By
Seedi Salman
University Of Groningen
MSc Marketing Management
June 23, 2014
First Supervisor: Dr. J.C. Hoekstra
Second Supervisor: Dr. J.W. Bolderdijk
Schoolhom 30
9711 JH Groningen
(06) 15942216
Seedi89@hotmail.com
Student Number s 1739220
Which Factors Influence The Giving Away of
Food and How Does Giving Food Affect
1
Abstract
This paper explores the concept of giving away food. Where the first part of the conceptual model tests the relationship of the drivers of giving food – Mindful Consumption, Pro-social Behavior, and Convenience- on giving food, the second part of the model relates to the field of happiness and tests the effect of giving food on happiness. The environmental values –Biospheric, Altruistic, Egoistic, and Hedonic- are included as moderators determining the effect they have on the relationship between the drivers of giving food on giving food (Part 1), as well as giving food on happiness (Part 2).
Results indicate that part 1 of the model is insignificant where no relationship is found between the drivers of giving food on giving food. Results for part 2 however illustrates that there is some significance. Although giving food has no direct effect on happiness, when values are included as moderators, there is a direct effect as well as a moderating effect due to the values. People become less happy when giving food to others while having more Biospheric and Altruistic values, and become happier when having Egoistic values.
Keywords: Mindful Consumption, Pro-social Behavior, Giving Food, Happiness,
2
Contents
1. Introduction ... 3
2. A Theory towards the Drivers of Giving Food, and the Effects on Happiness ... 7
2.1. Conceptual Model ... 7
2.2. Drivers to Giving Food ... 9
2.2.1. Giving Food. ... 9
2.2.2. Mindful Consumption. ... 10
2.2.3. Pro-social Behavior. ... 11
2.2.4. Convenience ... 12
2.2.5. The Moderating Effects of Values – Part 1. ... 14
2.2.6. Giving Trait. ... 19
2.3. Giving Food to Others and the Effect on Happiness ... 20
2.3.1. Happiness ... 20
2.3.2 The importance of gender. ... 21
2.3.3 The Moderating Effects of Values- Part 2 ... 22
3. Methodology ... 24
3.1 Data Collection, Procedure and Sample ... 24
3.2. Development of the Survey ... 26
3.3. Reliability and Validity ... 31
3.4 Method of Analysis ... 36
3.4.1. Econometric Model and the Main Effects for Part 1 ... 36
3.4.2. Econometric model and Moderating Effects for Part 1 ... 37
3.4.3. Econometric Model and the Main Effects for Part 2 ... 38
3.4.4. Econometric model and Moderating Effects for Part 2 ... 38
4. Results ... 39
5. Conclusion ... 41
6. Discussion ... 45
6.1 Discussion of Part 1 in the Conceptual Model ... 45
6.2 Discussion of Part 2 in the Conceptual Model ... 51
7. Managerial Implications ... 55
8. Limitations and Directions for Future Research ... 57
9. Sources: ... 59
3
1. Introduction
Large quantities of food along the distribution chain and by household consumers are being wasted every day (e.g. Gustavsson et al., 2011; Parfitt et al., 2010; Griffin et al., 2009). This has presented significant issues as wasting food presents the waste of limited resources leading to unnecessary impacts on the natural environment. Further, the discarding of food can have indirect impacts as large amounts of energy resources and materials are required to produce and package such goods. This fact contains issues that go beyond that of the environment as food waste represents ethical issues as a large part of the world is starving, and cannot afford such needs to survive.
Indications suggest that consumers and food institutions in Europe and the USA waste food which amounts to, and varies between 15% and 30% of all the purchased food (Kantor et al., 1997; Ventour, 2008; Quested and Johnson, 2009). What makes the matter even worse is the fact that most of the food being wasted is avoidable and can amount to about 4 kg per person per week in the UK, Norway and Sweden (Quested and Johnson, 2009; Fredriksen et al., 2010; SEPA, 2005). Reasons for this waste are diverse. Consumers can waste food due to a lack of plan, change of plan, buying too much food and not wanting to eat leftovers, or do not know what to do with the excess food they have (Cox and Downing, 2007).
4 enables other people to benefit from their food and provide nourishment for other people who might not be able to afford it. The question is why the majority of people do not give redundant food to others either. Drawing on literature in search of giving, little to no research has been found concerning the giving away of food. Most literature has focused on pro-social spending, donating money and blood (e.g. Brownell, et al., 2009; Zahn-Waxler et al., 1992; Aknin, et al., 2011; Borgonovi, 2008). This, in some aspect is relevant as this act has a focus on pro-social behavior which focuses on helping others. Pro-social behavior is defined as an act performed to benefit another person (Penner, Dovidio, Piliavin, and Schroeder, 2005). Giving food however, in addition, may have another dimension. This dimension is related to environmental concerns which is not covered by literature on pro-social spending.
5
This paper will not only investigate the giving of food to others, but also how giving food affects the giver’s happiness. This allows us to provide valuable insights into the human psychology as this might as well present new ways of people giving away products. Furthermore, due to the limited information available concerning the giving of food to others, it is unclear what the effects of such actions could have on a person’s wellbeing or happiness. This presents a different form of giving in relation to donating blood and money, which can encourage different levels of happiness than what is presented in literatures related to pro-social spending and conserving the environment.
Literature focused on happiness has shown that interventions such as pro-social spending can lead to sustainable increases in subjective well-being, and have found that there is a positive feedback loop between pro-social spending and happiness (Aknin et al., 2012). Although happiness levels are relatively stable across individuals, there are strategies which show to increase an individual’s happiness. Research has shown that spending money on others leads to an increase in happiness in relation to spending on ones’ self (Dunn et al. 2008). Further, Aknin et al. (2013) illustrates that people who chose to give a goody bag to someone who they did not know, reported higher levels of happiness which led to emotional rewards even when the receiver was unknown to the spender. This might have the same consequences on happiness when giving food to others.
As it is proven by research that pro-social spending, giving and sharing increase happiness, it has not confirmed whether giving food to others leads to happiness, or whether people who give food to others are happy.
6 while eating food and the amount of food being eaten by people while in a certain mood, no research has been conducted determining the effect of giving food on happiness. Theories such as mindful consumption, pro-social and environmental behavior exist but do not explain the effects of such factors on giving food. Giving food is related to both pro-environmental as well as pro-social behavior. Most research studied those behaviors in isolation where they did not consider the effects of such behavior on giving jointly. This study tests the hypothesis and will provide explanatory power related to those aspects to gain valuable insights in a relatively new domain. The main aim of this research is to determine the drivers of giving food to others in the field of giving, pro-social behavior, and mindful consumption, and build on literature in the field of happiness of how giving food affects a person’s happiness. This leads to the following question being presented:
Which factors influence the giving away of food, and how does giving food affect happiness?
In order to obtain insights to determine the effects of the drivers of giving away food on giving food, and the effects of giving away food on a person’s happiness, this research analyses the data collected from 53 respondents who have participated in filling in an online survey, which was sent through a food sharing platform. Furthermore, the data collected is analyzed by using the SPSS statistics program and where regression analysis have been conducted to determine the direct effects of the independent variables on the dependent variables, as well as the moderating effects of the moderators between the dependent and independent variables.
7 analyses the results obtained from the analysis. This is followed by Chapter 5 which provides the conclusions of the analysis and where Chapter 6 is based on the discussions of the obtained results. Next, Chapter 7 applies the findings to real life situations and its implications to managers and lastly, Chapter 8 provides the limitations of this study and directions for future research.
2. A Theory towards the Drivers of Giving Food, and the Effects on
Happiness
2.1. Conceptual Model
8 discard the food, the trash bin might be located further away from the people’s home, or might be due to a toll needed to be paid imposed by the government in order to throw away food and reduce waste (Schultz, 1999; Guagnano, Gregory, Stern and Thomas, 1995; Aadland and Caplan, 2003). These factors must be taken into account and are thus included in the conceptual model as factors which would directly influence giving food to others. Furthermore, a factor which might have an effect on how much food is given to others, as well as how happy an individual is, might be due to the importance they place on their values. The manner in which a person experiences happiness, as well as the set of factors which determines a person’s behavior is related to their values. They affect various beliefs, preferences and behavior simultaneously (Rohan, 2000; Rokeach, 1973) and are stable over time (Stern, 2000). However, different people prioritize among the different values and thus every person behaves differently. It plays an important role in how people view the world, the way they behave and thus the way they feel towards their actions. Steg et al. (2014) illustrate four values which are relevant to environmental beliefs, and they affect how a person acts and feels depending on the importance they place on those values. For this reason, they are illustrated as moderators in the conceptual model as they may have an effect between the relation of the drivers of giving food to giving food, and the relation between giving food and the effect on happiness.
Further, in literature related to happiness, evidence suggests the men and women pursue happiness in different manners. The article by Tkach and Lyubomirsky (2006) has studied this aspect and demonstrate which strategies of pursuing happiness can have an effect on happiness, depending on whether the person is a male or female. As giving food might be a strategy which suits one related more for women in relation to men or vice versa, it might have a differing effect on happiness. Due to this, gender might have an impact and influence the relationship between giving food and a person’s happiness, and is thus included as a moderator in the conceptual model.
9 This is illustrated in the conceptual model as a giving trait and is included as a control variable which is held constant, to test the relative impact of the drivers of giving food on giving food. Figure 1 below illustrates the conceptual model which provides an overview of the variables and the hypothesis which are discussed in the following sections.
Figure 1: Conceptual Model
The following sections provide the theoretical background and literature review concerning the variables, the reasons why they should be considered in the conceptual model, followed by the Hypothesis to be tested and the effect that each variables has on the dependent variables.
2.2. Drivers to Giving Food
2.2.1. Giving Food
As mentioned earlier, there is a scarce amount of research conducted which studies the giving away of food. As most research related to food elaborates on sharing food in
10 hunter- gatherer societies (e.g. Bird-David, 1990; Testart, 1982), the recycling of food to reduce food waste (e.g. Fehr et. al., 2002; Refsgaard et al., 2009) and the emotions which one experiences while eating food (Canetti et al., 2002 and Barthomeuf et al., 2009), research does not explain the giving away of food as well as the reasons for doing so. Furthermore, research related to giving mostly concerns the giving of gifts to others (e.g. Robben et al., 1994; Areni et al., 1998; Andreoni, 2007), blood donations or nature of the like. For this reason, it is important to explain what is meant by the giving away of food before we can proceed in explaining the reasons for doing so.
What is meant by giving away food in this context is the giving away of food to others for free, meaning that the person receiving the food does not have to pay for the food being given to them. Also, the food being given away refers to food which is edible, and which is not yet expired. It does not take into account spoiled food which is needed to be thrown away. This can be in the form of packaged goods which is still unpacked or not used, close to the expiration date, as well as food in the form of leftovers which is already prepared and where people do not want to dispose of it.
2.2.2. Mindful Consumption
11 defined as the processes through which materials were previously used are collected, processed, remanufactured and reused (Ruiz, 1993), and that the characteristics associated with recycling include attitudes concerning the environment as well as demographic variables and personality constructs. MC holds that people act in a manner which benefits society and the environment.
Seth, Sethia, and Srinivas’s (2011) theory on mindful consumption has been provided to explain environmental and social behaviors related to consumption. People act and consume in a manner in which the amount being consumed is controlled for and not taken for granted, keeping in mind their consumption patterns and how it would affect themselves as well as other people, and the environment. People would thus act in a manner to reduce food waste as well as benefit other people. Giving food to others instead of throwing it away would benefit both the environment as less waste is produced, as well as benefiting other people where nourishment is provided to them. It is postulated that mindful consumption would have a positive effect on giving food and leads to more food being given to others.
H1: Mindful Consumption has a positive effect on giving food.
2.2.3. Pro-social Behavior
12 Following the reasoning that is described which links giving money and donations to pro-social behavior (e.g. Brownell, Svetlova, and Nichols, 2009; Zahn-Waxler, Radke-Yarrow, Wagner, and Chapman, 1992; Ankin, Dunn and Norton, 2011), setting goals to reduce the waste of food could lead to behaviors which results in giving food to others, people give food because they want to help other people. They would help others without any rewards. By rewards it is meant to expect something in return as extrinsic rewards. Aside from this however, a person can experience intrinsic rewards which could motivates people to give more food to others (Ankin, Dunn and Norton, 2011). The social perspective is one of the intrinsic aspects which reflects the understanding of emotional states or the thinking of other people and is believed to be a cognitive skill for experiencing empathy for others (Eisenberg and Miller, 1987). As illustrated in the work by Ankin et al. (2011), spending on others increases subjective well-being and have found that there is a positive feedback loop between pro-social spending and happiness. Further, research has shown that spending money on others leads to an increase in level of happiness in relation to spending on one’s self (Dunn et al., 2008). As this paper does not study the effects of pro-social behavior and happiness, it is stated here to demonstrate that a feedback loop exists, that people would give because of pro-social behavior, and that it can be that people regularly give away food because of pro-social behavior. Giving food to others, as in the case of pro-social spending, is also a form or pro-social behavior. And as people act pro-socially to benefit and care for others, people are expected to give more. This leads to the following hypothesis being presented:
H2: Pro-social behavior has a positive effect on giving food to others.
2.2.4. Convenience
13 home which presents issues for people as they are not willing to spend time and effort. Further, governments also require that people separate their trash or waste in categories to promote the recycling of materials. However, people often have difficulties in separating edible from spoiled food as well as separating paper from glass, and other materials due the government not providing the facilities nearby which de-motivates people from doing so (e.g. Jahre, 1995; Martin, Williams and Clark; 2006; Gonzalez- Torre, Pilar and Adenso-Diaz, 2005). This issue might cause people to neglect using such kinds of alternatives to recycle food and can be a reason why people give food such as leftovers to others. This might motivate people to establish or follow goals related to reducing waste which curbside recycling constricts due to the factors stated above.
By giving food to others, people would simply have to package the food and hand them over to another person. They might give the food to a person who they might be visiting, or a person may come to the giver’s house and pick up the food. In doing so, people would give food to others which otherwise would be discarded of and which would benefit the receiver. Furthermore, people would not have to separate edible food from the spoiled food as they would only give away food which is not yet spoiled. This further reduces waste as the food is given to another person which otherwise would not be eaten, left to rot and which would then be discarded of, or throw the food simply due to not wanting to eat the same food again as leftovers.
We propose that giving food to others can solve for conveniency issues. It can be seen as convenient since little time, effort and no money is needed to give to others, or who are willing to pick up leftovers or packaged food which people are willing to give away. There is little to no effort needed for the provider as they are in the comfort of their homes and wait until the receiver can pick up the food, or no extra effort needed if the giver would give the food to a person who they might be visiting anyways. Due to this convenience, people give away food and benefit others as well as reduce their household waste in a convenient manner.
14 knows they will not consume, and otherwise be thrown away which in this case, other people benefit from. We therefore postulate that giving food to others is convenient and thus leads to people giving more food to others.
H3: The more convenient the giving of food is, the more food is given to others.
2.2.5. The Moderating Effects of Values – Part 1
15
certain situation is evaluated, and what the different alternatives are considered (Steg et al., 2014). Additionally, values affect various beliefs, preferences and behavior simultaneously (Rohan, 2000; Rokeach, 1973) and are stable over time (Stern, 2000). Values play an important role in how people view the world, the way they behave and the way they feel.
17
H4: A) Biospheric and altruistic values enhance the effect of mindful consumption on giving away food
B) Biospheric and altruistic values enhance the effect of pro-social behavior on giving away food
C) Biospheric and altruistic values enhance the effect of convenience on giving away food
In contrast to biospheric and altruistic values, Hedonic and Egoistic values are considered self enhancement values (Schwartz, 1992). The Egoistic value is concerned with one’s own self-interest. Those individuals focus more on their personal costs and their own benefits when making choices or decisions and reflect the costs that affect a person’s own resources such as money and power and act pro-environmentally when the perceived individual benefits of those actions exceed the costs of doing so, and vice versa. It is further stated by Steg et al. (2014) that egoistic and hedonic values are negatively correlated with pro-environmental beliefs, attitudes, preferences and behaviors (e.g. De Groot and Steg, 2008, 2010; Steg et al., 2005, 2011) which implies that people care more about personal gains than they are about the environment. In this line of thought, although people give food because of reasons such as pro-social behavior, mindful consumption or because it is more convenient, people who have egoistic and hedonic values would give less food to others.
18
19
food to others when its expiry date is almost due, in relation to when it is sufficiently before due, or they might throw away the food only when the food is expired and is rotting, as the actions of giving away food or throwing it away may come at a cost of comfort, time or money. Therefore, it is postulated that:
H5: A) Egoistic and hedonic values mitigate the effect of mindful consumption on giving away food to others.
B) Egoistic and hedonic values mitigate the effect of pro-social behavior on giving away food to others.
C) Egoistic and hedonic values mitigate the effect of convenience on giving away food to others
2.2.6. Giving Trait
20 independent variables (Mindful consumption, Pro-social Behavior and Convenience) may have a positive direct effect on giving food to others, we expect that the effect on giving food by the giving trait is constant. In other words, the independent variables and the moderators will have a greater effect on happiness than the Giving Trait which is the control variable in the model. This is a benchmark, and rests the notion that a person shares food because they are naturally a generous person. We will call this the Giving
trait as illustrated in the conceptual model in Figure 1 above.
2.3. Giving Food to Others and the Effect on Happiness
2.3.1. Happiness
21 or in other words, had the choice to buy a gift for someone, had a greater level of happiness in relation to the individuals who were assigned to buy a gift for them-selves. In this vein, giving food to others promotes and leads to happiness. People give away food to achieve goals whether it might be to help others or help in protecting the environment, which could lead to a sense of accomplishment and happiness.
Further, Aknin et al. (2013) study the effects of pro-social spending and the effect on happiness. They state that humans derive emotional benefits using their financial resources to help others and that in poor as well as in rich countries, pro-social spending lead to greater happiness. Children as young as two years old showed to exhibit increased happiness when a valued resource is given away (Aknin et al., 2012). They also mention that pro-social spending only reflects one form of generous behavior (Liu and Aaker, 2008) and it is possible that other kinds of behavior such as caring for the community and performing random acts of kindness may also promote the well-being of an individual. The act of giving food to others can also lead to such trends as they help others and create positive experiences increasing happiness. People continue to enjoy past pleasant experiences and may have a lasting impact on a person’s life (Frederick and Loewenstein, 1999; Van Boven and Gilovich, 2003; Frank, 2004, Scitovsky, 1976). We hypothesize that giving food to others would create happiness. Giving food to others may be a form of experiential purchases, meaning that people give away their food because it gives them internal satisfaction or happiness, and also, the act of giving may result in outcomes with positive experiences such as helping others. Nicolao et al. (2009) found that positive experiences tend to produce both more positive and less negative purchases than material purchases. The following hypothesis thus states that:
H6: Giving food to others is positively related to happiness.
2.3.2 The importance of gender.
22 this aspect and state that emotional benefits that a woman gains through affiliation (characterized by terms as helping others and communicating with friends) may be undercut by their greater tendency to ruminate about causes and/or consequences of their unhappiness (Nolen- Hoeksma, 1991). In contrast, men manage bad moods more through pleasurable and distracting activities such as sports (Thayer et al., 1994). It is thus likely that men and women have different strategies in which they prefer to increase and/or maintain their happiness. It is also found that males and females reported similar levels of happiness but they have confirmed that there are significant differences in the uses of happiness increasing strategies (Tkach and Lyubomirsky, 2006). Whereas women reported more social affiliation than men, men reported more frequently on active leisure such as going out, exercising and working on hobbies (Tkach and Lyubomirsky, 2006). It is proposed that men and women can experience different levels of happiness. When giving food to others, this might be related to a strategy which benefits and relates more to women than men as social affiliation fulfills a strategy which is characterized more for women in relation to men. Although both genders experience happiness when giving food, the amount of happiness might differ. This would provide valuable insights as it could have consequences and provide different strategies and programs to motivate men and women in giving food to others.
It is not to state that men do not experience happiness in conducting the same activities as well, however they may not experience as great a happiness as women do due to giving being more of a strategy which is more associated with social affiliation, and thus being a strategy more suited to women. The following hypothesis thus states that:
H7: Giving food leads to more happiness for women then for men.
2.3.3. The Moderating Effects of Values- Part 2
23 food fulfills the value whether it is to help people or reduce waste, which influences them to give more food to others due to a sense of joy or pride, or a sense of duty to do so.
Concerning the relationship between giving food and happiness, as giving food might be related to the recycling of food, it has an environmental aspect related to the reducing of waste which helps preserve the environment. People may become happier when they give food to others due to that act, and at the same time fulfilling their biospheric values where they have contributed to reduce food waste. Further, the act of giving food to another person can have an impact on happiness due to altruistic values which illustrates a concern of others who are not able to afford food, or have little resources which could hinder their purchases of food, where people might feel happier when they give food to others because they have the rewarding feeling that they have helped another person in the process of giving away food. In other words, giving food can be related to the helping of others as well, which is a concern for other’s well-being. This aspect is more related to the altruistic aspect. Being happy due to giving food to others would fulfill both values, whether it would be to help others, or to help preserve the environment. As the affect may be stronger and affects the relationship between the dependent variable and independent variable, this is included as a moderator to determine the effect of those values between the relationship of giving food on happiness. The following hypothesis states:
H8: Altruistic values and biospheric values enhance the effect of giving food on happiness.
Further, following the reasoning concerning the egoistic and hedonic behavior explained in section 2.1.5, egoistic and hedonic values can also affect a person’s happiness resulting from giving food to others as it is associated with financial and time costs. Hedonic values may be related to environmental beliefs, preferences and actions, and when in doing so, requires effort or comfort costs. A person may throw away their food even if there are leftovers because it requires more effort to package it and put it in the
24 others, the packaging of leftovers, putting it in the refrigerator, and the intent to share it or give it to others may be effortful and may also reduce comfort due to the processes of achieving this. As giving food requires effort, people with egoistic and hedonic values may be less happy when giving food to others as this might have presented a tradeoff between giving food and comfort or effort. The values represents one which has an effect on a person’s happiness when giving food to others as they associate such actions as effortful and which might decrease their comfort due to the process of going through the stages of giving food to others. For the reasons stated, hedonic and egoistic values can have implications on giving food and happiness and may affect the relationship between them. People may be less happy when giving food to others because in doing so, it creates effort and is not convenient at that time, and may come at a cost. In order to determine this, we have included those values in our conceptual model as a moderator.
H9: Egoistic and Hedonic Values mitigate the effect of giving food to others on happiness.
As limited research has been conducted relating the factors that lead to giving food and the effects on happiness, the theory will provide new insights and valuable information contributing to literatures of giving, food, waste and happiness among others. Due to the fact that this area of research is relatively new and many variables which have not been discussed, and which can affect the factors of giving food and happiness, the theory will give important insights and provide a benchmark and starting point in which future research can be built on.
3. Methodology
3.1 Data Collection, Procedure and Sample
25 items online which they do not need or would not consume. Further, people who are interested in the posted items are able to claim the items and collect them from the address which the giver provides. It is a website which encourages people to give food to others to reduce food waste. This presents a great opportunity for us to gain insights and information from people who have firsthand experience in going through this process, and some who actively or regularly give food to others.
By obtaining permission and conducting a meeting with the developers and owners of the foodsharing.nl website, we have obtained consent and were able to request 257 participants to participate in the online survey. After which, the link would be provided through foodsharing.nl after they have been informed about the survey.
It was assured that the informants were professionally conscientious, interested and committed to provide data by assuring them that the survey would be used to conduct a research on behalf of the University of Groningen to study a person’s food giving behavior, that the results would be used to provide valuable information to foodsharing.nl in order to improve their website and understand their users’ needs and wants, that the information obtained would be used for those purposes only and that their data would be kept confidential, and by offering them a report with the results of the study to whoever would be interested. Moreover, by sending the link to the survey which was included in the emails to the participants, it gave foodsharing.nl the opportunity to include questions in the survey which were of interest to them. This would provide valuable information to them as well as to create a more user friendly website, where participants were able to provide suggestions on how the website could be improved. Due to this, foodsharing.nl was interested and committed to this study. Further, a report would be provided to them with the relevant information which the participants provided, as well as a copy of the final results of this study.
26 and available online. Further, a reminder email was sent to the participants through foodsharing.nl requesting them to fill the survey if they have not yet done so before the deadline.
A total of 53 respondents have successfully and completely filled the online survey with a response rate of 20.62%. From the 53 respondents, 11.3% are men with an average age of 33 years old, and 88.7% are women with an average age of 39.26 years old. Moreover, 50% of the men are self- employed, 16.7% have a full time job, 16.7% are students and 16.7% are jobless.Most women, with an average of 25% have a part time job, 21.3% are jobless at the moment, 14.9% each are self-employed and have full time jobs, 4.3% each work at home or have home duties and 14.9% are students. Further, most men with an average of 33.3% have a university education and the rest have obtained a lower degree. For women, the majority which accounts of 44.7% has obtained the HBO degree which is similar to college degree and, 34% have obtained a university degree, 4.3% are post academics and the rest have obtained a lower degree than college. On average, a household consists of 2.13 people.
3.2. Development of the Survey
For most of the constructs in the model, existing scales were used which were obtained from academic and scholarly articles, and which have proven their reliability and validity in their respective research.
27 Mindful Mindset: Nature aspect Self-aspect Community aspect Mindful Behavior: Repetitive Acquisitive Aspirational
The formulated survey questions were constructed to represent the whole construct and were thus formulated by creating questions representing each of the characteristics associated with mindful consumption.
29
30 Most questions were formulated using a multi item7-point Likert Scale which provides for more accurate results and more diverse answers compared to a 5-point Likert Scale. This would significantly reduce the number of participants from choosing options in both extremes of the scale, which could result in biases and inaccurate results in subsequent analysis. For the survey questions relating to the constructs Mindful Consumption,
Prosocial Behavior, Happiness, and Giving trait, where the 7 means totally applicable to
me, 1 represents totally not applicable to me. For the construct Convenience, 1 means totally agree whereas 7 means totally disagree. The scale used to measure Values however was used with a multi item 9-point Likert Scale. This is consistent and adopted from the research conducted by Steg et al. (2014). The scale which was used in their research ranged from -1 to 7. Where -1 represents that the value is totally opposed to the principles that guide you and 0 means that the value is not important at all, 6 represents that the value is very important and 7 that the value is of supreme importance as a guiding principle in your life.
31 happiness, item 6.8 in Table 1 above, participants have had difficulties in understanding the question. After consultation with the professor, it was decided to remove this item and not include it in the final survey.
3.3. Reliability and Validity
In order to test the dimensionality and reliability of the scales a factor analysis was conducted.
Following the work of Malhotra (2010), a factor analysis (Principle Component Analysis)
with a Varimax (Orthogonal) rotation has been conducted for each of the 7 constructs included in the conceptual model, except for Gender. An examination of the Kaiser- Meyer Olkin measures of sampling adequacy suggests that the questions in their respective constructs were factorable where Biospheric, Egoistic, Hedonic, Altruistic,
Mindful Consumption, Pro-social Behavior, Giving, Giving Trait and Happiness all
obtained a KMO ≥.5. This was true for all constructs except for questions relating to the construct Convenience (KMO=.475). In order to inspect the low value of the KMO measure, frequency tables have been conducted in order to discover any large deviances in the filled questions by the respondents. However, there are no abnormalities found. Although there are answers which may vary in degree in some scales of the items, this is due to the relatively small amount of respondents obtained. However, by observing the Bartlett’s Test of Sphericity, all the items in their respective constructs were factorable, where they were all significant and below the threshold p-value <.05 (Malhotra, 2010).
The results of the orthogonal rotation of the solution are shown in Table 2 below. Loadings with values >.5 were kept in their respective factors which illustrates that there are high correlations between the variables and factors. A high value means that the factor draws significant information from the variable. Furthermore, we have removed the variables which had a cross loadings of >.3 as they are considered relevant to factor loadings and correlate with more than one factor. In selecting the number of factors, only those for which the eigenvalues greater >1, factors which explain >5% each, or where the total explained variance >60% were selected (Malhotra, 2010). In addition to, the number
32 from where the scree starts for all factors in their respective constructs. After the items with cross loading have been removed, the factor analysis was repeated to ensure that the items would still explain the same factors. Table 2 shows all the items and their loadings in their respective constructs. The bolded items in Table 2 illustrate the items which were used to measure the final constructs. The factors which explain the constructs most were used to represent the construct and were to be used in subsequent analysis. For the bolded items in Table 2, they represent the final loading on that factor meaning that those are the values after the items with low factor loadings were removed. Furthermore, the numbers that are allocated to the items refer to the questions of that construct shown in Table 1 in the previous page.
Construct and Items Factor 1 Factor 2 Factor 3 Factor
4 Communalities Explained Variance
Mindful Consumption:
Communalities
3 factors found: Factor 1 chosen
Mind_Cons_3 ,859 ,698
Mind_Cons_7 ,778 ,698 % of explained variance for chosen factor= 69,834 Mind_Cons_1 ,347 -,709 Mind_Cons_2 ,772 Mind_Cons_4 ,638 ,549 Mind_Cons_5 ,896 Mind_Cons_6 ,860 Prosocial Behavior: Communalities
2 Factors found: Factor 1 chosen
Prosocial_Beh_1 ,884 ,784
Prosocial_Beh_3 ,784 ,754 % of explained variance for chosen factor = 75,439 Prosocial_Beh_2 ,650 ,497 Prosocial_Beh_4 ,846 ,339 Prosocial_Beh_5 ,921 Prosocial_Beh_6 ,935 Convenience: Communalities
2 Factors found: Factor 1 chosen
Concenience_2 ,905 ,813
Concenience_3 ,898 ,813 % of explained variance for chosen factor = 54,368
Concenience_1 ,996
Values:
Communalities
4 Factors found: 4 Factors chosen
Biospheric_1 ,872 ,760 Factor 1 Biospheric Values: Altruistic_2 ,736
,542
% of explained variance for chosen factor = 73,004
33
Biospheric_4 ,913 ,833
Egoistic_2 ,678 ,460 Factor 2 Egoistic Values: Egoistic_3
,825 ,681
% of explained variance for chosen factor = 51,263
Egoistic_4 ,741 ,549
Egoistic_5 ,600 ,450
Hedonic_1 ,843 ,711 Factor 3 Hedonic Values:
Hedonic_2
,843 ,710
% of explained variance for chosen factor = 64,005
Hedonic_3 ,708 ,501
Altruistic_3 ,870 ,819
Altruistic_4 ,869 ,819 Factor 4 Altruistic Values:
Altruistic_1 ,448
,429
% of explained variance for chosen factor = 81,885
Egoistic_1 -,318 ,640
Biospheric_2 ,668 -,409
Giving_Trait
Communalities
2 Factors found: Factor 1 chosen
Giving_Trait_1 ,812 ,660
Giving_Trait_3 ,917 ,840 % of explained variance for chosen factor = 68,324
Giving_Trait_4 ,741 ,550
Giving_Trait_2 ,952
Happiness:
Communalities
2 Factors found: Factor 1 chosen
Happiness_1 ,870 ,862
Happiness_2 ,907 ,862 % of explained variance for chosen factor = 86,115 Happiness_3 -,398 ,470 Happiness_4 -,650 Happiness_5 ,515 -,622 Happiness_6 ,792 Happiness_7 -,370 ,338
Table 2: Constructs, Factor loadings, Communality, Validity and Reliability
Note: Bolded Items are used to measure the final constructs. Where the items which are Italic are removed
from the final constructs, the communalities, validity and reliability all represent the items which belong to the final constructs in bolded fonts. That is, the communalities and variance explained is the result of the factor analysis of the final constructs. The numbers associating the items relate to their respective numbered items in Table 1
34 whether they still measure the factor after all other irrelevant items have been removed. The final factor for Mindful Consumption consists of items 3 and 7 as they explain the factor most as shown by their high loadings where they are >.5. For this reason, those items are used to measure the construct of Mindful Consumption. The cumulative variance in Table 2 represents the final items used to measure the construct and excluding the other variables. This has been conducted to obtain all the constructs.
In general, 2 items load on Mindful Consumption, 2 items on Pro-social Behavior, 2 items load on Convenience, 2 items load on Giving, 3 items load on Giving Trait, and 2 items load on Happiness.
As for the factor analysis constructed for Values, it was mostly consistent with the work conducted by Steg et al. (2014) except for a few differences. Most items loaded in their respective constructs. However, since items Altruistic_1, Egoistic_1, and, Biospheric_2 load on more than two factors as shown Table 2, they were removed from the factor. Furthermore, another difference is that item Altruistic_2 loaded on the Biospheric value. For this reason, we have included the item Altruistic_2 with the Biospheric construct. This was followed by redoing the factor analysis with the relevant items to test the reliability of the final scale.
After the factor analysis was conducted, the internal consistency needed to be analyzed. For constructs which contain three items or more, the Cronbach Alpha was obtained to measure the internal consistency in order to proceed with those dimensions in the subsequent regression analysis. A Cronbach Alpha which is >.6 is required (Malhotra, 2010) to obtain a strong enough internal consistency. For the factors Biospheric, Egoistic, Hedonic and Giving Trait, the Cronbach Alpha was above the threshold of >.6. For
constructs which include less than three items, a bivariate correlation (Pearson Correlation) was conducted (Malhotra, 2010). As we cannot use the Cronbach Alpha for
35 be significant. The constructs Mindful Consumption, Pro-social Behavior, Convenience,
Giving, Altruistic Values and Happiness were all significant and are below the p-value of
.10. Table 3 below illustrates the Cronbach Alpha from the reliability analysis, and Pearson Correlations obtained from the bivariate correlations, as well as the means and standard deviations. Category Constructs # of items α Mean SD Moderators: Values: Biospheric 4 ,875 7,94 0,995 Egoistic 4 ,672 4,54 1,263 Hedonic 3 ,691 7,09 1,063 Control Var.: Trait: Giving Trait 3 ,760 14,57 3,651 Constructs Constructs # of items Pearson Corr. Mean SD Independent Var. Mind_Cons 2 ,397** 4,57 1,535 Posocial_Beh 2 ,509** 3,91 1,345 Convenience 2 ,626** 4,55 1,656 Moderator Altruistic 2 ,637** 6,98 1,109 Independent/Dependent Giving 2 ,686** 2,55 4,516 Dependent Var. Happiness 2 ,723** 4,89 1,257
Table 3: Cronbach’s Alpha, Peasron Correlations, Means, Standard Deviations (N=128)
36 1 2 3 4 5 6 7 8 9 10 1. Mindful_consumption Pearson Correlation 1 2. Pros_Beh Pearson Correlation ,073 1 3. Convenience Pearson Correlation -,136 ,147 1 4. Biospheric Pearson Correlation ,168 -,213 ,061 1 5.Egoistic Pearson Correlation -,109 ,154 ,076 ,121 1 6. Hedonic Pearson Correlation ,061 ,215 ,112 ,066 ,176 1 7. Altruistic Pearson Correlation ,083 ,173 -,023 ,341 b ,006 ,110 1 8. Giving Pearson Correlation ,118 ,129 ,140 ,032 -,169 -,027 ,130 1 9. Giving_Trait Pearson Correlation -,113 ,197 ,054 -,048 -,150 ,062 .451a ,136 1 10. Happiness Pearson Correlation -,089 ,102 .506 a -,172 ,075 ,075 -,115 ,020 ,005 1
Table 4: Correlation Coefficients
Note: a Sig. at p-value< .01, b Sig. at p-value< .05
Further, the variance inflation factor (VIF) is used in the regression analysis to indicate any multicollinearity issues. This is the reciprocal of the tolerance and where a high VIF >10 indicates high multicollinearity, and a VIF >4 indicates a moderate multicollinearity (Malhotra, 2010). This is important as it can adversely affect results associated in a
regression analysis. Due to this issue, the VIF scores have been analyzed and were all proven to be below a VIF score of 4. Mulicollinearity does not present any issue found in this model. Appendix 1 illustrates the associated VIF scores.
3.4 Method of Analysis
3.4.1. Econometric Model and the Main Effects for Part 1
Using the linear and multiple regression methods, the following equations to test the main effects in Part 1 are estimated:
(1) GF= α1+ β4GT+ ε
37 Equation (1) concerns the control variable Giving Trait (GT). Depending on the outcome of (1), GT will be included in the other models in the case of when a significant relationship would exist. This equation would measure whether GT has a main effect on GF. Equation (2) includes Mindful Consumption (MC), Pro-social Behavior (PB), and
Convenience(C) as antecedents of Giving Food (GF). This equation would estimate the
main effects of the independent variables on the dependent variable GF. All equations include a constant (α1) and an error estimate (ε).
3.4.2. Econometric model and Moderating Effects for Part 1
Due to the small sample size, it was not possible to include all moderators is one regression analysis. Thus the moderators and their interaction effects were conducted in a stepwise method and where the interaction effects were computed manually. This was also conducted in the same manner for the moderating effects of Part 2.
38
3.4.3. Econometric Model and the Main Effects for Part 2
The following equations used to test the main effects in Part 2 of the conceptual model are estimated:
(15) HA= α1+β1GF+ε
Equation (15) - Model 15 in Table 7 below- measures the direct relationship between the independent variable GF and the dependent variable Happiness (HA). Table 7 demonstrates the results of the regression analysis conducted to measure this relationship which will be explained in the results section below.
3.4.4. Econometric model and Moderating Effects for Part 2
The equations below are estimated to determine the moderating effects of the values on the relationship between the GF and HA.
(16) HA= α1+β1GF+ β2EV*GF+ β3EV+ ε (17) HA= α1+β1GF+ β2HV*GF+ β3HV+ ε (18) HA= α1+β1GF+ β2AV*GF+ β3AV+ ε (19) HA= α1+β1GF+ β2BV*GF+ β3BV+ ε (20) HA= α1+β1GF+ β2GEN*GF+ β3GEN+ ε
39
4. Results
4.1 Results Part 1
Concerning the control variable Giving Trait, Table 5 and Table 7 below illustrates the results of the regression analysis, and shows that GT has no effect on GF (Model 1). For this reason, it is not included as a control variable in the rest of the analysis. The regression results for hypothesis H1, H2 and H3, and which relates to the two equations above are labeled as Model 1 and Model 2 in Table 5. H1 which hypothesizes that MC has a positive effect on GF is not supported (p>.10) (Malhotra, 2010). MC has no effect
on GF. Further, there is no support for H2 which hypothesizes that PB leads to more giving food (p>.10). Furthermore, H3 which states that the more convenient it is to give food to others, the more will be given, is not supported by the results (p>.10) (Malhotra, 2010).
The regression results for equations 3 to 14, Table 5 and Table 6 below, which determines the moderating effects of values between the independent variables and the dependent variable GF, is not supported as well (p>.10)
40
Table 5:Regression Results for part 2 of the conceptual model (Dependent variable: Happiness) Note: ** Sig. at p-value < .10
Hypothesis (Effect) Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
(Constant) ,097 -2,278 -2,771 -3,463 4,499 ,198 -15,962
Main Variables: Beta Beta Beta Beta Beta Beta Beta
41
Table 6:Regression Results for part 2 of the conceptual model (Dependent variable: Happiness) Note: ** Sig. at p-value < .10
4.2 Results Part 2
Results indicate that the relationship between GF and HA is insignificant as the p-value is above the acceptable threshold (p>.10) - Model 15 in Table 7 below. This leads to the rejection of hypothesis H6 (giving food to other leads to happiness) where there is no relationship found between giving food and its effects on happiness.
Further, the results indicate that GEN has no effect either as a moderator or as a direct effect on HA (Model 20 in Table 7 below). It shows that it is above the acceptable significance level where p >.10. For this reason, hypothesis H7 is rejected. Gender does Hypothesis (Effect) Model 8 Model 9 Model 10 Model 11 Model 12 Model 13 Model 14
(Constant) 4,997 -16,478 -15,379 -22,515 -14,906 -8,584 -16,813
Main Variables: Beta Beta Beta Beta Beta Beta Beta
42 not show to have a moderation effect or influence the relationship between giving food on happiness.
The hypothesis regarding H8 and H9 are not supported by the results and are thus rejected as well as they assume that a positive relationship between GF and HA exists. See Table 7 below for the regression results. Results show however that there is a relation found on the effect among the moderators EV, AV, and BV between the independent variable GF on Happiness. The regression results for equation (17) - Model 17 in Table 7 -illustrates that the model is significant (p value= .059; R2adj = ,087). When the EV in included as a moderator in the regression analysis, the independent variable GF is shown to be significant and where there is a negative effect on HA (p= ,010; β= -,606). Further, a moderation interaction is found between the EV and GF on HA (p= ,008; β= ,160). When people give food, they become less happy. However, as illustrated, the interaction effect between Egoistic value and GF has a positive effect on Happiness. This shows that giving food to others leads to less happiness, but when they give food to others and the person has egoistic values, they become happier. The effect of GF on HA is only available when including the Egoistic value as a moderator in the model. There is no direct relationship found between EV and HA where it shows to be insignificant (p-value >.10)
The regression analysis results for equation (18) - see Table 7- illustrates that the model is insignificant (p-value>.10) (Malhotra, 2010). No effect has been found by the HV as having a moderation effect.
43 Altruistic values become less happy. The moderating effects of Altruistic values of GF on HA have a negative impact. Further, there is no direct relationship found of Altruistic Values on Happiness (p >.10).
Lastly, results from the regression conducted in relation to equation (20) shows that the model is highly significant (p = ,004; R2 adj = ,185). The interaction effect between BIO and GF on HA is highly significant and has a negative effect on HA (p= ,001; β= -,212) and where the effect of GF on happiness is also highly significant at (p= ,001; β= 1,645). Giving food to others has a positive relationship meaning the people who give food become happier. However, when people give food to others and have biospheric values, it has a negative effect on happiness meaning that people who give food and have biospheric values become less happy.
There is an effect of GF only when the values- Egoistic, Biospheric, and Altruistic- are included as moderators in the regression analysis. In the absence of the moderators however, there is no effect between GF and HA.
Hypothesis (Effect) Model 15 Model 16 Model 17 Model 18 Model 19 Model 20 (Constant) 4,882 5,748 5,843 3,729 3,931 4,509
Main Variables: Beta Beta Beta Beta Beta Beta
Giving H6 (+) ,005 -,606** -,721 1,046** 1,645** ,408 Moderators: Biospheric H8 (+) ,138 Egoistic H9 (+) -,222 Hedonic H9 (+) -,141 Altruistic H8 (+) ,188 Gender H7 (+) ,391 Interaction Effects: Inter_EGO_GIV ,160** Inter_HED_GIV ,105 Inter_ALT_GIV -,153** Inter_BIO_GIV -,212** Inter_Gen_GIV -,203 R2 ,000 ,140 ,062 ,154 ,232 ,010 (Adjusted R2) (-,091) (,087) (,004) (,103) (,185) (-,051) F-Value ,091 2,656** 1,075 2,982** 4,935** ,163
Table 7: Regression Results for part 2 of the conceptual model (Dependent variable: Happiness)
44 Aside from this, negative adjusted R2 have been encountered in the analysis which can be observed in tables 5, 6 and 7. There can be different reasons for this. A reason can be is that the items have been coded in the wrong manner where positive items are coded negatively or vice versa, which can produce inaccurate results. This however has been analyzed and the questions are coded correctly. This did not seem to be the reason for the negative adjusted R2. Although the adjusted R2 is not negative for any linear regression models with an intercept, an adjusted R2 canbe negative where a regression model may use data which contains outliers – influential observations (Kvålseth, 1985; Quinn et al., 2002; Norusis, 1990). For the construct Giving, from the population size of 53, 13 respondents have given away food above the norm which is 6-7 items. Where 4 respondents have given between 18 and 20 items, 4 respondents have given between 14 and 15 items, and 5 people have given between 9 and 13 items, and where the rest have given between 1 and 7 items. There is a large difference in the amount of food items given away where 75.45% have given less the 7 items, and where 24.5% have given food anywhere between 9 to 20 items. This causes for outliers to be present in the measure.
In order to determine whether the reason for the negative adjusted R2 is due to this, the respondents who have given items in a total of more than 7 have been removed from the sample. After this has been done, the regression analysis has been conducted without those respondents. However, the adjusted R2 was still negative. A reason for this negative adjusted R2 might be the reduction in the number of respondents from 53 to 40, which is a very small sample size.
5. Conclusion
45 the results are all above the significance level p>.10, it shows that it is not significant which leads to the rejection of hypothesis H4 (A, B, C) and H5 (A, B, C).
Concerning the regression analysis for Part 2 of the model, the results do not support the hypothesis. The effect of GF on HA is not significant (p>.10). Thus, H6 is rejected. Furthermore, H7 is rejected as the p>.10, meaning that no relationship has been found concerning GEN on HA. Although hypothesis H8 and H9 are rejected as well, which is due to the hypothesis stating that a relationship between GF and HA exists and that the values influence the relationship, there is a relationship found. A relationship between GF and HA exists when the values (EV, BV and AV) are included in the analysis. Model 17 in Table 7 shows that the model is significant (p= .059; R2 adj= ,087). EV shows to have a positive moderation effect between GF and HA (p= ,008; β= ,160), while there is a negative effect found between GF and HA (p=,010; β= -,606). Moreover, A moderating effect is found for AV where the model is significant (p= ,040 and R2 adj = ,103), as well as for BIO where it is significant at p= ,004 and R2 adj= ,185. This is illustrated in Table
7 as Model 18 and 19 respectively. When the ALT and BIO values are included in the
analysis, a positive relationship is found between GF and HA (p= ,006; β= 1.046 and p= ,001; β= 1,645 respectively). However a negative moderation effect is found between the values and the interaction between GF and HA (p= ,006; β= -,153 and p= ,001; β= -,212 respectively). Showing that although people are more happy when they share, they become less happy when they have ALT and BIO values.
All in all, the hypothesis is not supported however there are significant relationships found which are discussed in sections 6.1 and 6.2 below.
6. Discussion
6.1 Discussion of Part 1 in the Conceptual Model
47 create 6 different constructs of mindful consumption where they each capture 1 of the 6 characteristics. Results illustrate that there is no relationship found of giving trait on giving food. For this reason, giving trait is excluded as a factor in which the drivers to give food might not have an effect due to a person being a naturally generous person.
48 others have intended to consume the food for them-selves. They purchase the food for their own benefit. Due to there being leftovers and not wanting to waste food, they might decide to give it to another person. However this action was not intended to help another person in the first place but is a consequence or result of not wanting to throw away the food, and where helping another person is an outcome of not wanting to waste the food. Thus they may not feel that they have given another person food for pro-social reasons. Moreover, as mentioned in the section above, people might give for other reasons other than helping others such as preserving the environment and reducing waste. Pro-social behavior might be a reason to give however importance on other aspects might overshadow this intent.
Further, giving food on foodsharing.nl is different from giving food to others personally, or to family or friends. People who receive the food from others through a website are able to access computers and most of them probably own them. These people are able to afford such technology which is costly and are also able to afford food. Further, most of the respondent have educational backgrounds and have a job. Thus the need to give food are for other reasons mentioned such as preserving the environment and reducing food waste instead of pro-social behavior. In other words, the giver can see that the receiver is able to access a computer and who probably own it, so they know that the people receiving food are able to afford buying food for them-selves.
49 react to others, 39% of people do not agree that there is little effort needed to search for others to give away food on the internet, and where 15% of respondents are neutral to this. This shows that although it is easy to post the items on the internet and react to people’s comments who are interested in obtaining the items, there is effort needed. This effort is due to the search of people who are willing to obtain the item. The receiver might live in a location further away and might not be willing to travel the distance to obtain the item. Moreover, it is relatively difficult to find people to give away food who are living in the same district as the giver. This creates a convenience problem as well. Giving food on the internet creates effort during the whole process from posting the item online, to handing the item over to the receiver which completes the cycle. The giver first needs to package the food and store the item, then has to post the item online and provide a description of the items, and where it can be picked up from. After this step is taken, the giver has to wait for a person to respond and who is interested in the item and if they are interested, communication between the parties must be achieved until the receiver obtains the item from the designated location where the giver is present. This illustrates that although giving food to others can be done from the comfort of one’s home, effort is needed to go through the process. Furthermore, there might not be a person who is interested in the item which in turn causes more effort as the person went through the process of posting the item, and which needs to be disposed of due to no person wanting it.
50 people, giving food to others so they themselves eat less because they want to manage their diet, and do not see giving away of food as a waste of money in contrast to throwing it away. Giving away food so that the person giving it away eats less food, does not fit with the dimensions of either hedonic or egoistic. It may be however the items which are used to measure those constructs do not measure such an aspect. As hedonic is related to joy and pleasure, a person who gives away food so they eat less might not see it as hedonic. Further, egoistic does not measure this as well, as the items for measuring the construct egoistic values relate to power over others.
Further, the values do not consider the interaction of the different values. It is studied (Steg et al., 2014) that people have an either/or importance on a value and do not for instance combine the effects of values. People might place importance on more than one value equally inhibiting altruistic as well as biospheric values, or egoistic as well as hedonic values and so on.
