THE EFFECT OF TRAFFIC LIGHT LABELS ON THE
CONSUMPTION OF SUGAR-SWEETENED
BEVERAGES AND THE MODERATING ROLE OF
SELF-CONTROL IN A CANTEEN SETTING
Master Thesis, Msc Marketing Management
University of Groningen, Faculty of Economics and Business. June 17, 2019
MAX VAN POLL Student number: 2782855 Gelkingestraat 13a 9711NA Groningen Tel.: 0683533280 E-mail: m.m.van.poll@student.rug.nl Supervisor
Martine van der Heide Second supervisor
2
ABSTRACT
This research investigates the effect of traffic light labels on the consumption of sugar-sweetened beverages and the moderating role of self-control. A canteen setting was replicated in which participants could choose from a range of beverages. Participants in the intervention condition were presented a red, amber, or green colour label and subsequently, their self-control was assessed. A multinomial logistic regression analysis identified a significant positive effect of traffic light labels on the reduction of sugar consumption in sugar-sweetened beverages. Additionally, this analysis indicated a significant direct effect of self-control, high self-control leads to lower consumption of sugar in sugar-sweetened beverages in canteens. A PROCESS macro analysis indicated a marginal significant interaction effect between traffic light labels and self-control on sugar per beverage. The effect of traffic light labels on sugar consumption is contingent on the degree of self-control individuals possess. This research contributes to existing literature by investigating the moderation effect of traffic light labels and self-control on beverage consumption and highlights the need for more extensive research about traffic light labels and the moderating role of self-control in the food domain.
Keywords: SSB, Sugar, Self-Control, Traffic light labels, nutritional labels, FOP labels,
3 Table of Contents
INTRODUCTION ... 4
THEORETICAL FRAMEWORK ... 8
Nutritional Labelling ... 9
Traffic Light Labels ... 11
Self-Control ... 14 METHODOLOGY ... 19 Method ... 19 Design ... 21 RESULTS ... 26 Chi-Square test ... 26
Multinomial Logistic Regression analysis ... 27
Multicollinearity ... 31
PROCESS Macro ... 31
DISCUSSION ... 34
Public Policy and practical implications ... 39
Limitations and future research ... 41
CONCLUSION ... 43
APPENDIX ... 44
4
INTRODUCTION
Recently, the obesity epidemic has emerged as a growing health problem worldwide (Ludwig, Peterson & Gortmaker, 2001). According to WHO, (2017) over 650 million people aged 18 years and older are obese. In The Netherlands, the percentage of people with severe obesity has tripled over the last 40 years (CBS 2016). Obesity is caused by a positive energy balance, when calories consumed exceed calories expended (van Herpen and Trijp, 2011). Epidemiological studies have indicated that obesity can lead to cardiovascular diseases, diabetes type 2, cancer, and premature death (Wannamethee, 2008). Globally, 2.8 million people are estimated to die from the consequences of obesity per year (WHO 2017).
The consumption of sugar-sweetened beverages (SSB’s) such as lemonade, sports drinks, fruit drinks and sodas have been identified as an important contributor to obesity (Malik, Popkin, Bray, Després & Hu, 2010). Growing evidence encourages limiting the intake of SSB’s (Bleich, Herring, Flagg & Gary-Webb, 2012), (Thorndike, Riis, Sonnenberg & Levy, 2014). This is to improve the energy balance, and because SSB’s contain large amounts of refined sugars with a high glycaemic load that have low satiating capabilities, which has been associated with
excessive weight gain, metabolic syndrome and insulin resistance (Kass, Hecht, Paul & Birnbach, 2014). The American Heart Association has pointed out that SSB’s are the primary source of added sugars in the USA (Johnson et al., 2009).
5 calories. Generally, people underestimate the amount of calories in the food they have consumed (Livingstone and Black 2003). A recent study showed that 90% of participants who had to guess about the calorie amount of starters in a restaurant, underestimated this by an average of more than 600 calories (Burton, Creyer, Kees & Huggins, 2006). As a result individuals tend to overconsume, yet are not aware of this tendency (Chandon and Wansink, 2006).
Additionally, food consumed away from home has been associated with higher consumption of French fries and SSB’s and lower intake of fruit and vegetables (Guthrie, Lin & Frazao, 2002), (Harnack & French, 2008). In the US, over 50% of the food dollar is spent on food away from home such as restaurants and canteens (Saksena et al. 2018). Research has indicated that a change in the food environment that more strongly promotes healthier, less sugary foods and discourages unhealthy, calorie-dense foods, can be an effective way to reduce the obesity epidemic (Swinburn et al., 2011). Despite that many people try to engage in behaviour that promotes healthy eating, the majority finds it difficult to prolong this behaviour (Ogden, Karim, Choudry & Brown, 2006). One way of effectively changing the food environment is making use of food labels, addressing consumer knowledge by providing easily interpretable nutritional information at the point of purchase (Thorndike, Sonnenberg, Riis, Barraclough & Levy, 2012).
6
Figure 1: Traffic Light Labelling
Individuals very strongly associate colours with specific meanings. Traffic light labels were designed to draw upon customer associations between ‘red’ and ‘stop’ and ’green’ and ‘go’ and to minimize cognitive demands (Liu, Wisdom, Roberto, Liu & Ubel, 2013).
7 Metcalfe & Mischel, 1999; Vohs & Baumeister, 2004). To the author’s knowledge,
Koenigstorfer, Groeppel-Klein, and Kamm’s research is the only study that has
considered the moderation effect of self-control on TLL in making healthful decisions. Research by Dholakia et al. (2006) highlighted the need for better understanding of the role of self-control in consumer choices. In spite of this finding, no other research has focused on the moderation effect of self-control in reduction of SSB’s in an out-of-home setting. This study examines the potential of TLL to reduce canteen consumption of SSB’s. Furthermore, it aims at determining whether low- (vs. high-) self-control
individuals make healthier in-store beverage decisions if TLL are in effect. Therefore the following research questions are proposed:
RQ 1: What is the effect of Traffic Light Labels on the consumption of SSB’s in canteens? RQ 2: How is the effect of Traffic Light Labels on the consumption of SSB’s in canteens moderated by self-control?
This research contributes to research on traffic light labelling theory in the domain of eating by identifying the ability of TLL to counteract the tendency of individuals to purchase SSB’s in-store (Vohs, 2006).
In sum, the present research addresses existing gaps in knowledge by investigating the effects of TLL and the moderating role of the consumer characteristic self-control, on consumption
8
THEORETICAL FRAMEWORK
In this research it is proposed that the relationship between TLL and the consumption of SSB’s is moderated by self-control, the ability to change or override impulsive responses and regulate thoughts and behaviour (Carver & Scheier, 1981; Metcalfe & Mischel, 1999; Vohs &
Baumeister, 2004). First, there will be elaborated on the general arguments of TLL in the food environment and thereafter why the effect of TLL might be affected by consumer self-control (figure 2).
9 Nutritional Labelling
In light of the rising obesity epidemic, there has been increased research in the area of nutritional labelling as a measure to counteract unhealthy consumption decisions. Grunert and Wills (2007) suggested that providing customers with health information is a key element in making healthier eating decisions, while maintaining consumers’ freedom of choice.
One way of providing consumers with information is Back of Product (BOP) nutritional
labelling. BOP labelling has been supported by international legislation in the form of the Codex General Standard for the labelling of pre-packaged foods in USA and in EU, although in the EU only products that advertise health claims are required to carry nutritional information
(Aschemann-Witzel et al., 2013).
However, in a recent review of the broader literature it was presupposed that the effect of nutritional information remains weak, inconsistent or unclear (Ikonen, Sotgiu, Aydinli & Verlegh, 2019), (Harnack & French, 2008). The implementation of nutritional labels has had limited success (Bleich et al., 2012) for numerous reasons. First, BOP labels have been proven to be very difficult to interpret and are not easily understood, which reduces its effectiveness
(Easton, Entwistle & Williams, 2010). Furthermore, understanding of health information may also vary between different countries. Recent research that was conducted in Europe showed highly significant differences in understanding of health information between countries (Grunert, Wills, and Fernández-Celemín 2010). The research, that included six different European
10 2012). Research proposes that health information to consumers can be effective in changing consumer behaviour when it is provided in an understandable way (Fagerlin, Zikmund-Fisher & Ubel, 2011).
The limited effect of BOP labels may also be explained by motivation of consumers to make use of the information provided (Burton, Garretson & Velliquette, 1999; Szykman, Bloom & Levy, 1997). Interest in healthy eating more strongly predicts the use of BOP information than nutrition knowledge (Grunert & Wills, 2007). Thus, positive effects of BOP labels may only appear with those who are motivated to make use of the information. Therefore, labelling strategies that are easier to interpret and comprehend may be most effective in promoting healthy choices among individuals (Thorndike et al., 2012).
Front-of-Pack (FOP) nutritional labels have been widely introduced to provide information to individuals regarding the healthfulness of their foods and beverages and to complement
11 Traffic Light Labels
TLL are FOP labels that are implemented to signal the risks and benefits of the consumption of a certain product in an understandable way. TLL help consumers to evaluate the different
ingredients in food or beverages by using colours that summarize nutritional value (Drescher, Roosen and Marette, 2014). Specifically, TLL consist of the colours green, amber, and red on the front of the package to signal that the food or beverage contains low, medium, or high amounts of particular negative nutrients and may therefore be consumed regularly (green), most of the time (amber), or only occasionally (red) (Koenigstorfer et al., 2014).
12 On the contrary, the positive effect of traffic light labelling on healthful food choices has been underscored in a multitude of studies, as well as by health experts (Balcombe, Fraser & Falco, 2010). Customers have been found to have a relatively good comprehension of TLL. This might be because individuals have been accustomed to colours throughout their lives (Balcombe, Fraser & Falco, 2010). The colour red represents danger, caution or prohibition, for instance in traffic, where a red light means ‘stop’ (Moller, Elliot & Maier, 2009) Consumer studies have replicated the priming effect of red. Red decreases consumption of SSB’s and snacks when served on red-coloured (vs. blue red-coloured) cups or plates as was found by Genschow, Reutner and Wanke (2012). Conversely, the colour green is associated with safety and is considered quieting (Fehrman & Fehrman, 2004.). Furthermore, green is less arousing than red. It is the opposite colour to red in traffic light systems (Caivano, 1998). By presenting nutritional information with colours, individuals should be able to make a quick decision at the point of purchase that benefits their health (Balcombe, Fraser & Falco, 2010).
Research has indicated that plain graphics such as TLL are more salient to customers and are better capable of conveying strong health risks (Severtson & Henriques, 2009). Consumer comprehension of TLL is higher than any other nutritional label, so was found by the British Food Standards Agency (Malam, Clegg & Kirwan, 2009). This effect was emphasized by Grunert & Wills, (2007) who showed that 80% of respondents in a traffic light labelling
13 Feunekes, Gortemaker, Willems, Lion & van den Kommer, 2008; (van Herpen & Trijp, 2011); Jones & Richardson, 2007) (McLean, Hoek & Hedderley, 2012). Cowburn and Stockley (2005) found a significant influence of traffic light labelling on altering food purchase behaviour. Campos, Doxey & Hammond (2011) found that the use of nutritional labels is positively linked to healthier diets. This effect was emphasized by Thorndike, Riis, Sonnenberg and Levy (2014) who identified traffic light labelling as a factor contributing to increased sales of healthy items and decreased sales of unhealthy items in a hospital canteen. Additionally, Bollinger, Leslie and Sorensen, (2011) found that traffic light labels reduce the amount of calories purchased in a study conducted at Starbucks. Lastly Franckle, Levy, Macias-Navarro, Rimm & Thorndike, (2018) found evidence that TLL reduced purchases of SSB’s in a local supermarket. The large majority of studies on TLL suggests that TLL will have a significant positive effect on reduction of amount of sugar in SSB’s chosen by the consumer. In conclusion, research on TLL suggests that the labels have a positive effect on consumer behaviour because of its ability to convey
healthiness information in a simple yet strong way to consumers at the point of purchase. Therefore, it is hypothesized that:
H1: implementation of TLL will effectively reduce sugar consumption in SSB’s in canteens.
14 to overemphasize immediate benefits compared to long-term benefits. The characteristic self-control will be considered in depth in the next paragraph.
Self-Control
Self-control is defined in this study as the ability to change or override impulsive responses and regulate thoughts and behaviour (Carver & Scheier, 1981; Metcalfe & Mischel, 1999; Vohs & Baumeister, 2004). Self-control endeavours are aimed at overcoming temptation in favour of long-term goals, such as healthy food decisions (Koenigstorfer et al., 2014). High self-control is needed to consistently respond in an appropriate manner to temptations of unhealthy foods in an individual's environment (Baumeister, Heatherton, & Tice, 1994; Kuhl & Beckmann, 1985; Metcalfe & Mischel, 1999; Mischel, Cantor, & Feldman, 1996; Thaler, 1994; Wegner, 1994). Baumeister describes self-control as a character trait that is a predictor of a multitude of desirable outcomes with an appealing range of benefits to the individual (Fishbach, Friedman &
Kruglanski, 2003).
15 individuals must engage in self-regulatory processes regarding thoughts, emotions, impulsive behaviours and performances. If low self-control individuals want to refrain from constantly indulging in unhealthy foods, they must regulate their personal states and reactions.
The food industry often uses heuristics such as scarcity (e.g., limited edition) as a means to encourage unhealthy food choices (Fennis and Salmon 2014). Multiple studies showed low self-control individuals have a predisposition towards impulsive decision-making strategies, which include over-reliance on salient cues or heuristics (Fennis, Janssen, & Vohs, 2009; Janssen, Fennis, Pruyn, & Vohs, 2008). Consumers often fail to exert self-control, resulting in unhealthful food choices (Vohs, 2006). Individuals that are low in trait self-control may be permanently at risk of being seduced by and indulge in temptations. Individuals high and low in self-control are confronted with these temptations on a daily basis (Friese and Hofmann 2009). However,
whereas high self-control individuals are capable of controlling and overriding the urge to give in to these temptations, low in trait self-control individuals often fail to resist them (Baumeister & Heatherton, 1996).
To be able to regulate their reaction, low self-control individuals (Baumeister, 2002) depend imperatively on information. Research has underscored the ability of TLL to convey nutritional information fast and effective (Thorndike, et al., 2012). As discussed before, people low in self-control rely more on such fast and simple cues. Therefore, low self-self-control individuals are expected to benefit from choosing foods within a certain category (colour); this tool aids them to control their consumption behaviour (Koenigstorfer et al., 2014). Traffic light colours are
16 aware of their long-term goals (Koenigstorfer et al., 2014). Thus, TLL are expected to be more effective when a customer possesses low self-control.
Conversely, high self-control individuals are better capable of dealing with self-control
dilemma’s (Koenigstorfer et al., 2014). They already possess decision strategies that encourage healthful food decisions and do not have an implicit attraction to unhealthful foods (Fishbach, Friedman & Kruglanski, 2003). Moreover, low self-control individuals may be more susceptible to promotions accentuating immediate gratification, whereas individuals with high self-control are more likely to consume based on long-term values (Fishbach, Friedman & Kruglanski, 2003). As a result, coloured-coded priming will not have a strong effect on the reduction of consumption of SSB’s of high self-control individuals at the point of purchase.
Therefore, it is hypothesized that:
H2: TLL will have a more pronounced effect on the reduction of sugar consumption in SSB’s of
low self-control individuals than it will on individuals with high self-control.
17 Haws, 2009; Vohs, 2006). Because temptations are goal states that are intrinsically pleasurable, but obstruct over-arching long-term goals, they may naturally invoke self-control dilemma’s (Baumeister 2002). Previous studies have indicated that individuals use a variety of strategies to refrain from giving in to temptations.
One way individuals might do so is by keeping physical distance to temptations and proximity to items related to their long-term goals (Ainslie, 1992; Schelling, 1984; Thaler & Shefrin, 1981). People may vary in their ability to do so. Specifically, research suggests that high self-control individuals are more likely to approach goals and avoid temptations. Conversely, low self-control individuals are more likely to indulge in temptations in favour of their long-term goals (Fishbach and Shah 2006). Previous research highlighted that when presented with the options between a chocolate cake and a fruit salad, a low self-control individual is less likely to choose a less
appealing option with longer term benefits (fruit salad) in favour of a more immediately attractive but ultimately less beneficial option of chocolate cake (Baumeister et al., 1998; Muraven et al., 1998; van Dellen and Hoyle, 2010).
Therefore, it is hypothesized that:
H3: High self-control has a direct positive influence on reduction of sugar consumption in SSB’s
18 In this research there shall be clearly distinguished between low self-control and high self-control individuals, and difference in response to in-store temptations carrying a traffic light label will be assessed. This assessment may be important as a possible predictor of success to reduce obesity by approaching healthful foods and avoiding items strongly associated to obesity.
19
METHODOLOGY
In a 2x2 between-subject experiment, the moderation effect of low vs. high self-control
individuals on the consumption of SSB’s carrying TLL on the front of the beverages was tested. The study is built upon data from an online survey including a subsequent experiment focused on beverage choices (Drescher, Roosen and Marette, 2014). The study imitated a point of purchase canteen situation.
Method
Analysis of the survey returns begins by examining various socio-economic descriptive statistics (Balcombe, Fraser & Falco, 2010). A total of 159 consumers who were recruited through
20
Table 1 - Descriptives of the research population
Variable Mean SD Age 25.44 8.59 Nutritional knowledge consideration 4.43 1.67 Self-control 4.09 1.31
Gender Frequency Percentage
Male 52 38
Female 85 62
Education
High School 9 6.6
Secondary education 2 1.5
University of applied sciences 26 19
Bachelor’s degree 50 36,5
Master’s degree 45 32.8
Doctorate 4 2.9
21 Design
We applied a one-way factorial (Traffic light labelling yes [n =69] vs. no [n = 68] between-participants design. The moderator consumer self-control was included as measured variable. Specifically, self-control was measured on four items, using a seven-point rating scale where 1 = “strongly disagree” (low self-control) and 7 = “strongly agree” (high self-control) (Koenigstorfer et al., 2014).
A traffic light labelling intervention in a canteen was performed to see whether TLL influenced beverage choice. The beverages that participants could choose represented the average canteen selection of beverages in Holland. Participants were randomly assigned to the traffic light label / no traffic light label condition. Beverages were categorized as red, amber or green based on the criteria used for the Boston Public Health Commission’s ‘Rethink your drink’ campaign as was done in the study of Koenigstorfer, Groeppel-Klein, and Kamm (2014). Participants in the traffic light label condition were presented with beverages containing labels that showed either a ‘green’ ‘amber’ or ‘red’ light. Participants in the no labelling condition were presented with beverages without carrying a TLL. The beverage colours were determined based on the amount of sugar they contained. The main aim of the intervention was to identify the beverage choices in the labelling vs. no labelling condition.
22 & Thorndike, 2018). Artificially sweetened beverages were labelled green. Back of the product labels remained unchanged. Each virtual beverage was given a traffic light label to categorize it as healthy (green), less healthy (yellow) or an unhealthy (red) choice through computer artwork at a consistent position on the product (figure 2). The nutrition information was obtained from the product’s BOP label and was used to classify the beverages. No further description/explanations were given of the label to prevent participants from guessing the purpose of the study. The beverages were presented on the computer screen in a single row and in random order without any price tags.
23 Participants were recruited through Facebook and partook in the experiment and the subsequent questionnaire at any chosen time of their preference. Participants were informed that the study aimed at researching consumer behaviour in a canteen environment (Lee & Thompson, 2016). In addition, the explanation given to the participants afterwards was that this was done to see how consumers orient when shopping. No consumer was informed about the real goal of the study, in order to avoid priming and socially desirable responses. Before consumers started the
experiment, we obtained informed consent about their participation from them. Participants were randomly assigned to the manipulated condition of beverages carrying a traffic light label or to the control condition. After being assigned to a label condition, participants were directed to a page that displayed a hypothetical menu. Participants first received the following instructions; ‘’Imagine you are at a canteen and are ready to order a sandwich, a sweet snack, a savoury snack, and beverage for yourself. Please select what you would order from the menu on the next page based on the information provided’’ (Aschemann-Witzel et al., 2013). Participants could choose between a range of beverages, chocolate bars and savoury snacks, as these items are comparable between canteens in Holland. The items were randomly ordered and presented in a single row. The first three product categories are products not related to the research and are designed to bring customers in the mood of shopping. This was done to most adequately simulate a real-life shopping environment. The fourth product participants had to choose was a beverage.
Participants had the choice of 9 different beverages of which 3 were labelled as green, 3 as orange, and 3 as red. After the experiment participants were asked to fill in a survey about the moderator self-control, and thereafter, were debriefed.
24 tend to indulge more than I should,” and “I often wish I could get myself to avoid food
indulgences more often.” on a seven-point rating scale where 1 = “strongly disagree” (low self-control) and 7 = “strongly agree” (high self-control (Koenigstorfer et al., 2014). The last two items were reverse coded and a reliable overall self-control index was computed (α = .84) (table 2). Thereafter, customers in the manipulation group were asked to respond to the following two statements by answering either always, usually, rarely, or never: “I consider nutrition information when making a decision about which foods/beverages to eat/drink” and “I choose food that is healthy.” and if the response was “yes,” they were asked the follow-up question: “Did the labels influence your purchase today?’’ (Thorndike et al., 2012). Respondents in the label condition were also questioned about the familiarity with the nutritional labels (one-item, five-point rating scale; 1 = “very unfamiliar,” and 5 = “very familiar”) Koenigstorfer et al., 2014.
Table 2 - Reliability analysis of self-control
Cronbach’s Alpha Number of items
25
Figure 4: Survey beverage choice in the no label condition (above) vs. labelling condition
26
RESULTS
Chi-Square test
27
Figure 5: Percentage of beverage choices in the no label vs. labelling condition.
Multinomial Logistic Regression analysis
To test our main hypothesis, we ran a Multinomial Logistic Regression analysis. Included was a variable for the nutrition labelling condition (beverage without traffic light colours = -1, beverage with traffic light colours =1), perceived self-control (both mean-centered) and the interaction between the two as independent variables. Beverage choice colour was used (red, amber, green) as the dependent variable (Koenigstorfer et al., 2014). The model was significant as a whole (p = 0.044). Moreover, Pearson and Deviance Chi-square statistics were performed to assess goodness of fit of the model (table 4). Both tests were not significant, respectively. (p = 0.271 and p = 0.162). This implies that the observed distribution of the dependent variable, traffic light colours
0,00% 2,00% 4,00% 6,00% 8,00% 10,00% 12,00% 14,00% 16,00% 18,00% Spa Water Coca Col a Zero Cryst al Cl ea r Ice Tea Gr een Du bb el frisss Ac iv e O 2 Coca Col a R eg ul ar Fa nt a O ra ng e AA D ri nk
From left to right: Green beverages vs Orange beverages vs Red Beverages
Beverage choices
28 is not significantly different from the predicted hypothesis. Thus, the model is capable of
predicting the hypothesis. Furthermore, as can be seen in table 4 and 5, the analysis indicated a significant main effect for nutritional labelling on beverage choice (B= 1.875, p = 0.012), such that people presented with TLL compared to people not presented with TLL are 1.875 times less likely to go for a ‘red’ beverage relative to a ‘’green’’ beverageholding all other independent variables remain constant. Thus, this result supports hypothesis 1. For no other traffic light
colours a main effect was found to be significant. For example, people presented with traffic light labels are not significantly more likely to go for an ‘amber’ drink compared to a ‘green’ drink. Furthermore, there was no significant interaction between self-control and nutrition labelling (B = -.315, p = 0.210 > .05). Hypothesis 2 is therefore rejected. However, a significant direct effect of self-control on beverage choice was found as is also pointed out in table 5 (B= -.560, p= 0.026). To clarify, this indicates that as self-control increases with one unit, beverage choice for ‘red beverages’ decreases with .560 holding all other independent variables remain constant. This provides partial support for h3. Thus without considering traffic light labels, higher self-control reduces ‘red’ labelled beverage choices.
Table 4 – Goodness-of-Fit Test Multinomial Logistic Regression analysis
Goodness-of-Fit (p)
Pearson .271
29
Table 5 – Multinomial Logistic Regression (1/2)
Traffic Light B (p)
1 Condition No Label 1.875 .012*
Condition With Label . .
Self-control -.560 .026*
Interaction Label and Self-Control -.315 .210 2 No Label .593 .173 With label . . Self-control -.194 .257
Interaction Label and Self-Control
.011 .947
30
Table 6 – Multinomial Logistic Regression (2/2)
Traffic Light B (p)
2 Condition No Label 1.283 .107
Condition With Label . .
Self-control .366 .177
Interaction Label and Self-Control .327 .228 3 No Label -1.875 .012* With label . . Self-control .560 .026*
Interaction Label and Self-Control
.315 .210
Note: Reference category is 1.0. Significance level: * = significant at p < 0.05 Traffic light 1 = ‘red’ 2 = ‘amber’ 3 = ‘green’
31 self-control on beverage choice was also once more confirmed (B= -.709, p = 0.013) . Moreover, none of the control variables proved to have a significant influence on participants’ beverage choice in a canteen setting.
Multicollinearity
Precedently, decisions of consumers were analysed by considering the effect of traffic light labels on beverage choices and their corresponding colours. This analysis was replicated with sugar per beverage as dependent variable. A linear regression analysis with dependent variable sugar per beverage was ran to check for multicollinearity of the control variables. Multicollinearity happens when two or more independent variables within a model are highly correlated. Results for
multicollinearity resulted in VIF scores ranging from 1 to 1.77, indicating no concerns about multicollinearity.
PROCESS Macro
Furthermore, the moderation effect of self-control on the main effect of labelling condition on sugar per beverage was examined using Model 1 of the PROCESS Macro (Hayes 2013)
(appendix 3). The label condition, and the moderator self-control were mean-centred. The model as a whole was significant (𝑅² = .1931, F(9,90), p = 0.000). In support of h1, a significant main effect of the labelling condition on grams of sugar per beverage was found (B = -1.84. SE = .36 t = -5.1, p = 0.000). Participants in the labelling condition choose less sugary drinks than
33
Figure 6: Labelling condition by self-control on amount of sugar (g) per beverage choice.
Figure 7: Moderation effect of self-control on labelling vs. no labelling condition on sugar per
beverage.
0
1
2
3
4
5
6
7
8
9
10
- 1 SD
M
+1 SD
S
u
gar
p
er
b
eve
rage
(g)
Self-control
No label condition
Labelling condition
0 1 2 3 4 5 6 7Low self-control Medium self-control High self-control
Sugar
per
bever
age (g)
34
DISCUSSION
The last chapter of this study will provide a discussion of the outcomes of the analyses conducted, and formulate public policy and practical implications. Limitations and future research suggestions will be discussed and lastly a conclusion will be provided.
35 findings in this study are considered to provide support for the effectiveness of TLL as to
decrease consumption of highly sugary beverages.
For consumers who reported to indulge more often in food temptations than they desire and have trouble ignoring short-term rewards, captured as ‘’low self-control’’ a higher consumption of high sugary beverage is ascertained in comparison to high self-control consumers. Furthermore, a moderation effect was identified for self-control on the effect of traffic light labels on sugar per beverage. This research adds to the existing body of literature by identifying the traffic light labels strength to be contingent on consumers’ self-control.
Additionally, the effects were analysed across two dependent variables, to conduct the research as extensively as possible. Out of 3 hypotheses one was fully supported by the analyses conducted. Hypothesis 2 was significant, however the effect was opposite to what was hypothesized. Hypothesis 3 was partially supported, a significant effect was found for one of two dependent variables. An overview of the hypotheses and their results can be found in table 7.
Table 7 - An overview of the hypotheses and their results
Determinant Statistical Significance Coefficient estimate sign Hypothesis confirmed Beverage colour
H1: implementation of TLL will effectively
reduce sugar consumption in SSB’s in canteens
Significant positive ü
H2: TLL will have a more pronounced effect on
the reduction of sugar consumption of SSB’s of
36
Significance level: * = significant at p < 0.05
In accordance with earlier research, TLL significantly reduced consumption of sugary beverages both measured in grams per beverage and in beverage choice colour. This may be explained by the labels’ ability to provide automatic guidance in shopping environments(Dijksterhuis et al. 2005). The effect remained significant when controlling for age, education, gender, consideration
low self-control individuals than it will on individuals with high self-control
H3: High self-control has a direct positive
influence on reduction of sugar consumption in SSB’s in canteens
Significant negative Partially confirmed
Sugar per beverage
H1: implementation of TLL will effectively
reduce sugar consumption in SSB’s in canteens
Significant negative ü
H2: TLL will have a more pronounced effect on
the reduction of consumption of SSB’s of low self-control individuals than it will on
individuals with high self-control
Significant negative X
H3: High self-control has a direct positive
influence on reduction of SSB’s in canteens
37 of nutritional information and degree of hunger. The ‘red’ label showed to have a much stronger effect in comparison to the ‘amber’ coloured label. This indicates that TLL guide customers from ‘red beverages’ towards healthful ‘green’ beverages, over ‘semi’ healthful, ‘amber’ labelled beverages. This may be explained by participants’ interpretation of the ‘red’ label as ‘prohibited’ or ‘dangerous’ rather than a sign to limit the consumption of the beverage. ‘Amber’ labelled beverages may presumably confuse consumers, or not convey a strong enough message to achieve a statistical significant effect. The outcome of ‘red’ labelled beverages can be called desirable from a consumer welfare point of view, since indulging in too many red coloured
beverages may lead to obesity diabetes type 2 and cardiovascular diseases (Wannamethee, 2008). However, it remains unclear how consumers interpret the traffic light and it may be assumed from the results that consumers only make a distinction between the colour ’red’ and ’stop’ and ’green’ and ‘go’ rather than interpreting the different colours and their nutritional implications.
Previous research has related healthful decision making in response to nutrition information, to consumers motivation for healthful eating, and their ability to use nutrition information (Grunert, Wills, and Fernández-Celemín 2010; Moorman 1990, 1996; Van Herpen and Van Trijp 2011). However, these researches do not consider consumer ability to refrain from indulging in
38 labels is available, previous research has pointed at the ability of traffic light labels to stand out to individuals low in self-control at the point of purchase (Koenigstorfer et al). Additionally,
research on self-control highlights the ability of high self-control individuals to retract themselves from situations in which they are tempted to make a decision that is counterproductive for their long term wellbeing (Fishbach, Friedman & Kruglanski, 2003). Therefore the results are
surprising. An explanation of the contradicting effect may lie within the relatively small research group (138). There is a possibility that a part of the participants classified as high in self-control, had no conflicting values with regards to healthful eating vs. indulgence (Hassan et al., 2010). For example; if their goal was to gain weight, choosing a beverage that was high in sugar would have made an optimal decision for their health. A high self-control individuals’ indulgence is actually a healthful decision here. This could have resulted in the effect as seen in the current research. The non-significant interaction effect between TLL and self-control on beverage choice colour may be explained because changes in consumption behaviour have not reached the
magnitude to make participants reconsider their ‘red’ labelled beverage decision notably. Therefore it does reduce grams of sugar per beverage in both the dependent variables, but not enough to achieve statistical significance by falling within the parameters that are set for different traffic light colours in this research.
39 self-control individuals are expected to override the urge of being tempted by (food) temptations (Baumeister 2002, Hassan et al., 2010). An argument for this insignificant effect may be that participants were not provided with sufficient alternatives for the unhealthy options. Presenting consumers with a larger set of options has been proven an effective way of guiding them towards healthier food choices (Bucher et al., 2016). Another reason may be that one way for individuals to cope with temptations is to increase the physical distance between one-self and the
temptations. This may be especially true for in-store decisions in which the physical distance to satisfy short-term goals is small and unhealthful food is even more tempting (Geyskens et al. 2008). In our hypothetical canteen situation however, increasing physical distance was not an option which may be an explanation for this result.
Public Policy and practical implications
40 This study is one of the first to examine the role of self-control on the effect of TLL in a canteen setting. It was found that participants with more self-control (in the labelling condition) choose less sugary beverages when presented with TLL compared to those not presented with a TLL. In addition, it was identified in this study that as self-control increases in the labelling condition, consumption of sugar per beverage decreases. Opposed to hypothesis 2, this indicates that
especially high control individuals benefit from TLL relative to people who have lower self-control. However the effect is more pronounced for high self-control individuals, and public policy makers may advise implementation of TLL in places where low self-control individuals are especially tempted, such as canteens or cafeterias (Thorndike, Riis, Sonnenberg & Levy, 2014). As low self-control individuals have been proven not being able to resist temptations (Baumeister, 2002; Hassan et al., 2010). Here, traffic light colours make conflicts more remarkable to consumers at the point of purchase by priming them to stop vs. to go (Janssen, Fennis, Pruyn, & Vohs, 2008).
To optimize label efficiency, public policy makers may recommend keeping the format (place, size) of the labels constant (Bialkova and Van Trijp 2010). Also, implementation of TLL may also be extended to other product categories as well, to maximize their benefit. Food
41 Limitations and future research
This study has several limitations. This study was designed to assess the effect of TLL on in-store beverage decisions through imitating a canteen shopping experience. As with all stated
preference and methods studies, it is in this research not possible to rule out that any beverage decisions made in this study differ from consumers’ beverage preferences at the point of purchase in real life. Hypothetical consumer behaviour in this study is therefore not certain to be replicated in actual shopping decisions (Aschemann-Witzel et al., 2013). Additionally, in this study the majority of the participants were students between the ages of 20 to 30 years old. Most of the participants came from high social economic backgrounds, with a high education level. Future research may be aimed at conducting a similar study with a lower distribution of education level, as high literacy skills have been proven to enhance the ability to interpret nutritional information (Thorndike et al., 2014). Concerning the stimuli, 9 different beverages of which 3 were
considered ‘healthy’ were used. The study did not consider expanding the set of healthful beverages so that participants had more options to choose from. Expanding the choice of healthful beverages may underscore the idea of nudging participants towards more healthful choices, through changes in the environment (Thaler & Sunstein, 2008). Furthermore, options of predominantly green-labelled foods may encourage healthful food decisions and reduce
overconsumption. Research has already indicated that this is the case for low-fat product claims (Wansink and Chandon 2006).
Also, the effect of health information is difficult to measure. Like many other studies examining the effect of traffic light labels, results in this study have relied on self-reported outcome
42 consumers say to use nutritional information at the point of purchase (61% worldwide according to Grunert, Wills & Fernández-Celemín) (2010). Only a limited amount of research has measured the influence of health information on actual purchases in a real-life environment (Seymour 2004). According to Grunert, Wills, and Fernández-Celemín (2010) the actual percentage that uses nutritional information is much lower. They claim only 27% of consumers in supermarkets make use of nutritional information at the point of purchase. Thus, results may be biased
(Grunert, Wills & Fernández-Celemín, 2010), (van 't Riet, 2012). Future research may be conducted in real life canteen situations, with a more thorough assessment of the use of healthiness information.
Finally, this research did not find a consistent hypothesized moderating effect of self-control on the labelling condition and beverage choice colour. One reason to explain this may be that although self-control contributes to a significant reduction of sugar in beverage choice, the reduction effect may not have the magnitude to fall within the parameters of a different traffic light colour. Further research may be dedicated to the specific mechanism that leads consumers with low self-control to make more healthful decisions when presented with TLL. More attention may be directed to how the traffic light colours are interpreted by consumers at the point of purchase or if they only acknowledge the ’red’ colour on the traffic light.
This research did however find a moderating effect of self-control on the labelling condition and sugar per beverage. Surprisingly, this effect contrasted previous findings by
43 decisions. Future research may make a more specific assessment of individuals food goals, and when they are indulging in a food temptation that they cannot resist.
CONCLUSION
In line with previous research this study has identified a positive effect of traffic light labels on the reduction of sugar consumption in SSB’s in canteens, across two dependent variables. The positive effect of the label condition on beverage choice colour was found while controlling for age, education, gender, hunger, and consideration of nutritional information. This is to prove that TLL have a positive influence on consumers’ healthful decisions in a canteen situation. Also, a positive relationship between self-control and the consumption of SSB’s in canteens was found with beverage choice colour as dependent variable. Indicating that people with higher self-control make more healthful food decisions. Furthermore, a marginal significant interaction effect
44 consumers possess and underscore the need for more knowledge on the moderating role of self-control in the food domain.
APPENDIX
Appendix 1 – Chi-Square test
45
Total Count 19 31 78 137
Within condition
14,8 24,2 60.9 100
Appendix 2 Multinomial Logistic Regression
Traffic Light Colour
Exp. (B) (p)
1 Condition No Label 5.031 .042*
Condition With Label . .
Self-control .492 .013*
Interaction Label Self-control .440 .440
Age 1.011 .772
Consider Nutrition Info .923 .648
Degree of Hungriness 1.218 .354
Education 1.622 .128
Gender .357 .128
2 Condition No Label 2.140 .136
46
Self-control .763 .185
Interaction Label Self-control .996 .985
Age 1.013 .647
Consider Nutrition Info .874 .374
Degree of Hungriness 1.164 .380
Education 1.307 .276
Gender 1.028 .961
Note: Reference category is 3.0. Significance level: * = significant at p < 0.05 Traffic Light Colour 1 = red 2 = amber 3 = green
Appendix 3 Process Macro output
47 Interaction -.5138 .2799 -1.8355 .0688* Constant 3.2198 .3604 8.9339 .0000
49
REFERENCES
Aschemann-Witzel, J., Grunert, K., van Trijp, H., Bialkova, S., Raats, M., & Hodgkins, C. et al. (2013). Effects of nutrition label format and product assortment on the healthfulness of food choice. Appetite, 71, 63-74. doi: 10.1016/j.appet.2013.07.004
50 Balcombe, K., Fraser, I., & Falco, S. (2010). Traffic lights and food choice: A choice experiment
examining the relationship between nutritional food labels and price. Food Policy, 35(3), 211-220. doi: 10.1016/j.foodpol.2009.12.005
Baumeister, Roy F. (2002), “Yielding to Temptation: Self-Control Failure, Impulsive Purchasing, and Consumer Behavior,” Journal of Consumer Research, 28 (4), 670–76
Bleich, S., Herring, B., Flagg, D., & Gary-Webb, T. (2012). Reduction in Purchases of Sugar-Sweetened Beverages Among Low-Income Black Adolescents After Exposure to Caloric Information. American Journal Of Public Health, 102(2), 329-335. doi: 10.2105/ajph.2011.300350
Borgmeier, I., & Westenhoefer, J. (2009). Impact of different food label formats on healthiness evaluation and food choice of consumers: a randomized-controlled study. BMC Public Health, 9(1). doi:
10.1186/1471-2458-9-184
Burton, S., Garretson, J., & Velliquette, A. (1999). Implications of Accurate Usage of Nutrition Facts Panel Information for Food Product Evaluations and Purchase Intentions. Journal Of The Academy Of Marketing Science, 27(4), 470-480. doi: 10.1177/0092070399274006
Burton, S., Creyer, E., Kees, J., & Huggins, K. (2006). Attacking the Obesity Epidemic: The Potential Health Benefits of Providing Nutrition Information in Restaurants. American Journal Of Public Health, 96(9), 1669-1675. doi: 10.2105/ajph.2004.054973
51 Caivano, J. (1998). Color and semiotics: A two-way street. Color Research & Application, 23(6), 390-401. doi: 10.1002/(sici)1520-6378(199812)23:6<390::aid-col7>3.0.co;2-#
Campos, S., Doxey, J., & Hammond, D. (2011). Nutrition labels on pre-packaged foods: a systematic review. Public Health Nutrition, 14(08), 1496-1506. doi: 10.1017/s1368980010003290
Chen, L., Appel, L., Loria, C., Lin, P., Champagne, C., & Elmer, P. et al. (2009). Reduction in consumption of sugar-sweetened beverages is associated with weight loss: the PREMIER trial. The American Journal Of Clinical Nutrition, 89(5), 1299-1306. doi: 10.3945/ajcn.2008.27240
Cowburn, G., & Stockley, L. (2005). Consumer understanding and use of nutrition labelling: a systematic review. Public Health Nutrition, 8(01). doi: 10.1079/phn2004666
Chou, S., Grossman, M., & Saffer, H. (2004). An economic analysis of adult obesity: results from the Behavioral Risk Factor Surveillance System. Journal Of Health Economics, 23(3), 565-587. doi: 10.1016/j.jhealeco.2003.10.003
Dijksterhuis, A., Smith, P., van Baaren, R. and Wigboldus, D. (2005). The Unconscious Consumer: Effects of Environment on Consumer Behavior. Journal of Consumer Psychology, 15(3), pp.193-202.
Drescher, L., Roosen, J., & Marette, S. (2014). The effects of traffic light labels and involvement on consumer choices for food and financial products. International Journal Of Consumer Studies, 38(3), 217-227. doi: 10.1111/ijcs.12086
52 Easton, P., Entwistle, V., & Williams, B. (2010). Health in the 'hidden population' of people with low literacy. A systematic review of the literature. BMC Public Health, 10(1). doi: 10.1186/1471-2458-10-459
Fagerlin, A., Zikmund-Fisher, B., & Ubel, P. (2011). Helping Patients Decide: Ten Steps to Better Risk Communication. JNCI Journal Of The National Cancer Institute, 103(19), 1436-1443. doi:
10.1093/jnci/djr318
Fehrman, C., & Fehrman, K. Color: The Secret Influence (2nd ed.). Upper Saddle River: NJ: Prentice Hall.
Feunekes, G., Gortemaker, I., Willems, A., Lion, R., & van den Kommer, M. (2008). Front-of-pack nutrition labelling: Testing effectiveness of different nutrition labelling formats front-of-pack in four European countries. Appetite, 50(1), 57-70. doi: 10.1016/j.appet.2007.05.009
Fishbach, A., Friedman, R., & Kruglanski, A. (2003). Leading us not into temptation: Momentary allurements elicit overriding goal activation. Journal Of Personality And Social Psychology, 84(2), 296-309. doi: 10.1037/0022-3514.84.2.296
Franckle, R., Levy, D., Macias-Navarro, L., Rimm, E., & Thorndike, A. (2018). Traffic-light labels and financial incentives to reduce sugar-sweetened beverage purchases by low-income Latino families: a randomized controlled trial. Public Health Nutrition, 21(08), 1426-1434. doi:
10.1017/s1368980018000319
53 French, S., Story, M., Neumark-Sztainer, D., Fulkerson, J., & Hannan, P. (2001). Fast food restaurant use among adolescents: associations with nutrient intake, food choices and behavioral and psychosocial variables. International Journal Of Obesity, 25(12), 1823-1833. doi: 10.1038/sj.ijo.0801820
Genschow, O., Reutner, L., & Wänke, M. (2012). The colour red reduces snack food and soft drink intake. Appetite, 58(2), 699-702. doi: 10.1016/j.appet.2011.12.023
Geyskens, K., Dewitte, S., Pandelaere, M. and Warlop, L. (2008). Tempt Me Just a Little Bit More: The Effect of Prior Food Temptation Actionability on Goal Activation and Consumption. Journal of Consumer Research, 35(4), pp.600-610.
Glanz, K., M. Basil, E. Maibach, J. Goldberg, and D. Snyder. 1998. Why Americans Eat What They Do: Taste, Nutrition, Cost, Convenience, and Weight Control Concerns as Influences on Food Consumption. Journal of the American Dietetic Association 98 (10): 1118–1126.
Grunert, K., & Wills, J. (2007). A review of European research on consumer response to nutrition information on food labels. Journal Of Public Health, 15(5), 385-399. doi: 10.1007/s10389-007-0101-9
Grunert, K., Wills, J., & Fernández-Celemín, L. (2010). Nutrition knowledge, and use and understanding of nutrition information on food labels among consumers in the UK. Appetite, 55(2), 177-189. doi: 10.1016/j.appet.2010.05.045
54 Hassan, L., Shiu, E., & Michaelidou, N. (2010). The Influence of Nutrition Information on Choice: The Roles of Temptation, Conflict and Self-Control. Journal Of Consumer Affairs, 44(3), 499-515. doi: 10.1111/j.1745-6606.2010.01180.x
Harnack, L., & French, S. (2008). Effect of point-of-purchase calorie labeling on restaurant and cafeteria food choices: A review of the literature. International Journal Of Behavioral Nutrition And Physical Activity, 5(1), 51. doi: 10.1186/1479-5868-5-51
Johnson, R., Appel, L., Brands, M., Howard, B., Lefevre, M., & Lustig, R. et al. (2009). Dietary Sugars Intake and Cardiovascular Health. Circulation, 120(11), 1011-1020. doi:
10.1161/circulationaha.109.192627
Jones, G., & Richardson, M. (2007). An objective examination of consumer perception of nutrition information based on healthiness ratings and eye movements. Public Health Nutrition, 10(03). doi: 10.1017/s1368980007258513
Kant, A., & Graubard, B. (2004). Eating out in America, 1987–2000: trends and nutritional correlates. Preventive Medicine, 38(2), 243-249. doi: 10.1016/j.ypmed.2003.10.004
Kass, N., Hecht, K., Paul, A., & Birnbach, K. (2014). Ethics and Obesity Prevention: Ethical
Considerations in 3 Approaches to Reducing Consumption of Sugar-Sweetened Beverages. American Journal Of Public Health, 104(5), 787-795. doi: 10.2105/ajph.2013.301708
55 Koenigstorfer, J., Groeppel-Klein, A., & Kamm, F. (2014). Healthful Food Decision Making in Response to Traffic Light Color-Coded Nutrition Labeling. Journal Of Public Policy & Marketing, 33(1), 65-77. doi: 10.1509/jppm.12.091
Kristal, A., Levy, L., Patterson, R., Li, S., & White, E. (1998). Trends in food label use associated with new nutrition labeling regulations. American Journal Of Public Health, 88(8), 1212-1215. doi: 10.2105/ajph.88.8.1212
Lee, M., & Thompson, J. (2016). Exploring enhanced menu labels’ influence on fast food selections and exercise-related attitudes, perceptions, and intentions. Appetite, 105, 416-422. doi:
10.1016/j.appet.2016.06.007
Levy, A., & Fein, S. (1998). Consumers’ Ability to Perform Tasks Using Nutrition Labels. Journal Of Nutrition Education, 30(4), 210-217. doi: 10.1016/s0022-3182(98)70321-8
Liu, P., Wisdom, J., Roberto, C., Liu, L., & Ubel, P. (2013). Using Behavioral Economics to Design More Effective Food Policies to Address Obesity. Applied Economic Perspectives And Policy, 36(1), 6-24. doi: 10.1093/aepp/ppt027
56 Luszczynska, A., Scholz, U., & Schwarzer, R. (2005). The General Self-Efficacy Scale: Multicultural Validation Studies. The Journal Of Psychology, 139(5), 439-457. doi: 10.3200/jrlp.139.5.439-457
Malam, S., Clegg, S., & Kirwan, S. (2009). Comprehension and use of UK nutrition signpost labelling schemes. Social Research, FSA
Malik, V., Popkin, B., Bray, G., Després, J., & Hu, F. (2010). Sugar-Sweetened Beverages, Obesity, Type 2 Diabetes Mellitus, and Cardiovascular Disease Risk. Circulation, 121(11), 1356-1364. doi:
10.1161/circulationaha.109.876185
McLean, R., Hoek, J., & Hedderley, D. (2012). Effects of alternative label formats on choice of high- and low-sodium products in a New Zealand population sample. Public Health Nutrition, 15(05), 783-791. doi: 10.1017/s1368980011003508
Mattes, R. (1996). Dietary Compensation by Humans for Supplemental Energy Provided as Ethanol or Carbohydrate in Fluids. Physiology & Behavior, 59(1), 179-187. doi: 10.1016/0031-9384(95)02007-1
Metcalfe, J. and Mischel, W. (1999). A hot/cool-system analysis of delay of gratification: Dynamics of willpower. Psychological Review, 106(1), pp.3-19.
Moller, A., Elliot, A., & Maier, M. (2009). Basic hue-meaning associations. Emotion, 9(6), 898-902. doi: 10.1037/a0017811
O'Donoghue, T., & Rabin, M. (2000). The economics of immediate gratification. Journal Of Behavioral Decision Making, 13(2), 233-250. doi:
57 Ogden, J., Karim, L., Choudry, A., & Brown, K. (2006). Understanding successful behaviour change: the role of intentions, attitudes to the target and motivations and the example of diet. Health Education Research, 22(3), 397-405. doi: 10.1093/her/cyl090
Rick, S., & Loewenstein, G. (2008). Intangibility in Intertemporal Choice. SSRN Electronic Journal. doi: 10.2139/ssrn.1082056
Rothman RL, Housam R, Weiss H, Davis D, Gregory R, Gebretsadik T, Shintani A, Elasy TA. Patient Understanding of Food Labels: The Role of Literacy and Numeracy. Am J Prev Med 2006; 31(5): 391-398. (2015). American Journal Of Preventive Medicine, 49(2), 332. doi: 10.1016/j.amepre.2015.04.010
Saksena, M., Okrent, A., Anekwe, T., Cho, C., & Hyman, J. (2018). America’s Eating Habits:Food Away From Home. United States Department of Agriculture (USDA).
Sacks, G., Rayner, M., & Swinburn, B. (2009). Impact of front-of-pack 'traffic-light' nutrition labelling on consumer food purchases in the UK. Health Promotion International, 24(4), 344-352. doi:
10.1093/heapro/dap032
Severtson, D., & Henriques, J. (2009). The Effect of Graphics on Environmental Health Risk Beliefs, Emotions, Behavioral Intentions, and Recall. Risk Analysis, 29(11), 1549-1565. doi: 10.1111/j.1539-6924.2009.01299.x
58 Swinburn, B., Sacks, G., Hall, K., McPherson, K., Finegood, D., Moodie, M., & Gortmaker, S. (2011). The global obesity pandemic: shaped by global drivers and local environments. The Lancet, 378(9793), 804-814. doi: 10.1016/s0140-6736(11)60813-1
Szykman, L., Bloom, P., & Levy, A. (1997). A Proposed Model of the Use of Package Claims and Nutrition Labels. Journal Of Public Policy & Marketing, 16(2), 228-241. doi: 10.1177/074391569701600204
Tangney, J., Baumeister, R., & Boone, A. (2004). High Self-Control Predicts Good Adjustment, Less Pathology, Better Grades, and Interpersonal Success. Journal Of Personality, 72(2), 271-324. doi: 10.1111/j.0022-3506.2004.00263.x
Nudge: improving decisions about health, wealth, and happiness. (2008). Choice Reviews Online, 46(02), pp.46-0977-46-0977.
Thorndike, A., Riis, J., Sonnenberg, L., & Levy, D. (2014). Traffic-Light Labels and Choice Architecture. American Journal Of Preventive Medicine, 46(2), 143-149. doi: 10.1016/j.amepre.2013.10.002
Thorndike, A., Sonnenberg, L., Riis, J., Barraclough, S., & Levy, D. (2012). A 2-Phase Labeling and Choice Architecture Intervention to Improve Healthy Food and Beverage Choices. American Journal Of Public Health, 102(3), 527-533. doi: 10.2105/ajph.2011.300391
Tice, D., Baumeister, R. and Heatherthon, T. (1995). Losing control: how and why people fail at self-regulation. Choice Reviews Online, 32(11), pp.32-6508-32-6508.
59 van Herpen, E., & Trijp, H. (2011). Front-of-pack nutrition labels. Their effect on attention and choices when consumers have varying goals and time constraints. Appetite, 57(1), 148-160. doi:
10.1016/j.appet.2011.04.011
van 't Riet, J. (2012). Sales effects of product health information at points of purchase: a systematic review. Public Health Nutrition, 16(03), 418-429. doi: 10.1017/s1368980012001103
Vohs, K. (2006). Self-Regulatory Resources Power the Reflective System: Evidence From Five Domains. Journal Of Consumer Psychology, 16(3), 217-223. doi: 10.1207/s15327663jcp1603_3
Wannamethee, S. (2008). Obesity Epidemiology. Frank B Hu. International Journal Of Epidemiology, 38(1), 325-326. doi: 10.1093/ije/dyn227
Wegner, D. (1994). Ironic processes of mental control. Psychological Review, 101(1), pp.34-52.
Wiesenfeld, B., Reyt, J., Brockner, J., & Trope, Y. (2017). Construal Level Theory in Organizational Research. Annual Review Of Organizational Psychology And Organizational Behavior, 4(1), 367-400. doi: 10.1146/annurev-orgpsych-032516-113115
