that when consumers do stick to the products on their shopping list, when one of these items is on sale, sale-prone consumers will nevertheless buy too much of it, which is caused by the promotion.
2.5 Conceptual model
Based on previously discussed literature, a positive relationship is expected between the use of an enhanced shopping list and food waste reduction. In short, consumers will plan their
shopping trips in advance and will take into account the inventories they have at home, thereby preventing consumers from purchasing products they already bought in a previous shopping trip. Moreover, because it is argued that using a shopping list reduces impulse purchases (Baumeister, 2002; Block et al., 1999; Bell et al, 2011), which are identified as an important cause of food waste generation (Evans, 2012), the main effect is hypothesized to be positive.
21
Moreover, consumers with higher cooking enjoyment are already planning more in advance and are more likely to use shopping lists (de Boer et al., 2004). Therefore, the interaction effect that is included in the conceptual model is hypothesized to be negative.
The interaction effect of sale proneness on the hypothesized positive effect of using an enhanced shopping list on food waste reduction is based on the finding that highly sale-prone consumers experience higher buying intention (Alford & Biswas, 2002) and derive pleasure and a feeling of achievement when purchasing a product on special (Garretson & Burton, 2003; Lichtenstein et al., 1993). Therefore, these consumers are more likely to react to price promotions, which were found to be an important driver of food waste (Halloran et al., 2014). Moreover, price promotions distort the goals or intentions to plan and manage food purchases regarding needs and stocks (Borgne et al., 2015). This planning and managing food purchases is realized by means of the shopping list. However, this effect is offset by in-store price promotions that overrule the positive effect of the shopping list. In other words, the temptation that is caused by promotions can’t be resisted and the sale-prone consumers will still buy the promotion on impulse.
The three constructs that are reviewed based on previous literate are included in the conceptual model that will be tested in this study that can be found below in Figure 1.
22
3. METHODOLOGY
3.1 Research design
To be able to test the hypotheses that were developed in the theoretical framework, an
experimental field research study was set up. A major advantage of a field experiment that gives it preference above a laboratory setting is that the causal influence of variables can be
determined with more certainty (Aronson et al., 1998). Using this research design also ensures high external validity. Because individual food waste behaviors occur in people’s home
environment, a field experiment that records input inside the homes of participants best fits the goal of this research.
Because the research is undertaken to measure the effect of an enhanced shopping list
intervention, the independent variable has two levels: use vs. non-use. Therefore, the experiment will consist of an experimental group (i.e. the group that will use the enhanced shopping list intervention) and a control group (i.e. the group that will not use the enhanced shopping list intervention). Finally, the participants are randomly assigned to one of the two groups. This ensures that the result is caused by the intervention and not based on pre-existing differences between the participants (Aronson et al., 1998). When the sample size is sufficient enough, individual differences will be fairly distributed across the two groups and are therefore not able to influence the results of the experiment.
3.2 Procedure
The experiment took place in an Albert Heijn supermarket located in Eelde-Paterswolde in the north of the Netherlands. Participants were recruited on two consecutive Fridays (Friday, 11th
November 2016 and Friday, 18th November 2016) between 10am and 7pm. In order to recruit
23
This was done because otherwise it might have led to different responses of participants during the experiment and this could negatively bias the results of the study. The cover story for the shopping list group described that the researchers were interested in consumer’s experiences with certain tools developed by the Dutch Nutrition Foundation (e.g. the enhanced shopping list) and would like consumers to evaluate them and come up with suggestion to improve these tools. After visitors agreed to participate in the study, they were asked to read and sign the informed consent form. By signing this form, participants declared they were clearly informed about the purpose of the study and that they participated voluntary. Moreover, the form mentioned that participation was completely anonymous and that participants could drop out of the study at any moment in time without the need to provide a reason for this. Please refer to Appendix I for a copy of the informed consent form. After signing the form, participants wrote down their e-mail address and phone number to allow the researchers to notify them when they had won one of the gift cards. The first 70 participants were recruited for the control group and the second 70 participants were recruited for the shopping list group.
Respondents were asked to keep track of their food waste in a food waste diary for a period of 7 sequential days and fill in a supplementary questionnaire. The specific content of the
questionnaire will be discussed in more detail later in this section. The questionnaire could be completed online via the survey platform Qualtrics or filled out on paper and handed back in in the supermarket in a drop box. To ensure anonymity of the participants, each food waste diary and questionnaire was marked with a number. By doing so, participants could only be identified with this number. Moreover, the shopping list group also received an ‘enhanced shopping list’ developed by the Dutch Nutrition Foundation. This shopping list is comprised of two parts. On the left side of the list, consumers are reminded to manage their inventories of the products they have at home by the phrase “Check your inventories”. Moreover, the purpose of this part of the list is to plan meals in advance and write down for how many people this meal is meant to be prepared. On the right side of the list, consumers can write down which products they still need to purchase based on their inventories. This section is divided into different categories to provide more structure in the list: Fruit/vegetables, bread, meat/fish/vegetarian,
24
To motivate participant to hand in their materials, a little present was given to them when they handed in both the questionnaire and the food waste diary. Moreover, they would then
participate in a lottery where they had the chance to win one of five gift cards worth 50 Euros. Furthermore, when the response rate of the study was 75 percent and up, the researchers would also donate to the food charity.
3.3 Data collection
As mentioned in the previous section, the data collection was undertaken by means of two different elements: (1) a food waste diary and (2) a supplementary questionnaire that included question to obtain additional information about the other constructs of the study.
3.3.1 Food waste diary
To be able to collect food waste data from the participants, each participant was given a food waste diary (Appendix III). This approach has been used by previous researchers in other food waste studies (e.g. Koivopuro et al., 2012). On this diary, participants could keep track of their food waste for a period of seven consecutive days. Participants assigned to the control group were allowed to start filling in the diary from the first day. Participants assigned to the shopping list group had to wait filling in the diary after they had once used the enhanced shopping list, otherwise it could bias the results of the study.
The food waste diary consisted of two parts. On the upper section of the diary, participants needed to report their frequency of ordering take-away food and eating in a restaurant. Moreover, they needed to list their shopping frequency in terms of main shopping and top-up shopping. On the bottom section of the diary, participants could write down their food waste within one of the 24 food categories, or report it in the “others” category. For every grocery item that was thrown away, participants needed to report the amount of grams. For each category, an indication of the weight of each given product was provided, making it easier for participants to fill out the waste diary.
3.3.2 Questionnaire
25
This difference was because the shopping list group was also asked to come up with suggestions about the enhanced shopping list intervention.
The first part of the questionnaire consisted of questions that were used to obtain insights about the household’s socio-demographics, e.g. age, gender, employment status, income and
household size. Also, several questions were asked related to shopping and eating habits. Next, participants were asked to report their attitude toward statements about cooking enjoyment and sale proneness. Also, two question were included about identifying the main reasons for over-purchasing and food waste. Other questions that were included, for instance about health motivation, are not relevant for this study and will not be elaborated further.
3.4 Measurements
3.4.1 Food waste
The dependent variable ‘food waste’ that can be found in the conceptual model previously presented will be measured with the data collected through the food waste diaries.
3.4.2 Cooking enjoyment
The participant’s ‘cooking enjoyment’ is measured by means of asking them to evaluate their attitude towards different statements regarding cooking enjoyment on a 7-point Likert scale ranging from 1 (completely disagree) to 7 (completely agree). Examples of statements that were included are: “I like trying out new recipes”, “Cooking is my passion”. The scales are adopted from previous studies, which have already tested their validity (e.g. Candel, 2001; Hartmann et al., 2013). A total of 7 items was included to measure this construct. Higher scores on these items indicate greater cooking enjoyment
3.4.3 Sale proneness
26
3.4.4. Control variables
Several control variables were included in the study, being age, gender, education level, employment status, net income and household size. Taken from previous research, these
variables might possibly affect a household’s food waste generation (Evans, 2012; Lyndhurst et al., 2007; Parfitt et al., 2010) and should therefore be accounted for in the model to rule out their influence on the hypothesized relationships between the variables of the model.
4. RESULTS
4.1 Data4.1.1 Data preparation
The first step in analyzing the results of the food waste diary and supplementary questionnaire, the data needs to be prepared for further analysis. The paper questionnaires that were retrieved from the drop box in Albert Heijn were first filled in online and exported to allow for data analysis. For the two conditions, a total of 140 participants was recruited to participate in the study. However, only 33 packages of materials (i.e. food waste diary and questionnaire) were retrieved from the control group participants and 21 packages of materials from the shopping list group; Equivalent to a response rate of 47.14% for the control group and 30 percent for the shopping list group.
27
From the 21 responses retrieved from the shopping list group, only 13 food waste diaries were useful for further analysis; 8 participants filled in their daily consumption of food in the food waste diary and not their actual waste. Within these 13 responses, there were 6 missing values found for the variable ‘age’ and 1 missing value for the variable ‘household size’. These 7 missing values in total were replaced by the mean response to the variable.
Reviewing the data has left a total of 31 responses suitable for further data analysis. The drawbacks of this small sample size will be discussed in the limitations sections.
4.1.2 Sample descriptives
Overall, 21 women (67.7%) participated in the study and 10 men (32.2%). This difference might be because in general, women are the ones responsible for the household work (Bittman et al., 2003; Bianchi et al., 2000). The age of the respondents ranges from 25 up to 84 years and has an average of 56.32 years (SD = 12.724).
The largest part of the participants has an HBO degree (41.9%), followed by a university degree (32.3%). After this comes MBO (9.7%) together with other (9.7%), and finally HAVO (3.2%) together with MAVO/VMBO (3.2%). Moreover, most of the participants are employed (51.6%), followed by retired (16.1%). Next comes housemen/housewife (12.9%), and finally unemployed (9.7%) together with other (9.7%).
The monthly income of the study participants ranges from 1000 – 1499 Euros to 4000 Euros or more. There were no respondents that reported a monthly income below 1000 Euros. However, 4 participants (12.9%) indicated they would rather not disclose their income. Additionally, the household size of the respondents ranges from 1 to 5, with an average household size of 2.42 persons (SD = .992). Finally, the average food waste of a participant was 797.2581 grams (SD = 840.75744).
28
29
4.1.3 Random assignment
After reviewing the data and removing cases that weren’t useful, there were 31 respondents left to continue the analysis with. The distribution of the respondents over these two groups differs slightly: The control group has 18 respondents left and the shopping list group has 13
respondents left after the data editing.
However, it needs to be ensured that these participants are randomly assigned to one of the two groups in the experiment. This can be done by validated by comparing the means of the
different descriptives across the two groups. For the variable ‘age’, an ANOVA was conducted that showed no significant difference (p >.05) for this variable (p= .487). The other descriptives were tested with Pearson’s chi-squared test. This resulted into no significant difference between the two groups on the variables gender (p= .353), education (p= .328), employment (.056), household size ((p= .377) and monthly income (p= .905). This confirms the random assignment of the different respondents over the two groups of the experiment.
Table 2. Random assignment tests
4.2 Pre-analysis
4.2.1. Reliability analysis
30
This will determine whether the scale items measure the same construct and can be summed together, or if there are any uncorrelated underlying dimensions (Malhotra, 2010). Moreover, Cronbach’s Alpha is computed to assess the internal consistency of the different scale items for the two constructs (Malhotra, 2010). When these tests provide reliable results, the scale items are summed together and computed into two variables (i.e. the variables ‘cooking enjoyment’ and ‘sale proneness’).
First, the Kaiser-Mayer-Olkin (KMO) measure of sampling adequacy needs to be above .5 for a factor analysis to be appropriate. This value was .711 and thus enough. Second, Bartlett’s test of sphericity determines the correlation between the variables. When the significance level is below .05, a correlation is confirmed. This value was .000 and thus significant. Third, the communalities of the scale items need to be larger than .4 to ensure the factors explain a sufficient amount of variance. This was the case for all the 13 scale items. Finally, each value for the factor loading needs to be larger than .5. This critical value was also met for all the scale items.
When the scale items have passed these different tests to determine the appropriateness of factor analysis, it can be determined into how many factors the different scale items should be
combined. A principal component analysis was conducted with a Varimax rotation. An Eigenvalue (EV) of at least 1 is necessary in order for a factor to be considered.
Moreover, the final amount of factors that is chosen should explain at least 60% of the total variance explained (Malhotra, 2010). The total variance explained by the scale items was 60.267%.
Finally, the internal consistency of the different scale items is determined by computing Cronbach’s Alpha. This measure should have a value of at least .6 to allow for combining the scale items into a reliable variable (Malhotra, 2010). For the construct ‘cooking enjoyment’, Cronbach’s Alpha was computed to be .921 and for the construct ‘sale proneness’, Cronbach’s Alpha was computed to be .835.
31
Table 3. Factor analysis and reliability analysis
4.2.2 Normality test
The conclusions that are drawn from a regression analysis need to be validated. This means that first, the assumption of normality must be met (Field, 2009). This test is performed on the dependent variable ‘food waste’ and the interaction variables ‘cooking enjoyment’ and ‘sale proneness’.
Because the independent variable is a dummy variable (0 = control group or 1 = shopping list group), normality tests can’t be applied to this variable. The two tests that will be used to assess normality are the Kolmogorov-Smirnov test and Shapiro-Wilk test. These tests demonstrate a significant result (p < .05) for the dependent variable ‘food waste’, which means the distribution of this variable is significantly non-normal. This means the null hypothesis of a normal
32
these variables is insignificantly non-normal. This mean the null hypothesis of a normal distribution is accepted. An overview of the tests can be found in Table. 4.
Table 4. Normality test
4.2.3 Homogeneity of variances test
The second assumption that needs to be met in order to be allowed to perform a regression analysis, is that there must be homogeneity of variances between the control group and shopping list group. To test this, Levene’s test is performed on the variables (Field, 2009). For the
dependent variable ‘food waste’ and the moderating variables ‘cooking enjoyment’ and ‘sale proneness’, the result of the test is insignificant (p > .05). This means that the hypothesis that the variances are significantly different across the two groups is accepted. An overview of the test can be found in Table. 5.
33
4.3 Hypotheses testing
In this section, a multiple regression analysis will be conducted to test the hypotheses that were developed based on the literature review and the actual core of the research will be analyzed. The main effect (i.e. the effect of using an enhanced shopping list on food waste reduction) will be interpret. Next, the two interaction variables ‘cooking enjoyment’ and ‘sale proneness’ will be added in the regression analysis.
4.3.1. The effect of using an enhanced shopping list on food waste
To start with, the effect of using an enhanced shopping list for the planning of a shopping trip and applying it during the actual grocery shopping on the amount of food waste is analyzed by means of a multiple regression. A summary of the results can be found in Table. 6a.
The control variables, i.e. age, gender, education, employment, household size and in monthly net income, are also included in the multiple regression analysis. These variables can potentially impact the main effect of the study. This impact will be assessed with the results of the
regression. This showed that none of the VIF-scores of the control variables is above the critical value of 4. However the effect of the control variable ‘household size’ is significant. This can be explained by the fact that logically more people will generate more waste. The influence of the other control variables on the main effect is ruled out. In other words, the main effect will be explained by the independent variable and not partially by (some of the) control variables.
For the main effect, the model has an explained variance of R² = .390, which means that 39.0% of the variance is explained by the model. However, the model is insignificant when using an alpha level of 5%, but significant when using an alpha level of 10% (F = 2.104, p = .084). This entails a significant fit of the data overall when continuing with the 10% alpha level. However, the effect of the independent dummy variable control group or shopping list group is
34
Table 6a. Multiple regression - main effect
In order to try to find any significant results for the main effect, the variable ‘food waste’ is further divided into three categories: the total amount of food waste, the liquid food waste and the solid food waste. An ANOVA was conducted to assess the overall fit of the model when this categorization is made. Unfortunately, the two additional models turned out to be insignificant. Therefore, making a distinction between liquid and solid food waste doesn’t change anything with regards to rejecting the first hypothesis. Moreover, it was attempted to put in the control variables one by one to find significant results, but this wasn’t helpful either. The results can be found in Table 6.b
35
4.3.2 The interaction effect of cooking enjoyment
Besides the main effect of the study, i.e. the effect of using an enhanced shopping list on the amount of food waste, it was hypothesized that cooking enjoyment decreases the effect of this relationship. This effect is analyzed by means of a multiple regression analysis. Again, the dummy variable control group vs. shopping list group is the independent variable and the amount of food waste is the dependent variable. Also, the variable cooking enjoyment is included as an independent variable, to assess its direct effect on the amount of food waste. To test the interaction effect of cooking enjoyment between the dummy variable and the dependent variable, it is also included as a moderator. The moderating variable is computed by multiplying the independent dummy variable with cooking enjoyment (i.e. dummy variable control group or shopping list * cooking enjoyment).
With regards to the control variables, the VIF scores indicate no values above the critical value of 4. Continuing, only one of the control variables demonstrate a significant influence, which is again household size. The overall model explains 43.4% of the variance (R² = .434). The model is significant when applying an alpha level of 10% (F = 2.109, p = .079).
However, the interaction effect of the variable ‘cooking enjoyment’ has no significant result (p = .206). Hypothesis 2, cooking enjoyment minimizes the beneficial effect of using an
enhanced shopping list on the reduction of food waste, should therefore be rejected. There is also no significant direct effect of cooking enjoyment on the amount of food waste (p = .272).
36
Table 7a. Multiple regression – Interaction effect cooking enjoyment
37
Table 7b. Multiple regression – interaction effect cooking enjoyment additional tests
4.3.3 The interaction effect of sale proneness
For the second moderator, it was hypothesized that sale proneness has a weakening moderating effect on this relationship. This effect is analyzed by means of a multiple regression analysis. Again, the dummy variable control group vs. shopping list group is the independent variable and the amount of food waste is the dependent variable. Also, the variable sale proneness is included as an independent variable, to assess its direct effect on the amount of food waste. To test the interaction effect of sale proneness between the dummy variable and the dependent variable, it is also included as a moderator. The moderating variable is computed by multiplying the independent dummy variable with sale proneness (dummy variable control group or shopping list group * sale proneness).
38
Table 8a. Multiple regression – Interaction effect sale proneness
Next, in an attempt to find significant results, food waste was again divided into the three categories total food waste, liquid food waste and solid food waste. Unfortunately, this didn’t change anything to the overall fit of the model, i.e. the ANOVA. This statistical test remained insignificant. It was also tried to compute the z-scores of the sale proneness variable and use these scores when running the linear regression. Then, the interaction variable became the independent dummy variable multiplied with the z-scores of the sale proneness variable
(dummy variable control group or shopping list group * z-scores sale proneness. However, this also didn’t result into any significant results for the second hypothesis. Moreover, it was attempted to put in the control variables one by one to find significant results, but this wasn’t helpful either. The results can be found in Table 8.b.
39
5. DISCUSSION
Because the amount of food waste generated by households has increased dramatically, its environmental and economic consequences has increased public awareness about the topic. It not only has negative consequences like monetary losses, but also affects limited natural resources. Consequently, multiple questions raise about the determinants of food waste and possible solutions to help overcome this food waste issue. Nonetheless, in the literature to date, not much attention has been paid to food waste generation and there are few academic studies conducted. So far, there are just a couple of insights about the underlying factors of consumer’s food waste behaviors.
The current study has used a field experiment that made use of a food waste diary and
supplementary questionnaire to examine the effectiveness of using an enhanced shopping list to reduce the amount of food waste generated by households. Additionally, the potential
moderating effect of the two constructs cooking enjoyment and sale proneness is investigated by means of a multiple regression analysis. The results were meant to answer important questions about the precedents of food waste generation.
With the conceptual model as their foundation, three hypotheses were developed and tested. However, these conceptualized hypotheses could not be demonstrated with this study. The hypotheses, their expected relationships and the study outcomes can be found in Table 10.
Expected Result H1: Using an enhanced shopping list during grocery shopping
positively contributed to food waste reduction +
Not supported H2: Cooking enjoyment minimizes the beneficial effect of using an
enhanced shopping list on the reduction of food waste -
Not supported H3: Sale proneness negatively affects the positive relationship
between using an enhanced shopping list and food waste reduction -
Not supported
40
Previous research has already identified some causes for the complex behaviors that cause food waste generation. However, to this date, cooking enjoyment and sale proneness has not yet been linked to food waste. The literature that was discussed in the literature review section suggested that better planning before going on a shopping trip and developing and using a shopping list during grocery shopping would results into a reduction of the amount of food waste. The failure to follow a shopping list was identified as a reason for household food waste to occur (Evans, 2012; Lyndhurst et al., 2007; Graham-Rowe et al., 2014). Specifically, over-purchasing was found to be a large contributor to food waste (Evans, 2012). Therefore, the first hypothesis states that using an enhanced shopping list positively contributes to food waste reduction However, the empirical results of this study do not support this notion.
Regarding cooking enjoyment, it was implied that products bought for specific purposes (e.g. a recipe) would have a high probability of ending up in the trash bin (Wansink, 2000). Because consumers who experience high cooking enjoyment are more willing to try out new recipes (Brunso & Grunert, 1995; Labrecque et al., 2006) and products bought for specific purposes (e.g. a recipe) have a high probability of ending up in the trash bin (Wansink, 2000), it was hypothesized that cooking enjoyment minimizes the beneficial effect of using an enhanced shopping list on the reduction of food waste. However, this hypothesis could not be confirmed with the results of this study, contrary to the findings in the literature. A possible explanation for this could be that consumers with high cooking enjoyment find ways to use left-overs by trying out new recipes and thus throw away less. This notion deserves more attention in further research.
41
However, one thing that can be concluded from the food waste diaries is that the average food waste per person was higher for the control group than for the shopping list group; control group respondents reported on average 839.72 grams of food waste and shopping list group
respondents reported on average 738.46 grams of food waste. This demonstrates that using a shopping list has the potential to reduce food waste generation and that this topic needs to be investigated in more detail to identify the consumer behaviors that cause food waste.
5.1 Limitations
Unfortunately, the author of this thesis was not able to confirm any of the hypothesized concepts due to insignificant statistical results. A possible explanation for this fact could be the limited sample size (N=29) that was used to generate the results. This sample size is not representative for a large population. Therefore, the results can’t be generalized. Moreover, regarding the age of the respondents, there were no younger participants in the study and the elderly participants are overrepresented. Also, more females than males participated in the study. This also leads to low generalizability of the data. To be able to confirm the hypotheses, an additional study is necessary that will need a larger and more representative sample.
Another major limitation of the study was the food waste diary. Because many respondents filled in their consumption instead of their food waste, a lot of cases weren’t useful for the data analysis. The researchers included a form with a detailed description of how to fill in the waste diary, but perhaps not all of the participants read this form carefully. In the future, this needs to be explained better when recruiting respondents for the study.
Thirdly, the study period might not be sufficient enough to capture the whole cycle of planning, buying, consumption and disposing. By the way the study was organized now, food waste of previous purchases might be included in the study period and/or food waste that occurred after the study period is not included in the food waste diary.
42
43
REFERENCES
Ailawadi, K.L. and Neslin S.A. (1998), “The Effect of Promotion on Consumption: Buying More and Consuming It Faster”, Journal of Marketing Research, 35(3), pp.390–398.
Alford, B. L. and Biswas, A. The effects of discount level, price consciousness and sale proneness on consumers' price perception and behavioral intention. Journal of Business Research. Sep2002, Vol. 55 Issue 9, p775-783
Arnaud, A., Kollmann, A., & Berndt, A. (2015). Generation Y: The Development and Use of Shopping Lists. Advances in Social Sciences Research Journal, 2(9) 1B13.
Aschemann-Witzel, J., de Hooge, I., Amani, P., Bech-Larsen, T. and Oostindjer, M. (2015) ‘Consumer-related food waste: Causes and potential for action’, Sustainability, 7(6), pp. 6457–6477. doi: 10.3390/su7066457.
Baumeister, R.F. (2002) ‘Yielding to temptation: Self‐Control failure, impulsive purchasing, and consumer behavior’, Journal of Consumer Research, 28(4), pp. 670–676. doi: 10.1086/338209.
Bell, D. R., Corsten, D., & Knox, G. (2011). From point of purchase to path to purchase: How preshopping factors drive unplanned buying. Journal of Marketing, 75(1), 31–45.
Bianchi, S.M., Milkie, M.A., Sayer, L.C. and Robinson, J.P. (2000). Is anyone doing the housework? Trends in the gender division of household labor. Social Forces, 79 (1), pp. 191-228.
Bittman, M., England, P., Folbre, N., Sayer, L. and Matheson, G. (2003) When does gender trump money? Bargaining and time in household work. The American Journal of Sociology, Jul 2003, 109, pp. 186.
Blattberg R.C. and Neslin S.A. (1989), “Sales promotion: The long and the short of it”, Marketing Letters, 1(1), pp.81–97.
Block, L.G., Morwitz, V.G., 1999. Shopping lists as an external memory aid for grocery shopping: influences on list writing and list fulfilment. Journal of Consumer Psychology 8 (4), 343–375.
de Boer, M., McCarthy, M., Cowan, C. and Ryan, I., 2004. The influence of lifestyle characteristics and beliefs about convenience food on the demand for convenience foods in the Irish market. Food quality and preference, 15(2), pp.155-165.
Borgne, G.L., Sirieix, L. and Costa-Migeon, S. (2015) Food waste and promotions. Available at: https://hal.archives-ouvertes.fr/hal-01140919/document (Accessed: 10 October 2016).
Brunso, K., & Grunert, K. G. (1995). Development and testing of a cross-culturally valid instrument. Food-related life style. Advances in Consumer Research, 22, 475–480.
Buckley, M., Cowan, C. and McCarthy, M., 2007. The convenience food market in Great Britain: Convenience food lifestyle (CFL) segments.
Appetite, 49(3), pp.600-617.
Burton, S., Lichtenstein, D.R. and Netemeyer, R.G., 1999. Exposure to sales flyers and increased purchases in retail supermarkets. Journal of
Advertising Research, 39(5), pp.7-15.
Candel, M. J. J. M. (2001). Consumers’ convenience orientation towards meal preparation: conceptualization and measurement. Appetite, 36, 15–28.
