Sasja van der Leeuw
University of Groningen Faculty of Economics and Business MSc Marketing
June 14, 2016 Lingestraat 9b 9725 GL Groningen
(06) 554 08 258
N.m.van.der.leeuw@student.rug.nl Student number 2075962 Supervisor: Koert van Ittersum
CAN-Approach; leading children towards
healthy choices in sports canteens
2
Table of Contents
PREFACE ... 4
ABSTRACT ... 5
1. INTRODUCTION ... 6
BACKGROUND: ... 6
INFLUENCE OF ENVIRONMENT ON FOOD CHOICE: ... 6
PROBLEM STATEMENT AND RESEARCH QUESTION: ... 8
CONTRIBUTION TO EXISTING KNOWLEDGE: ... 9
STRUCTURE OF THESIS: ... 9
2. RESEARCH FRAMEWORK: ... 10
LITERATURE REVIEW ... 10
FOOD CHOICES ARE INFLUENCED BY OUR ENVIRONMENT ... 11
2.1.PRODUCT: ... 12
2.1.1CHANGE IN ASSORTMENT ... 12
2.1.2CALORIE/NUTRITION LABELING ... 13
2.2.PLACE ... 14
2.2.1AMBIANCE CHANGE ... 14
2.2.2DINNERWARE SIZE ... 15
2.3.PRICE ... 16
2.3.1PRICE STRATEGY ... 16
2.3.2PAYMENT METHOD ... 17
2.4.PROMOTION ... 18
2.4.1HEALTH MESSAGES / SOCIAL MARKETING ... 18
2.4.2MENU STRATEGIES ... 19
2.4.3CAN-APPROACH ... 20
LICENSED CHARACTERS ... 22
PRODUCT BUNDLING ... 23
3. HYPOTHESES ... 24
3.1COLLABORATION WITH KNVB AND DISNEY ... 24
3.2THE INTERVENTION –CHARACTER LICENSED COMBO BOX ... 25
3.3MODERATOR - PERMISSIVE EATING ... 27
4.RESEARCH DESIGN ... 30
4.1RESEARCH METHOD ... 30
4.2DATA COLLECTION ... 31
SURVEY ... 31
SALES DATA ... 31
4.3MEASUREMENTS ... 31
4.4PLAN OF ANALYSIS ... 33
5. ANALYSIS AND RESULTS ... 34
5.1PREPARING THE DATASET ... 34
5.2SAMPLE ... 34
5.2FACTOR ANALYSIS MODERATOR PERMISSIVE EATING ... 38
5.3EVALUATION OF REGRESSION ASSUMPTIONS ... 40
5.4REGRESSION ... 41
MODEL 1-CONTROL VARIABLES ... 41
3
MODEL 2-MULTIPLE REGRESSION (HYPOTHESIS 1A) ... 42
MODEL 3-MULTIPLE REGRESSION (CONTENT BOX) ... 43
5.4INTERACTION –PERMISSIVE EATING ... 45
MODEL 4– INTERACTION EFFECT ... 45
5.5T-TEST –ACTUAL PURCHASE HEALTHY AND UNHEALTHY OPTIONS ... 46
RESULTS ... 47
DIFFERENCE IN ATTITUDE BOX BUYERS VERSUS NON-BUYERS ... 48
5.6OBSERVATIONS ... 49
6. CONCLUSION AND RECOMMENDATIONS ... 50
6.1CONCLUSION ... 50
6.2DISCUSSION ... 52
6.3MANAGERIAL IMPLICATIONS ... 56
6.4THEORETICAL CONTRIBUTION ... 57
6.5LIMITATIONS ... 57
6.6FURTHER RESEARCH ... 59
REFERENCES: ... 60
APPENDICES ... 69
APPENDIX I:THE HEALTHY COMBO BOX ... 69
APPENDIX II:NUTRITION FACTS ... 70
APPENDIX III:SURVEY ... 71
APPENDIX IV–DESCRIPTIVE STATISTICS ... 73
APPENDIX V–FACTOR ANALYSIS ... 74
APPENDIX VI–REGRESSION ASSUMPTIONS ... 77
APPENDIX VII–SALES DATA ... 79
APPENDIX VIII–OBSERVATIONS ... 82
APPENDIX IX–BOOKLET WITH TIPS ... 84
APPENDIX X–PHOTOS OF THE EXPERIMENT ... 87
APPENDIX XI-SPSSOUTPUT ... 88
4
Preface
This paper has been written for the completion of my Master Marketing at the Rijks Universiteit Groningen, which started in February 2016 and ended in June 2016. In this study I was aiming to contribute to a solution to the obesity problem we face nowadays, by conducting a field study in which I introduced a ‘healthy combo box’ licensed by Donald Duck to make children eat healthier in soccer canteens. I would like to express my gratitude to the KNVB and especially to P. Legters - project leader of the KNVB - for giving me the opportunity to work with them in such a challenging project. I also want to thank my supervisor Dr. K van Ittersum for all the pleasant corporation and guidance to complete this project and VV Heino for the opportunity to use their canteen to conduct the research. Even though a field experiment comes with some difficulties once in a while, I am very glad that I have been able to complete this study for an actual organization.
5
Abstract
The growing number of children who are suffering from overweight is of great public concern. The physical environment plays a key role when it comes to consumption decisions leading to overweight. The sports canteen is a place that is often visited by children, but still often promotes unhealthy food choices. In order to contribute to healthier eating habits of children in sports canteens, this research is done in commission of KNVB (Royal Dutch Soccer Association).
The aim of this study is to test if the CAN-approach in the form of a combination of character licensing and product bundling (‘healthy Happy Meal’) stimulates children to make healthier consumption choices. The healthy combo box consists out of two healthy options;
popcorn and DubbelFrisss Free and is licensed by Donald Duck. A field study is conducted in which the healthy combo box was sold in an actual soccer canteen. Sales data of the healthy and unhealthy items of two weeks before and two weeks during the intervention were compared. A paired samples t test was used to see if the consumption of healthy foods increased and the consumption of unhealthy foods decreased during the introduction of the box. Besides a survey is conducted among 95 children between the ages of 4-14 to measure the influence of their attitude towards the box on their purchase intention, making use of a regression analysis.
Results show that a positive attitude towards the box significantly increased the purchase intention of the box. The box partly increased the consumption of healthy foods;
sales of DubbelFrisss Free significantly increased during the intervention with 200%. The consumption of unhealthy foods only significantly decreased for the catagory crisps, during the intervention 56% less bags of crisps were sold. Only 6 children bought the box, because the majority (86%) did not notice the box. In conclusion, the healthy combo box seems to be a potentially successful strategy for sports canteens to stimulate healthy consumption. In combination with promotion the box might increase the positive effects on food consumption even more.
Key words: CAN-approach, character licensing, product-bundling, healthy consumption, children, overweight.
6
1. Introduction
Background:
The number of people that are obese worldwide is increasing. In 2014 worldwide more than 1,9 billion adults were overweight and 600 million individuals fell in the category of obese (World Health Organization, 2015). When it comes to children, worldwide 43 million under the age of 5 were overweight in 2013 (World Health Organization, 2015). Recent research on overweight in The Netherlands indicates that on average more than 11% of children between 4-16 years are either overweight or obese (CBS, 2014). These numbers are alarming; children with extra weight have an increased risk of serious health problems including diabetes and cardiovascular diseases and cancer, but can also have a negative impact on psychological health (World Health Organization, 2015). They are at risk of getting bullied and have lower perceived competences in physical activities and peer engagement, which can lead to a lower self-esteem (Franklin, Denyer, Steinbeck, Caterson & Hill, 2006). If measures are not been taken in time, two-third of the Dutch is estimated to be overweight in 2024, with all its consequences that will entail (NIVEL, 2014).
Influence of environment on food choice:
Numerous studies have found environmental factors to be a driver for the choices consumers make regarding to food (Hill & Peters, 1998; Hill, Wyatt & Melanson, 2000; Wansink, 2010).
Some believe that the growing obesity problem is largely caused by an environment that stimulates the consumption of too many calories (French, Story & Jeffery, 2001). An important concept to understanding the influence of environments on obesity is the
“obesogenicity” of an environment which is defined as “the sum of influences that the surroundings, opportunities, or conditions of life have on promoting obesity in individuals or populations” (Swinburn, Eggers & Raza, 1999. p. 564).
7
These environmental cues are consisting of but not limited to; assortment composition, point-of-purchase health messages, nutrition-labeling, presentation of the food, size of dinnerware and sense cues such as lighting and music. Research has shown that health-messages are associated with healthier consumption choices for both children and adults (Buscher, Martin & Crocker, 2001; Gordon, McDermott, Stead & Angus, 2006; Shive
& Neyman Morris, 2006). Numerous of studies have confirmed that nutrition labeling in eating environments is an effective way to stimulate healthy food intake (Dubbert et al., 1984;
Levin, 1996; Vyth et al., 2011).
Another factor that influences food choice is the composition of the assortment.
Simply increasing the availability of healthy options, while still permitting access to unhealthy variants, can create healthier eating patterns (van Kleef, Otten & van Trijp, 2012).
Besides, the presentation of food plays a crucial role in selecting food. Wansink (2015) has introduced the CAN-Approach, which stands for convenient, attractive and normal. Making healthier choices more convenient (easier to get), attractive (in the form of packaging, food names or display/presentation form) and normal (make it appear as an normal/popular choice), can help transform our eating environment into one that stimulates healthy consumption (Wansink, 2015). His framework will be the basis of the current study and will be applied in the form of a character licensed combo food box (“Healthy Happy Meal”).
Character licensing is the endorsements of products by famous cartoons. This method can be assigned to the CAN-approach. In the context of brands or food in general, studies evidenced that children show an increased liking of food items that have a famous cartoon character on the packaging (Mizerski, 1995; Garretson & Niedrich, 2004, Roberto, Baik, Harris and Brownell, 2010). Product bundling is also applied in the combo box, as research proved this method to be effective in increasing healthy consumption (Wansink & Hanks, 2014).
8
Problem statement and research question:
The growing amount of children that are overweight is of great concern for public health. The environment plays a key role in forming and maintaining healthy eating habits and keeping the energy intake and energy use in balance. However, public foodservice settings are often designed to stimulate unhealthy choices (Brug, Kremers, van Lenthe, Ball & Crawford, 2008;
Swinburn, Egger & Raza, 1999; Wansink, 2010; Buscher, Martin & Crocker, 2001). An example of such public settings is the sport canteen, a place where children come and purchase foods/drinks on a regular basis. In The Netherlands, on average 50,1% of children between 4-12 years old is member of a sport association and in the category 12-16 year-olds 60,7% is member (CBS, 2014). Nevertheless, sports canteens often promote unhealthy food choices by selling fried snacks, candy, crisps and sugary drinks.
This research is done in commission of KNVB, Koninklijke Nederlandse Voetbal Bond (Royal Dutch Soccer Association), located in The Netherlands. This is the governing body of soccer, which organizes the Dutch soccer leagues as well as the amateur leagues and the national teams. The KNVB has 1,2 million members of which more than half is considered youth (KNVB, 2015). Since it is the biggest Dutch sports association, where children are participating in competitions on a weekly basis, it wants to take its social responsibility to make the sports canteen environment healthier. Therefore this study is interested in finding environmental cues in line with the CAN-approach that stimulate healthy food intake by making healthy options more Convenient, Attractive and Normal. More specifically, this research wants to find out if character licensing in combination with food- bundling in the form of a “healthy Happy Meal” has an effect on healthy food intake of children. This leads to the following research question:
• Does a character licensed combo food box stimulate children to make healthier consumption choices?
9
Contribution to existing knowledge:
This research contributes to the urgent need to change the environment of children that stimulate unhealthy food habits to an environment that stimulates healthy choices. There is a lot of research on environmental cues and its influence on healthy food choices in the context of school, university and worksite cafeteria. Not much is known yet about the effect of environmental changes in a sports canteen setting. What is known is that these venues are of great influence on shaping the image of children of what is acceptable regarding nutrition choices. Nevertheless, the food environment at sports canteens often does not stimulate healthy food choices. After practicing sports kids often go to the canteen and buy some candy and soda or sometimes even a fried snack. With the growing problem of obesity in mind, it is time to gain understanding on how soccer canteens can be designed to stimulate healthy choices and by doing so contribute to the solution of the overweight problem. That is why the focus of this study is on environmental changes in sports canteens that stimulate healthy food choices among children between the age of 4-14 years.
The contribution to the literature is as follows: a new way of stimulating healthy choices among children will be tested. Little is known yet about the effect of character licensing on promoting healthy food to children, especially in sports settings. This research combines character licensing with a default choice in the form of a combo box with two pre- selected healthier food items. This combo box will be sold in a soccer canteen to see if this stimulates children to pick healthier food than they normally do. Data to answer this question will be provided with a field experiment in a sports canteen in combination with a survey.
Structure of thesis:
The structure of the remainder of this article is as follows: To create the foundation of this study, research on relevant environmental cues is discussed in the literature review. Based on this literature review a conceptual model and hypotheses will be derived. After that the
10
research design will provide the methodology of the experiment, followed by the analysis and discussion of the results obtained in this study. The final part will consist of practical implications of environmental changes in sports canteens and the theoretical contribution to the existing theory will be discussed.
2. Research framework:
Literature review
Much research is conducted to find out how healthy eating can be influenced by interventions in the environment. This section reviews the previous literature on ways to affect consumers’
behavioral preferences towards healthier food choices. The strategies to steer people towards healthy eating are very diverse. Among the type of interventions are changing the assortment, point-of-purchase messages, calorie labeling, changing plate or cutlery size and change the presentation order of the food. These strategies will be categorized on the basis of the four P’s of marketing: Product, Place, Price and Promotion.
In addition to type of intervention a distinction can be made between the type of setting such as school canteens, university canteens, worksite canteens and restaurants.
Results of these various settings will be discussed. In this study we are focusing on changing the eating behavior of children in sports canteens. A healthy alternative of a “Happy Meal”
will be introduced in a soccer canteen, in the form of a character licensed combo box.
Figure 1 shows a conceptual framework of how the variables are related to each other and how these constructs are measured. A more general discussion of determinants of the environment that influence individuals’ food choices are given in the first paragraph. The following subsections discuss the several types of interventions that are already investigated, subdivided in the four P’s of marketing, ending with the type of intervention studied this paper.
11
Food choices are influenced by our environment
In order to promote healthy food choices, knowledge about the determinants of food choices is necessary. Research has shown that socio-cultural and physical environments may influence the eating behavior of individuals (Swinburn, Egger & Raza, 1999; Hill, Wyatt &
Melanson, 2000; Brug et al., 2008).
Swinburn, Egger & Raza (1999) came up with The ANGELO framework, a framework to identify potential intervention strategies to influence food intake. It makes a distinction between micro and macro environments on the one axis and the other axis is dissected into four types of environments: physical, economic; political and socio-cultural. To the microenvironment belong settings such as schools, universities, workplaces, sports canteens and restaurants. Macro-environments are sectors that influence energy-intake behavior on a higher level and are less controllable by individuals. Examples of macro- environments are the health systems, government institutions, nutrition policies and the food industry (Swinburn, Egger & Raza, 1999). In this study the focus will be on the physical environment on a micro-level.
Microenvironment settings can be designed or set-up in such a way that people are stimulated to make healthier choices. The assortment composition (Kahn & Wansink, 2004), point-of-purchase health messages (Buscher et al., 2001), nutrition information (Cranage, Conklin & Lambert, 2005; Freedman, 2011) and the size of dishes and cutlery (Mishra, Mishra & Masters, 2012; Van Ittersum & Wansink, 2012) are all examples of environmental determinants that have been demonstrated to influence consumers’ eating behavior. These elements can suggest which food choices are seen as normal, it can make certain choices more obvious or attractive or it can steer the choice decision of people towards a certain alternative.
12
2.1. Product:
2.1.1 Change in assortment
The assortment of a foodservice place determines the purchase options you have as a consumer. If there can be only chosen between nutrition pore, energy rich foods it is impossible to make healthy choices. Research had been conducted on the effect of assortment change on healthy food intake. Jeffery, French, Raether & Baxter (1994) conducted a study in a cafeteria for University employees, where they doubled the number of fruit choices and added three ingredients to the salad ingredient selection. The results were three times higher sales of fruit and salad during the intervention period. During the study the price of those healthy choices was also reduced, so it cannot be assured that assortment change accounted for the increase of sales. Similar results are found in a study of Bucher, van der Horst and Siegrist (2011) at a University of Zwitserland, where increasing the vegetable variety influenced the students towards healthier consumption without increasing the total energy served. The classical psychological theory of human motivation might be the explanation of this effect; having more choice variety is more desirable and intrinsically motivating (Iyengar, Lepper, 2000; Cordova & Lepper, 1996).
Shelf arrangement also plays an important role. Research has shown that increasing the prominence of healthy food in the shelves promotes healthy consumption (van Kleef, Otten & van Trijp, 2012; French, Story, Fulkerson & Hannan, 2004). A possible explanation for this is that consumers think important brands/products are given priority in the shelves and therefore they presume that products with a prominent place are better (Buchanan, Simmons
& Bickart, 1999).
Furthermore, providing attractive opportunities to try new healthy food as part of a larger nutrition campaign is evidenced to increase students’ fruit intake (Perry et al., 2004).
Freedman and Brochado (2010) showed that decreasing the package size of pre-packed foods
13
is also a way to decrease calorie intake by assortment change. It can be concluded that there is extensive support of the influence of assortment on changing food choice behavior.
2.1.2 Calorie/nutrition labeling
Providing nutrition information at the point selection is a way to make people more aware of the healthiness of the food choices they make. Research has shown that nutrition information can be an effective way of stimulating healthy eating decisions. A study of Jeffery, French, Raether, & Baxter (1994) executed in a worksite environment supported the hypothesis that increasing the amount of nutritious food choices influenced the dietary intake in a positive way. Similar results are found in a restaurant setting by Cranage, Conklin and Lambert (2005); entrees high in fat were bought less and entrees low in fat were bought more when nutrition information was displayed. The reasoning behind this is that consumers are able to make more informed choices, when they have access to the nutrition facts (Cranage, Conklin
& Lambert, 2005).
Nutrition information in the form of symbols is proved to increase healthy food intake in worksite settings (Dubbert et al., 1984; Levin, 1996; Holdsworth & Haslam, 1998).
Evidence also exists in a University setting where calorie information was provided next to the French fries and salad dressing in an all-you-can-eat dining hall (Freedman, 2011). The calorie content information resulted in an increase of people choosing small portion size instead of the larger option.
Hoerr and Louden (1993) showed contradicting results in a two-year during study on a University campus. After adding nutrition information with bar graphs on the snacks that were sold in the vending machines, sales increased. Nevertheless, this was mainly the case for the less healthy options. There is more evidence for the positive effect of nutrition labeling on the healthiness of food choices. Therefore can be concluded that nutrition labeling is a potentially effective way to stimulate healthy food intake in soccer canteens as well.
14
2.2. Place
2.2.1 Ambiance change
Research has shown that the ambience of the dining location plays an important role in the tempo and amount of food intake. Ambience can be defined as “parts of the environment that are hard to localize and that surround the organism integrally like temperature, sound, smell”
(Stroebele & Castro, 2004).
Music
Music is used in restaurants to neutralize sounds from other guests to make the dining experience more pleasant. However, music can contribute to more than covering up unpleasant sounds. There is extensive evidence that demonstrates that music can influence an individual’s eating behavior. Prior research among students in a natural setting evidenced that soft drinks consumption increased with increased auditory stimulation (McCarron & Tierney, 1989). They could serve themselves freely of a supply of soft drinks and found out that the higher the volume of the music, the more people consumed. Similar results were found for the consumption of food; the tempo of the music influences chewing intensity (Roballey et al., 1985). Faster music was related to an increased number of bites. An experiment in a dining area setting demonstrated that slow music led to a longer meal duration, which led to higher energy intake mostly in the form of drinks (Milliman, 1986). Findings of a study of Stroebele and Castro (2006) were not in line with those results. In their experiment the tempo of the music did not have a significant impact on the food intake nor the meal duration.
Nevertheless, in line with Roballey et al. (1985) they did find that the presence of music leads to a higher food intake (Stroebele & Castro, 2006).
There is no clear answer on how to decrease food intake by using music. Part of the studies found that loud and fast music seems to increase food intake. However, other researchers showed that slower music at a lower volume is associated with a longer meal
15
duration, which in turn can also lead to an increased energy intake (Milliman, 1986). This means that if you strive to decrease the food intake, a potential strategy is to play no music at all.
Lighting
Research has provided evidence that individuals are less inhibited and less self-conscious when there is less lighting (Kasof, 2002). With soft or warm lighting people take more time to eat and are more prone to taking an extra desert or drink, because their decreased self- consiousness makes them consume more than they otherwise would (Ragneskog, Brane, Karlsson & Kihlgren, 1996). Controversially; bright light is associated with the increase of speed in which people finish their meals (Birren, 1988). An example of this is the use of very bright lights in fast food restaurants, to speed up the cycle time of customers. At the same time this quick way of dining in fast-food restaurants or cafeterias is associated with a higher total energy intake than when you eat at home in dimmed lighting (McCrory et al., 1999).
Dimmed light seems to increase unplanned food intake because of decreased self- consciousness and dining by bright light leads to quick eating but at the same time an increase of total calorie intake overall. Research so far is inconclusive, thus no clear conclusion can be drawn about an effective strategy to decrease food intake.
2.2.2 Dinnerware size
Another environmental factor that unknowingly increases individuals’ energy intake and therefore contributes to unhealthy eating habits is dish and cutlery size. Insights on the plate size on paintings of The Last Supper show that plate size has steadily increased over the last millennium (Wansink & Wansink, 2010). Extensive research has been done on the effects of plate size on calorie intake. Wansink and van Ittersum (2013) suggest that people have the tendency to fill their plates until a level of 70% full as this seems an appropriate consumption standard. This could be the reason why a bigger plate size leads to more consumption.
16
A laboratory study of Wansink, Van Ittersum and Painter (2006) evidenced that larger bowls are also associated with bigger self-serving of ice cream, but there was weak grounds to assume that bigger spoons led to bigger self-serving. Several other studies have shown that a bigger size of dinnerware leads to increased food intake (Mishra et al., 2012; Wansink &
Van Ittersum, 2013; Van Ittersum & Wansink, 2012). Research has evidenced that bigger dinnerware size also makes children increase their consumption (Wansink, Van Ittersum &
Payne, 2014). When young children were given a larger bowl they asked nearly twice as much cereal compared to the smaller size and they ate 52% more cereal when presented in a large bowl. Similar results were found in a study with children in a natural setting (DiSantis et al., 2013). It can be concluded that introducing smaller dinnerware in sport canteens is a potential strategy to reduce overeating.
2.3. Price
2.3.1 Price strategy
It is evidenced in various studies that price reduction can be effective in promoting the purchase of healthy foods. French et al. (1997) studied the role of price on the purchase of low-fat options from vending machines in a University. They demonstrated that after reducing the price of low-fat snacks, the proportion of low-fat snacks sales was increased while the total snack sales did not vary. This means that the price reduction shifted the choice to low-fat alternatives without increasing the total snack intake. The results of a similar study in a high school cafeteria are consistent with the before-mentioned study; lower prices for fruit and vegetables, with minimal promotion, increase purchase of these more healthy options among high school students (French et al., 1997).
Horgan and Brownell (2002) compared the effect of health messages, reduced prices and a combination of the two on the purchase of healthy food. The study was done in a restaurant setting over a 14-week period. They found that price reduction on its own might be
17
more effective in increasing healthy food intake, rather than a combination of price reduction and health messages (Horgan & Brownell, 2002). Consistent with previous research, a study conducted in 12 high schools and 12 worksite canteens showed the effectiveness of stimulating healthy consumption by lowering prices (French et al., 2001). They used four pricing levels; normal price, 10% reduction, 25% reduction and 50% reduction, and showed that both adults as well as adolescents purchased more low-fat snacks when the prices where lowered. The higher the price reduction, the higher proportion of low-fat snacks purchased.
The evidence for the effectiveness of price reduction on increase of healthy consumption is shown in various settings and age-categories; worksites, universities and high schools.
Therefore can be assumed that this could also be an effective strategy in sports canteens.
2.3.2 Payment method
In most public consumption places people can choose in which form they pay for their food, using cash, prepaid systems or a type of credit or debit card. Research has shown that the way you pay can influence what food you pick (Just, Wansink, Mancino & Guthrie, 2008).
During the experiment that was done at a University, students could choose between three payment methods; prepaid cards for the whole assortment (unrestricted debit cards), prepaid cards only for healthy products (restricted debit cards) and cash. They discovered that in general people paying with an unrestricted prepaid card ordered less healthy food and consumed more calories than those paying with cash. This effect is probably due to the discount effect; dollars on a prepaid card feel of less value to the people than cash. Thus, paying with cash leads to more cognitive decision-making, resulting in healthier choices.
Besides, there was a difference between food consumption of the restricted and unrestricted prepaid card, individuals paying with the restricted version bought about twice as much healthy food and consumed less calories (Just, Wansink, Mancino & Guthrie, 2008). Putting this knowledge into practice by introducing prepaid cards where you can only buy healthy
18
items with or letting people only pay by cash, can be a way to increase healthy consumption in sports canteens.
2.4. Promotion
2.4.1 Health messages / social marketing
Promotion can also be used to achieve behavior change such as making people eat healthier.
Consider seeing a poster with somebody stressing the importance of eating healthy when entering a sports canteen. There is proof that such health-messages can influence someone’s food choice towards healthier food (Buscher et al., 2001; Gordon et al., 2006; Shive &
Morris, 2006). This is called social marketing, defined as “the application of commercial marketing technologies to the analysis, planning, execution, and evaluation of programs designed to influence the voluntary behavior of target audiences, in order to improve their personal welfare and that of society” (Andreasen 1995, p. 7).
Buscher et al. (2001) used catchy posters to promote healthy options in a school cafeteria setting. The promotion was found to significantly increase sales of 2 out of the 4 products during the intervention. A healthy eating campaign led to similar results in a restaurant setting (Acharya, Patterson, Hill, Schmitz & Bohm, 2006). It had a positive influence on the attitude towards healthy dining and also indirectly increased sales of healthy options. Health messages also seem to work on a younger target group, namely pre-school children (Richards et al., 2009). A brand with the name “Snack Right” was created to improve the attitude, knowledge and behavior towards healthy eating. The attitude of children towards the majority of the key foods (15 out of 24) had positively changed, and the sales of fruit significantly increased. In a review of Carins and Rundle-Thiele (2014) on social marketing, positive change in healthy consumption behavior was evidenced in fourteen out of sixteen studies. It can be concluded that health-messages can be an effective way of changing people’s choices towards healthier ones.
19
2.4.2 Menu strategies
Menu psychology, defined by Wansink and Love (2014, p.3) as “the way in which people perceive, interpret and react to different menu elements”, is a concept that is widely used in restaurants and cafeterias over the world to maximize profit. In other words, changing the way in which the menu is designed can steer people towards certain more profitable choices.
Wansink and Love (2014) conclude in their paper that menu psychology can also be used to realize healthier food intake by 1) shifting attention, 2) improving taste expectations, and 3) enhancing the perception of value.
Shifting attention
Research has shown that attention can be shifted making use of graphics such as contrasting fonts, descriptive names, pictures and colors (Hensdill, 1998 Wansink, Van Ittersum &
Painter, 2001; Zwicky & Zwicky, 1980). Labeling healthy choices as the ‘Chef’s recommendation’, or ‘Specialty of the house’ have shown to increase sales (Wansink, van Ittersum & Painter, 2001). In a field study with 140 adults they showed that these descriptive menu labels might have a positive halo effect to the items. The labels works as a social norm, suggesting those options are a good choice. Furthermore, the placement of healthy options on the menu influences the amount of attention a menu item gets (Yang, Kimes & Sessarego, 2009; Hanks, Just & Wansink, 2012). Items that are placed on top of the menu will get more attention due to the primacy effect, but items listed at the end may benefit because they are processed most recently (Haughvedt & Wegener, 1994; Freund & Kruglanski, 1985).
Improving expectations and perception of the value
When ordering food or drinks from a menu, consumers first form an expectation of the taste of the different items. This process of imagining the taste experience can be quite subjective, based on cues such as attractive or complex names or labels that they associate with existing beliefs (McCall & Lynn, 2008, Wansink, Van Ittersum & Painter, 2005; Wansink et al., 2001;
20
Wansink, Just, Payne & Klinger, 2012). Therefore, using complex and long descriptive names for food items can increase the perceived quality (McCall & Lynn, 2008). A study in a restaurant setting expressive named menu items (such as Satin Chocolate Pudding) led to more positive ratings in terms of taste and appeal than for its normally-named counterparts (Chocolate Pudding) (Wansink, Van Ittersum & Painter, 2005). Similar research is done in the context of healthy food. Wansink et al. (2012) proved that attractive names made the intake of healthy food significantly increase in a primary school. When naming the carrots
““X-ray Vision Carrots,” the children consumed twice as much carrots as without attractive name. Using more attractive names for healthy food seems to be a way to improve expectations in taste and perception of the value and in its turn stimulate healthy eating behaviour.
2.4.3 CAN-Approach
Wansink is a researcher specialized in changing eating behavior and has introduced a framework that makes it easier for consumers to make healthy choices. This framework is called the CAN-Approach, which stands for convenient (easy to pick), attractive (appealing presentation/packaging/names) and normal (make it look like a typical/reasonable option) (Wansink, 2015).
When it comes to food, convenience is often associated with less healthy foods. As convenience is a way to safe effort, in a physical and/or cognitive way, it is attractive to people to choose the convenient option. If you are hungry and you are in a hurry it seems logic to choose the quick convenient option, which most likely is an unhealthy choice. This convenience-seeking behavior can also be used as an opportunity to steer individuals towards healthy choices, by making healthy food more convenient. Research has shown that this is an effective strategy to encourage healthy behavior (Wansink, Just & Payne, 2012; Hank, Just, Smith & Wansink, 2012).
21
Convenience
Changing the presentation of the food is a way to encourage healthy food consumption.
Especially when it comes to children it takes some effort to make them want the healthy option. Using the convenience part of the CAN-approach is one way to do so. Displaying healthier food in the convenience line in a school cafeteria nudges students to eat less unhealthy food, in two weeks after the intervention the sales of the healthy options increased with 77% (Hanks et al., 2012). This is a form of nudging; “to adapt the environment in which consumers make decisions to help them make better choices, without forcing certain outcomes upon them” (Van Kleef, Otten & Trijp, 2012, p.1).
Attractive and Normal
Research has shown that children are 40% more likely to pick the healthy choice, when this option is made more attractive and “fun” (Wansink, Just & Payne, 2012). In their research including 208 boys and girls between 8 and 11 years, they tested the effect of branding on children’s food choices. When putting an Elmo sticker on apples, children almost doubled their choice for an apple instead of a cookie, showing that healthy eating behaviour can be nudged by making the healthy option more attractive (Wansink, Just & Payne, 2012). Similar results were found for fruit in a school environment; when presented in a nice bowl and placed in a part that lit properly the fruit sales grew with 103% (Just & Wansink, 2009). At the same time the “normal” part of the CAN-approach is used in this strategy, fruit bowls make grabbing a piece of fruit seem like a typical choice. A verbal suggestion to children in school cafeterias to take a piece of fruit, makes it feel more normal to eat fruit. Research showed it almost doubled the fruit intake from 40% to 70% (Swartz, 2007). Furthermore, giving healthy food an enticing name makes it more appealing to order healthy food, because of higher taste expectations and the creation of higher awareness (Cardello, 1996; Tuorila, Meiselman,Cardello & Lesher, 1989).
22
Licensed Characters
The use of promotional characters to promote food to children is a marketing strategy that has become common nowadays, also known as character licensing. It has been applied extensively with movie or television show characters such as Spongebob Squarepants, Winnie the Pooh, Teletubbies, Dora and The Minions. A study from Roberto et al. (2010) tested the influence of character licensing on taste preference on fourty children between 4-6 years old.
The children had to try food from packages with a famous cartoon figure on it and foods without. They found that children significantly liked the taste better of the food with the cartoon on the package. Another study with eighty 4-6 year-olds showed similar results;
Children that got confronted with a cereal box with a movie character on it liked the cereal more than those who saw the packaging without a character (Lapierre, Vaala & Linebarger, 2011).
The reasoning behind this phenomenon is the familiarity of the children with the cartoons which gives them a feeling of trust, which in its turn leads to increased liking of the foods with cartoon characters on the packaging (Mizerski, 1995; Garretson & Niedrich, 2004;
Lapierre et al., 2011). This increased liking is manifested in increased attention to the licensed products, an increased probability of selecting the licensed products or even preference in taste of the food with the cartoon packaging (Roberto et al., 2010; Garretson & Niedrich, 2004; Lapierre et al., 2011).
Until very recently these characters were mainly used to endorse unhealthy food such as Happy Meals, candy, crisps and other energy dense nutrition-poor products. With the growing obesity problem in mind, there is a need for a more responsible use of licensed characters to market food products to children. Nevertheless little research has been done so far to confirm if character licensing has a positive effect on the consumption of healthy items.
23
A study of de Droog, Buijzen and Valkenburg (2012) did research on the effect of brand characters (such as the red and yellow M&M cartoon figures of M&M’s packaging) on children’s’ food choice. For this study 166 children between 4-6 years old were included. The authors found that those brand characters triggered an automatic positive affective response towards those licensed products, which in its turn resulted in an increased choice of these healthy products. In other words, they showed that brand character licensing can stimulate healthy eating behavior. No difference was found between familiar and unfamiliar characters in augmenting an automatic effective response (de Droog, Buijzen & Valkenburg, 2012).
Earlier research of the same authors already evidenced that brand characters can increase the purchase intention for fruit up to a level comparable to candy (de Droog, Valkenburg, Buijzen, 2011). A study of Wansink, Just & Payne (2012) discovered that the usage of a familiar character nudged children towards healthier eating behavior. Children between 8 and 11 years old could choose between picking an apple or a cookie. Wansink evidenced that the children chose almost twice as much for the apple when it had an Elmo sticker on it, compared to the control setting. The line of thought is that the reasoning power of young children is not completely developed yet (Reiher & Acuff, 2008). Therefore they often base their product choice on perceptual elements such as familiar characters or colorful packages, instead of having the healthiness of the products as choice motivation.
Product bundling
Another alike strategy has been demonstrated to decrease the calorie intake of children, namely combining two healthy products in a “Happy Meal” kind of box (Wansink & Hanks, 2014). By already putting products containing fewer calories in a box, the default-option is steering to less calorie intake. They concluded that this default combo offer could reduce calorie consumption and improve the overall food intake.This so-called status quo bias - the tendency to stick with the default option – can influence consumer behavior through social
24
norms or effort saving (Samuelson & Zeckhauser, 1988; Hartman, Doane & Woo, 1991;
Madrian & Shea, 2000).
This is the end of the overview of methods to stimulate healthy consumption behavior in canteens categorized by the four P’s of marketing. We decided to continue with the CAN- Approach, standing for Convenient, Attractive and Normal. Based on interactions with the KNVB and Disney we came to the following intervention; we will make healthy options in the soccer canteen more attractive by using a Disney character and more convenient and normal to already bundle the healthier snack option with the healthier drink, so the effort to choose is lower. Besides a “Happy Meal” kind of box will feel more normal to children to buy, because they are already familiar with this concept. The intervention will be further discussed in the following chapter.
3. Hypotheses
The CAN-Approach suggests that nudging children towards healthier options by making it more attractive, convenient and normal for them is an effective strategy (De Droog et al., 2011; De Droog et al., 2012; Wansink, Shimizu & Camps, 2012; Hank et al., 2012; Wansink, 2015). Based on research discussed in the literature review the CAN-Approach will be applied in this study by combining character licensing with product bundling to investigate if this indeed leads children towards healthier choices.
3.1 Collaboration with KNVB and Disney
One of the major media brands, Disney, recently revised its nutrition guidelines. In 2012 it established more strict criteria for all food and beverage products it promotes towards families with younger children, which should be met by 2015 (The Walt Disney Company, 2012).
This means Walt Disney decided to only license their characters for healthy food and
25
beverages that meet these criteria. The KNVB is on the same page when it comes to stimulating healthy eating behavior of children. Since it is the biggest Dutch sports association, where a lot of children eat something in a soccer canteen after the match, they want to take the social responsibility to make the sports canteen a healthier environment.
KNVB and Disney joined hands in helping children to eat healthier. If character licensing works for unhealthy food, it seems presumable that this will also be the case for healthy products.
3.2 The intervention – Character licensed combo box
Based on the existing literature of the CAN-approach on character licensing and product bundling and in close consultation with the KNVB and Disney, the current study will introduce a character licensed healthy version of a “Happy Meal” in sport canteens. A combination of two healthier options (Popcorn and Dubbel Frisss Free, a soft drink without added sugars and sweeteners) will be put into a combo box (product bundling) that will be licensed by one of Disney’s characters, namely Donald Duck. The packaging of the box has a cartoon of Donald duck playing soccer with his three little nephews on it (see Appendix I).
Together with the two healthy options, a Donald Duck key ring will be included as the surprise element of the healthy combo box. Donald Duck is expected to be familiar and attractive to the target group and making the box look like a kind of ‘Happy Meal’ is expected to attract the attention of children.
The healthy options
DubbelFrisss Free is considered a healthy option because it does not contain added sugars nor sweeteners and it only has 16 calories per 100 milliliter, while the average soft drinks easily contain around 35-50 calories/ml. Popcorn is considered a healthy option, due to the fact that it only contains way less calories than a normal bag of crisps (79 calories versus 216 calories)
26
and less fat as well (3.4 grams versus 13.6 grams of fat per bag). For the comparison of the nutrition facts see Appendix II.
Hypotheses
Aforementioned research shows character licensing can lead to an increased liking and preference of the licensed food. Besides, combining healthier food in a box (healthy default- option) can lead to less calorie intake. Based on this knowledge we will examine if the character licensed combo box will stimulate children to make healthier choices when it comes to consumption in sports canteens (see figure 1 for the conceptual framework). Therefore the following hypothesis has been formed:
H1a: The positive attitude towards a healthy combo box licensed by a cartoon character increases the purchase intention of the box (and hence healthier options inside).
H2: The offering of a healthy combo box licensed by a cartoon character leads to an increase of purchase (and hence consumption) of healthy products.
H3: The offering of a healthy combo box licensed by a cartoon character leads to a decrease of purchase (and hence consumption) of unhealthy products.
27
Figure 1: Conceptual Framework
3.3 Moderator - permissive eating
When deciding what snack to pick in a sports canteen, children are often surrounded by their friends and/or family. Influences of family and pears are likely to play a role in their consumption behavior and subsequently their weight (Baranowski, Cullen & Baranowski, 1999). Parents shape their children’s consumption behavior not only by what they serve them for dinner, but also through child-feeding strategies. Some parents let their children decide for themselves what they eat as a snack or even let them decide what they want to eat for dinner, others are more controlling when it comes to the food intake of their kids.
Research found a clear relationship between parental control in the context of food intake and the weight outcome of their children (Birch, McPheee, Shoba, Steinberg &
Krehbiel, 1987; De Bruijn et al., 2007; Johnson, & Birch, 1994). De Bruijn et al. (2007) found that restricting children from consuming unhealthy food led to a decrease of intake of
H1b(+)
28
those products. Zabinski et al. (2006) showed that parental control also has a positive effect on the consumption of healthy food; family support led to an increased intake of fruits and vegetables. Furthermore, restrictive parenting practices were associated to a decreased intake of the restricted products (Gubbels et al., 2009). Cullen et al., 2001 have conducted a study to measure family and peer influences on the intake of fruit, juice and vegetables of children.
One of the scales identified in this study conducting a factor analysis, was the scale for permissive eating; until what extent a child can decide what he/she eats on a day. This scale seemed to significantly influence children’s intake of healthy food (Vereecken, Van Damme
& Maes, 2005).
In the context of this research, children’s food intake in sports canteens, this scale seems very applicable. Some parents give money to their children so they can buy something on their own in the sports canteen, others are more controlling. In addition the ‘theory of planned behavior’ (Azjen, 1985) states that behavior is influenced by perceived behavioral control, together with attitude and subjective norm. It is an extension of the theory of reasoned action (Fishbein and Ajzen, 1975), by incorporating the influence of perceived behavioral control on intentions. This theory states that perceived behavioral control is a determinant of people’s intentions to perform certain behavior; when people believe to be hindered to do something this results in a decreased intention to exert this specific behavior (Azjen, 1985).
Based on literature on parental control and planned behavior is expected that a low score on permissive eating leads to a reduction of the relationship between the attitude towards the box and the purchase intention of the box. In other words; if a child has little permission to decide what to consume, he/she will be less intended to buy the box. The reasoning behind this is that a child that scores low on permissive eating, knows that he/she is restricted to freely choose what to eat and thus perceives less behavioral control to actually
29
buy the box, because he/she in general is more hindered by parents when making decisions on consumption. Therefore the child will also indicate to be less intended to purchasing the box.
Based on this, the current study assumes the following hypothesis:
H1b: A low score on permissive eating reduces the relationship between the attitude towards a character licensed healthy combo box and the purchase intention of healthier options.
30
4.Research Design
This chapter provides the methodology and data collection of this research. To test the hypotheses a field study is conducted in combination with a survey. In this section details of the field study, sampling method, questionnaire and data collection are provided.
4.1 Research method
This study wants to test if the introduction of a character licensed combo box leads to an increase in purchases of healthier food options. In order to examine this a field study is conducted in which this Disney licensed combo box is introduced and sold in an actual soccer canteen in Heino, a town in the Dutch province Overijssel. The soccer club V.V. Heino has around 1000 members and has four playing fields of which two are artificial grass fields. In this club the healthy combo box is added to the assortment and is placed on the shelves next to the other products such as candy and crisps.
Besides the attitude of the respondents towards the box, its content and the purchase intention of the respondents are measured. A survey is conducted among children between 4- 14 years old who are in this specific soccer canteen during two weekends of the intervention.
On the other hand sales data of the key food and drink items of the two weeks before the intervention are compared with the sales data during the two intervention weeks. Note that in all 4 weeks the healthy options ‘popcorn’ and DubbelFrisss Free’ are sold. The first two weeks only as part of the normal assortment and in the two weeks of the intervention the healthy options were part of the normal assortment as well as they formed the content of the healthy combo box introduced. With this data is examined if the purchase of healthier food has increased and if this increase is at the expense of the sales of unhealthy items.
31
4.2 Data collection Survey
In order to collect data for this research a survey was created. The questions and scales applied in this survey are adjusted to the target group, children between 4-14 years old who are visiting a soccer canteen. The survey is filled in by children that have participated in a soccer game as well as children who are there to support or for other reasons. To make the 5- point Likert-scales more understandable for children, smileys are introduced from a very sad face to a very happy face. In total 95 children filled in the survey during the two weekends of the intervention. The older children were able to fill in the survey themselves, the younger ones were provided with help to assure the questions were filled in correctly. The survey is used to measure the dependent variable ‘purchase intention of healthy items’ and the moderator ‘permissive eating’. The survey can be found in Appendix III.
Sales data
The collection of the sales data took place during four weeks in the weekends of April 2, 2016 until April 23, 2016. This collection was done manually, due to the lack of a cash register system in the soccer canteen. During the four weeks of the field study all the relevant products were counted at the beginning and end of the day. The difference was noted to be the sales of these specific products during this particular day.
4.3 Measurements
Independent variable – Attitude towards healthy combo box
The independent variable ‘attitude towards box’ is measured with a 5-point likert scale. The respondents were asked the following; What do you think about the Donald Duck box? They were asked to rate their attitude where 1 means ‘I don’t like it very much’ and 5 means ‘I like it very much’. This question will provide the data for the independent variable ‘attitude box’
in hypothesis 1a and 1b.
32
Dependent variable – Purchase intention
Purchase intention of healthier items is measured with the question; ‘Would you like to buy the Donald Duck combo Box (again)?’ with a 5-point likert scale where 1 means ‘Not at all’
and 5 ‘ Very much’. This will create the dependent variable ‘purchase intention box’.
Moderator – Permissive eating
For the moderator ‘permissive eating’ the scale of Cullen et al., (2001) is slightly adjusted and used in this study. The scale consists out of three statements; ‘My parents let me eat what I want for breakfast’ ‘My parents let me eat what I want as a snack’. The third question is adjusted to measure permissive eating in the soccer canteen; ‘My parents let me eat what I want in the soccer canteen’. These statements are rated on a 5-point likert scale where 1 stands for never and 5 stands for always. Note that a low score indicates that children are more controlled by their parents and a high score means that children are less controlled by their parents. These questions provide data to test hypothesis 1b.
Dependent variables - Actual purchase of healthy and unhealthy products As before mentioned the sales data is measured manually during 4 weekends and serve as the input of the dependent variables ‘Actual purchase of healthier options’ and ‘Actual purchase of unhealthier options’ is measured and form the input to answer hypotheses 3 and 4. The category food consists of cookies, crisps, candy and candy bars where the healthy alternative is popcorn. The drinks category is ordered by amount of calories per 100 milliliter and consists of drinks over 50 kcal/100ml, between 35-50 kcal/100ml and drinks with sweeteners.
There is a separate category for sweeteners as health benefits are still inconclusive (Gardner et al., 2012; Raben et al., 2002). The healthy option for this category is DubbelFrisss Free. If a product is sold it is assumed that this product is also consumed in the soccer canteen.
Therefore consumption will be equal to sales in this study.
33
Control variables
To get a better understanding of the sample, various control variables are included. The age, gender, membership (member of VV Heino/member of other soccer club) and the reason of playing soccer of respondent is controlled for in the data analyses. The variables that turned out to be significant are taken into the analysis as covariates.
4.4 Plan of analysis
First a regression of the control variables with the dependent variable ‘purchase intention box’
is performed to see if one or more of these variables significantly influence the dependent variable. If this is the case they are taken into the subsequent regressions.
In order to test hypotheses 1a a regression analysis is performed. The dependent measure
‘purchase intention box’ is a 5-point likert scale and therefore ordinal, but is treated as an interval variable (Norman, 2010). The independent variable ‘attitude box’ is also a 5-point likert scale that is treated as interval data.
To create the moderator variable a factor analysis is performed on the three statements to check if they measure the same construct. If this is the case they will be merged into one variable, namely ‘permissive eating’. For 1b regression analysis is performed with de independent variable ‘attitude box’, the moderator ‘permissive eating’ and the dependent variable ‘purchase intention box’.
To test H2 and H3 a paired sample t test is executed. This test is applicable, because two groups of similar data are compared in two different points of time. The results indicate whether the healthy combo box significantly led to an increase of purchase of healthy products and a decrease of purchase of unhealthy products.
34
5. Analysis and results
Within this chapter the results of the survey and sales data measurements are shown. Firstly, the descriptive statistics of the sample are presented. Then, the main effects of the regression are shown and will be used for answering hypothesis 1a. The interaction effect will be discussed, which belongs to hypothesis 1b. Finally the sales data will be examined to test hypotheses 2 and 3.
5.1 Preparing the dataset
Before starting the analyses all the surveys are manually inserted in SPSS (the survey can be found in appendix III). The data are checked for outliers and missing values. There are no missing values, all respondents filled in all questions. When a question was not applicable for example in the case that a child did not practice soccer we had the option ‘not applicable’
included. There are also no outliers in the dataset. 5 children filled in 1 or 2 for purchase intention while most of the children filled in 4 or 5. These data points could be seen as outliers, but some children just did not like the combo box and therefore stated they are not very intended to purchase the box.
5.2 Sample
The survey was filled in by 95 children, of which 65 were boys and 30 were girls. The descriptive statistics of the sample can be found in table 1. The participants were between the age of 4 to 14 years old. Of the respondents 36 were member of VV Heino (the club where the experiment was conducted) 45 children were from visiting teams and 14 were not playing soccer at all. In Appendix X photos can be found of the experimental setting.
35
Descriptive statistic Frequency Percentage Mean
Gender N=95
Boy 65 68,4%
Girl 30 31,6%
Age N=95 9,34
4-5 5 5,3%
6-7 18 19%
8-9 23 24,2%
10-11 32 33,7%
12-13 13 13,7%
14 4 4,2%
Weight N=95
Normal 91 95,8%
Expected overweight 4 4,2%
Member of VV Heino N=95
Yes 36 37,9%
No, visiting team 45 47,4%
No, not playing soccer 14 14,7%
Table 1: Descriptive statistics
When it comes to the normal consumption after a match almost 70% of the respondents indicated drinking high calorie and sugary drinks (lemonade, sports drinks and soda) and almost 60% of the children normally eat calorie-dense food (candy, crisps, candy bars), see figures 2 and 3 (for the table with the data see Appendix IV).
Figure 2: Food consumption after sports
