Master thesis Psychology, Economic and Consumer Psychology Institute of Psychology
Faculty of Social and Behavioral Sciences – Leiden University Date: February 28th, 2019
Student number: s1506862
First examiner of the university: Hilmar Zech Second examiner of the university: Lotte van Dillen
Approach and avoidance
tendencies towards food and
color
The effect of hue, saturation and value on approach
and avoidance tendencies towards healthy and
unhealthy food.
Louise Dillies
Abstract
Rohr et al. (2015) found that the color red facilitated avoidance tendencies towards unhealthy foods in a traffic light context (compared to green). This experiment aimed to verify and extend these findings on the effect of hue, saturation and value in the approach or avoidance tendency towards healthy and unhealthy food items in and out of the traffic light context. One hundred and forty-five participants completed a mobile approach-avoidance task (AAT) in which they were presented with healthy or unhealthy food items surrounded by a colored circle that was either of red, green or blue hue and earthy or popping in shade (or saturation and value). They were asked to approach or avoid these food items according to the instruction given and reaction time was measured. No effect of hue, saturation and value was found on approach or avoidance behavior towards the healthy and unhealthy foods. However, a general approach tendency towards healthy food was found and an approach tendency was found to be stronger towards food items surrounded by earthy blue or green and popping red compared to popping blue or green and earthy red regardless of healthiness of food.
Approach and Avoidance Tendencies towards Food and Color
When going through the aisles of a supermarket, we rely on several sensory cues to make a decision on what we choose to purchase (e.g.: color, package design, labels, etc.). We use these cues to make decisions that are in line with our personal motivations. These goals could be, for example, whether we want to take care of our bodies by buying healthy food or indulge in something tasty by choosing the unhealthy option. These goals can be conscious ones or unconscious.
The Approach Tendency
The act of approaching a certain stimulus faster than when avoiding it without conscious control, like an impulse, is called an approach tendency (Klein, Becker, & Rinck, 2011). This can occur in many situations. For example, Kakoschke, Kemps and Tiggeman (2017) pointed out in their research on approach tendencies that people have an automatic tendency to reach out toward rather than away from appetitive cues in their environment (e.g.: a piece of cake). Approach tendencies are also especially salient in addictive behaviors such as with alcohol or drugs (Eberl, Wiers, Pawelczack, Rinck, Becker, & Lindenmeyer, 2013; Wiers, Eberl, Rinck, Becker, & Lindenmeyer, 2011). People who are addicted to these substances will have an approach tendency towards them that can be very hard to control. Watson, de Wit, Cousijn, Hommel and Wiers (2012) state that researchers have found that people with drug approach tendencies were therefore more likely to engage in drug use.
Rohr, Kamm, Koenigstorfer, Groeppel-Klein and Wentura (2015) carried out a study to research the effect of the color red and green on approach and avoidance
tendencies towards healthy and unhealthy food. This was done by measuring the reaction time for participants to categorize the pictures as healthy or unhealthy by pushing or pulling the images away from or towards them. They studied the effect of color on this movement by varying the color of a circle surrounding these food items. Red and green were the two colors they focused on but they also ran a control study on pink and blue. Measuring people’s reaction times allows researchers to determine the unconscious or innate associations people may have with healthy or unhealthy foods and how color may have an effect on this; the approach or avoidance tendency.
One of the initial findings of this study was that participants had an approach tendency towards healthy food items. Added to this, Rohr et al. (2015) also found that red led participants to have an avoidance tendency towards unhealthy foods. However, there was no effect of green on approach tendencies towards healthy foods. This study was the most complete study on this topic and therefore we chose to replicate it. However, we found some of Rohr et al.’s (2015) findings unintuitive and decided to extend the study in order to determine which combination of colors (red or green but also adding blue) led to approach and avoidance tendencies. Additionally, we want to take into account shade, which was not manipulated in Rohr et al.’s (2015) study, to determine whether this also plays a role in people’s approach or avoidance tendencies towards healthy or unhealthy food.
Rohr et al.’s (2015) study firstly found evidence that people have an approach tendency towards healthy foods. However, there is an existing disagreement on people’s general approach tendency towards healthy and unhealthy food. On one hand, obesity rates are climbing in America and it has been found that Americans find healthy food to be less tasty and less indulgent (Devaux, Graf, Goryakin, Cecchini, Huber, & Colombo, 2017). Americans were found to choose food for its taste and indulgence rather than for its healthiness (Turnwald, Jurafsky, Conner, & Crum, 2017). Raghunathan, Naylor, and Hoyer (2018) named this occurrence the unhealthy – tasty intuition and stated that the consumption of unhealthy foods is more exciting and more of a hedonic experience than that of healthy foods.
On the other hand, there is the increasing trend of healthy eating, vegetarianism and veganism. This has led to an increasing presence of superfoods (e.g.: chia, kale, acai berries, etc.) in supermarkets and restaurant. Additionally, Gillebaart, Kroese, & Ridder (2014) suggested that people who are successful at self-control would approach healthy foods quicker compared to unhealthy foods. It is necessary to discern these two different views as it is important information marketers could use when trying to improve the sale of healthy and unhealthy foods. As Rohr et al. (2015) found evidence that there may be an approach tendency towards healthy foods, we predict that participants will have an approach tendency towards healthy food items independently of the effect of color and shade.
H1: People will be more likely to have an approach tendency towards healthy food items compared to unhealthy food items.
As Rohr et al. (2015) found that participants had an avoidance tendency towards unhealthy foods surrounded by a red-colored circle, this leads us to believe color may play a role on our approach or avoidance tendencies towards healthy and unhealthy foods. Consumers rely on sensory cues in packaging such as design, labels, word choice and colors. Packaging is what essentially appeals to the consumer emotionally and what communicate the brand’s identity (Baik, & Suk, 2010). Our research will also concentrate on the role of colors and how their use in packaging may affect approach or avoidance tendencies towards healthy and unhealthy foods. Colors communicate specific
information outside of conscious awareness: these associations can be culturally learned or grounded in biology or both (Genschow, Reutner, & Wanke, 2012). For example, red is often linked with anger, blood and danger in nature and for animals, therefore it is an innate association we may have with that color that has existed for centuries
(Pravossoudovitch, Cury, Young, & Elliot, 2014).
Previous research on color has led to different theories about what people associate certain colors with and therefore their approach or avoidance tendencies towards that color. For example, green is a color that has been uniformly accepted as having positive associations. It is a color often associated with growth, fertility and nature (Lichtenfeld et al., 2012). Additionally, green is associated with success and achievement (Moller, Elliot, & Maier, 2009). In Rohr et al.’s (2015) study, however, no effect of the color green on approach tendencies towards healthy food items was found. Knowing the inborn associations people have with the color green, it could have been expected that Rohr et al. (2015) would have found that presenting green alongside healthy food items might have led to approach tendencies towards healthy foods. However this was not the
case, green did not intensify approach tendencies towards healthy foods (Rohr et al., 2015).
The color red is argued to have negative associations such as being a signal for danger, prohibition, warning or fear (Genschow, Reutner, & Wänke, 2011; Kliger, & Gilad, 2012; Mehta, & Zhu, 2009). This has been demonstrated in experiments which showed that drinking from red cups, eating from red plates or exhibiting a TV with a red background all led to less consumption and more avoidance tendencies of the products compared to other colors such as green (Bruno, Martani, Corsini, & Oleari, 2013; Bellizzi, & Hite, 1992; Genschow, Reutner, & Wänke, 2012). However, other research has found that red, in certain cultures, signifies joy and fortune such as in China (Kliger, & Gilad, 2012). It can also be a sign of sexual attraction; Young women were rated as more sexually desirable when they were wearing red (Lichtenfeld, Elliot, Maier, &
Pekrun, 2012). Goldstein’s (1942) warm vs. cold theory states that warm colors (e.g.: red) stimulate, focus people on the outward environment and lead people to produce forceful movements (Bellizzi, & Hite, 1992). Supporting this, Danger (1969, as cited in Bellizzi, Crowley, & Hasty, 1983) stated that warm colors attract consumers’ attention. Red, specifically, was found to elicit exultant moods, emotional arousal, excitement and increased blood pressure (Schaie, & Heiss, 1964; Bellizzi, Crowley, & Hasty, 1983; Nakshian, 1964). All of these are more positive associations people may have with red. Overall, Valdez and Mehrabian (1994) summarize the contradiction about the color red and its associations because it is an “exciting” and “stimulating” color, which could have positive and negative connotations.
Overall, when looking at the literature on the colors red and green we can conclude that red’s connotations depend on the situation but that green has an overall positive connotation. Rohr et al.’s (2015) research looked into this red-green distinction as they introduced the traffic-light theory. This theory states that due to the repeated exposure to red and green associated together in traffic lights, we experience these colors as we would in traffic. This means that we would associate red as a sign to stop and green as a sign to go. This is eventually reflected in our approach and avoidance tendencies, where stop in a consumption context would mean not buying a certain product and where go would mean buying the product (Rohr et al, 2015). This can then be interpreted as stop being the act of avoiding and go the act of approaching (Rohr et al., 2015). Taking into account the traffic light context and Rohr et al.’s (2015) findings, we also expect that people will be more likely to avoid unhealthy foods when associated with the color red. H2: People will be more likely to have an avoidance tendency towards unhealthy food
stimuli surrounded with a red circle (high saturation and high value) compared to a green circle (high saturation and high value).
Removing the Traffic Light Context
Due to red’s positive-negative debate and Rohr et al.’s (2015) finding that red leads to avoidance tendencies towards unhealthy foods only in the traffic-light context (when red is presented alongside green), we decided to determine whether it was because of the traffic-light context that the color red led to avoidance tendencies towards
unhealthy food. Therefore, we expect that manipulating the context by replacing green with blue, in order to eliminate the traffic-light context, may lead us to understand whether red is linked to an avoidance tendency only in the traffic-light context.
We chose to use blue as a replacement of green in order to take the color red out of the traffic-light context. Blue is a cold color and is associated with feeling
“secure/comfortable” and “tender/soothing” (Goldstein, 1942; Valdez, & Mehrabian, 1994). This indicates an approach tendency in a consumption-context. We decided to keep red so as to decipher whether it led to more avoidance tendencies towards unhealthy foods or if this was due to the fact that it was in the traffic-light context when presented alongside green.
H3: People will be more likely to have an approach tendency towards unhealthy food surrounded by a red circle (high saturation and high value) when taken out of the traffic-light context, compared to food surrounded by a blue circle (high saturation and high value).
Manipulating shade
Finally, a drawback Rohr et al. (2015) had in their research is that they only took into account hue, which is the attributes of a color that make it discernible from other colors, and left out the manipulation of saturation and value (Valdez, & Mehrabian, 1994). Hue refers to Saturation refers to the purity of the color whereas value refers to the lightness or darkness of the color (Adamson, 2018). Both of these combined can be referred to the shade of the color. This is something that Rohr et al. (2015) has not manipulated in their study as they used colors that were rather bright and high in
saturation. Guilford and Smith (1959) found that brighter and more saturated colors led to more pleasure. This could have an important implication in packaging color research especially on how colors are associated to healthiness of food. Research has shown that paler, less saturated shades of green and beige were symbolic of nature, well-being and
conveyed a feeling of warmth (Baik, & Suk, 2010). These colors were found to be present on organic food packaging and Baik and Suk (2010) referred to them as earthy tones of color. If we were to manipulate colors’ shade, this could allow us to study which shades of color are more congruent with our idea of healthy and unhealthy food. In order to obtain the earthy shades of red, green and blue we will manipulate the shade with the aim of examining whether this has an effect on approach or avoidance tendencies towards healthy and unhealthy foods. We will compare these to popping shade (high-saturation and value) of the same colors. We expect that earthy shades of the three colors will be more congruent with participants’ concept of healthy food and therefore help them approach healthy food compared to popping shades.
H4: People will be more likely to have an approach tendency towards healthy food surrounded by earthy colors (low saturation and low value) compared stimuli surrounded by popping colors (high saturation and high value).
Mobile Approach Avoidance Task
The way we will measure these approach and avoidance tendencies will be by using the mobile Approach Avoidance Task. This task measures the amount of time it takes for participants to pull (approach) or push (avoid) a stimulus between when the stimulus is presented and the beginning of their movement (Klein, Becker, & Rinck, 2011). This shows whether there is a conflict or congruency between the goal given to the participant (push or pull) and their innate reaction to the stimulus; their innate impulse. The mobile AAT works with the underlying assumption that the bigger the conflict, the longer the reaction time (Kakoschke, Kemps, & Tiggemann, 2017). An approach
when avoiding these stimuli (Eberl, Wiers, Pawelczack, Rinck, Becker, & Lindenmeyer, 2013).
We will be using the mobile version of this task. The mobile version allows the participants to replicate the action they would engage in when shopping for food in the supermarket. In other words, participants pulling the phone towards themselves would equate picking up a product from the shelf. Pushing the phone away from themselves would equate putting the product away when it does not appeal to them. This would allow us to add possible significant findings that are more in-line with actual
consumption contexts compared to the traditional computer and joy-stick AAT, which was used in Rohr et al.’s (2015) study. This latter AAT consists of using a joy-stick to execute the approach and avoidance movements and therefore is less representative of the real-life movements customers may execute in a supermarket.
Method Participants
149 participants were recruited from Prolific, an online scientific recruitment platform (86 females and 59 males and 4 unknown due to a technical error; Mage= 29.22 years, where 7 participants’ age remained unknown due to technical error). These participants had respond to the following criteria: originate from the European Union, speak fluent English, own an Android phone or tablet, have normal or corrected-to-normal vision and no color-vision deficiencies. We recruited only European participants as Rohr et al.’s (2015) participants were from Germany and therefore we were interested in extending their findings to countries that may have similar associations with the colors used in the experiment. The participants were rewarded £1 for their participation.
Research design
The design of the study was a two (food type: healthy vs. unhealthy) by two (color: red vs. not red) by two (shade: earthy vs. popping) repeated measures ANOVA with traffic-light condition as a between-subjects variable and approach tendency as a dependent variable. This was in the form of an approach-avoidance tendency (subtracting the reaction time of approach movements from the reaction time of avoidance movements for each stimulus category). The independent variables were food type (healthy vs.
unhealthy), color (red vs. not-red) and shade (popping vs. earthy). Each participant completed the mobile approach-avoidance task and rated the stimuli on healthiness. Participants were randomly assigned to one of the two conditions but both conditions had the same design making it a balanced design.
Materials and Instructions
Mobile AAT. The mobile approach-avoidance task (AAT) was used in this study.
This task consists of showing stimuli along with an instruction to push or pull the stimuli and by then measuring the participants’ reaction time to carry out the movement,
researchers can obtain an approach tendency towards that particular stimulus. Participants are instructed to position the phone or tablet at a starting position (Figure 1) and to return to it after each movement. The stimulus is presented in the center of the screen and when it appears, participants have to move the phone or tablet as straight as possible until the stimulus disappears from the screen. However, if reaction times are longer than 2,000 ms participants were informed that time had run out. The mobile AAT is equipped with sensors that record the reaction time of a participant to initiate a pushing or pulling movement from the moment the stimulus appears on the phone. Therefore, the reaction
times of participants’ movements are directly recorded on the phone or tablet. Using this mobile version of the AAT allows the reproduction of movements that consumers carry out in a shopping related setting.
Figure 1. Starting position, approach position and avoid position of the participant during the experiment.
Stimuli. In this experiment, we used the same stimuli set as Rohr et al. (2015).
These consisted of 20 black-and-white drawings of healthy food items and unhealthy food items (Appendix B). These were pretested by Rohr et al. (2015) on students so that each stimulus was clearly classifiable as healthy or unhealthy.
In the practice trials, the food item stimuli were presented on a white background like in the main experimental trials but the food items were surrounded by a grey circle. These were not the same stimuli shown in the experimental trials as to eliminate carry-over effects. The circles surrounding the food items had an inner diameter of 79mm and a thickness of 13mm. In the experimental trials, the circles surrounding the food item stimuli were either red or green (for the traffic-light condition) or red or blue (for the non-traffic light condition). We then manipulated the saturation and value as to make the colors a popping or earthy shade. The popping colors of these circles were based on the ones used in Rohr et al.’s (2015) study. We manipulated saturation and value for the
earthy colors (see Appendix A for example of stimuli and hue, saturation and value settings).
Procedure
Once having signed up for the study on the Prolific website, participants were asked to download the mobile AAT to their Android mobile phone or tablet. Instructions were given on the screen and participants were informed that food items would be presented to them and that they would have to push or pull them towards each other according to the instruction given to them at the beginning of each trial. Additionally, participants were informed there would be a colored circle surrounding the food items but that these were not important and could be ignored. Before starting the AAT, they were asked to fill out an informed consent (Appendix C).
The whole AAT had a total of 160 randomized experimental trials divided into two blocks of 80 trials. In the first block, participants had instructions to either push or pull healthy food or unhealthy food and the instructions were reversed in the second block. These instructions were randomly assigned to each participant. Each block
contained 40 trials for healthy food items and 40 trials for unhealthy food items. Of these 40 trials, 20 were surrounded by the color red (10 being a popping shade and 10 being an earthy shade) and the other 20 trials were non-red (i.e., green in traffic-light or blue in non-traffic light condition). Each set of 40 trials was preceded by 6 practice trials. In these practice trials, participants were given feedback on the movements they performed according to the instructions given with “correct” or “incorrect” as feedback. There was no feedback in the main trials.
After completing the 160 trials, participants were asked to provide a rating for each of the 20 food item drawings used on a 5-point scale ranging from very unhealthy to very healthy. Participants were finally debriefed (Appendix D) and paid.
Results
This study aimed to investigate which colors (red, green and blue) combined with which shade (high or low saturation and value) would lead participants to approach or avoid healthy or unhealthy food items. From these, approach-avoidance tendencies were calculated. This means that positive scores represent the participants’ quicker approach movements and a negative score a quicker avoidance movements. However, it is important to note that approach and avoidance tendencies should not be interpreted in absolute terms because it is possible that there is an effect of movement that makes one movement easier to initiate due to the ease of its motor components (Rohr et al., 2015).
Main Effects
We carried out at two (food type: healthy vs. unhealthy) by two (color: red vs. not red) by two (shade: earthy vs. popping) repeated measures ANOVA with traffic-light condition as a between-subjects variable and the approach tendency as a dependent variable. The main effect of food type yielded as significant, F(1, 148) = 33.39, p < .001, where participants’ approach tendency was quicker for healthy food stimuli compared to unhealthy food stimuli (MH = 48.81 vs. MUH = -10.21). This finding indicates that people
have a general tendency to approach healthy foods compared to unhealthy foods. This supports our first hypothesis that predicted that participants would have an approach tendency towards healthy foods. The other main effects were both insignificant: Shade, F(1, 148) = .244, p = .244; Color, F(1, 148) = 0.15, p = .697. This indicates that shade
and color are not significant predictors for approach or avoidance reaction times in the mobile AAT.
Figure 2. Mean approach tendency of foodtype (healthy vs. unhealthy) in ms.
Hypothesized results
We also hypothesized that participants would have an avoidance tendency towards unhealthy foods surrounded by a red circle of high saturation and value. In order to verify this second hypothesis, we looked at the two-way interaction between Food type and Color in the traffic light context, F(1, 72) = 0.06, p = .809. This interaction suggests there is no significant difference in approach tendency towards healthy or unhealthy foods when presented with a red or green circle around it. This finding does not confirm our second hypothesis.
Our third hypothesis predicted that participants would be more likely to approach unhealthy food surrounded by red when taken out of the traffic light context. For this, we analyzed the two-way interaction between Food type and Color in the non-traffic light context, F(1, 75) = 0.02, p = .898. This interaction was not significant suggesting that participants’ approach tendencies towards healthy or unhealthy foods did not differ
according to whether the food items were presented with a red or a blue circle around them. Therefore, our third hypothesis was not supported.
Finally, we looked at the two-way interaction of Food type and Shade in order to verify our fourth hypothesis: that participants would have a greater approach tendency towards healthy foods surrounded by an earthy shade rather than a popping shade. This interaction did not yield significant results, F(1, 148) = 0.01 =, p = .934. This, therefore, does not support our final hypothesis but rather indicates that the earthy or popping shade was not a significant predictor for reaction time towards healthy foods within or outside of the Traffic-Light context.
Explorative results
All the other interactions of this study were not significant with F < 1.5 except for two two-way interactions that were either marginally significant or significant. The interaction between Food type and Traffic Light condition tended towards significance (F[1, 148] = 3.45, p = .062). This may indicate that there may be a trend towards
participants having a greater approach tendency towards healthy food items compared to unhealthy food items surrounded by a green circle but even more when surrounded by a blue circle (non-traffic light condition; MHealthyTrafficLight = 43.87, MHealthyNonTrafficLight = 53.76 vs. MUnhealthyTrafficLight = 4.06, MUnhealthyNonTrafficLight = -24.47). The Shade and Color interaction was significant, F(1, 148) = 4.57, p = .034. As this interaction does not take into account the Traffic Light condition, the colors can only be interpreted as red or non-red; in other words, as warm (red) or cold (green or blue) colors. In this case, the findings show that there is a greater approach tendency towards stimuli surrounded by a non-red circle that is an earthy shade than a red circle of an earthy shade (MNon-RedEarthy = 26.48 vs.
MRedEarthy = 17.75). However, for the popping shade, there was a greated approach tendency towards stimuli surrounded by a red circle than a non-red circle (MRedPopping = 22.90 vs. MNon-RedPopping = 10.09). Overall, food items surrounded by an earthy non-red circle led to the greatest approach tendency.
Figure 3. Two-way interaction between Shade (earthy andpopping) and Color (red and not red).
Healthiness questionnaire
In order to ensure that the stimuli Rohr et al. (2015) used was still classifiable as healthy or unhealthy by our participants, we administered a questionnaire at the end of the game. Participants had to rate all the healthy stimuli as very unhealthy (a 1 on the scale) or very healthy (a 5 on the scale). By calculating the average ratings for each participant on each stimuli, the results indicate that the stimuli were clearly classifiable as healthy or unhealthy.
Figure 4. Average rating on the 5-point scale (1 = very unhealthy and 5 = very healthy) for each food type stimuli.
Discussion
This study aimed to verify Rohr et al.’s (2015) findings on approach and
avoidance tendencies towards healthy and unhealthy foods and how these can be affected
by the colors red and green in a food packaging context. Rohr et al. (2015) had found that people approached healthy foods more than unhealthy foods that red led to more
avoidance tendencies towards unhealthy foods compared to green. In addition to
verifying Rohr’s findings, this study aimed to extend these findings to a larger population (Europeans instead of only German participants) and to look at whether manipulating the traffic-light context (red vs. blue as opposed to red vs. green) and the shade of the colors could affect participants’ approach or avoidance tendencies towards the food items. Firstly, in line with Rohr, we predicted that participants would approach healthy food items more than unhealthy food items. Secondly, we predicted that participants, in a traffic-light context, would avoid unhealthy foods surrounded by the color red. Thirdly, we predicted that taking away the traffic-light context, by replacing green with blue,
would lead participants to approach unhealthy food items surrounded by red more than in the traffic-light context. Finally, we predicted that manipulating the shade to an earthy shade would lead to more approach tendencies towards healthy food items compared to popping colors. We tested this by administering a mobile AAT to participants which measured their reactions times to approach (pull) or avoid (push) food items as instructed to.
In line with our prediction, participants approached healthy food items more than unhealthy food items. This supports Rohr et al.’s (2015) findings in their research as well as the overall rise of conscience in healthy eating. Contrary to our predictions, we found that participants did not differ significantly on approach tendencies between healthy and unhealthy food items presented with red or green circles (high saturation and high value). Therefore, this did not support the claims that the colors red and green could have an effect on participants’ perception of the food items according to their healthiness. When we took away the Traffic Light context, participants still did not differ significantly on approach-avoidance tendencies between healthy and unhealthy food items encircled by red or blue circles, either. Overall, these two results may indicate that colors (regardless of being in the Traffic light context or not) do not have an association with participants’ approach or avoidance towards the food items. Even though there is evidence that blue was found to lead to more consumption of food and drinks when served in blue plates or cups compared to red in a study by Bruno and colleagues (2012). Finally, when
manipulating the shade of the colored circles, participants still did not differ significantly between healthy and unhealthy food approach-avoidance tendencies on earthy or popping versions of the green, red and blue.
Despite not being able to support our main hypotheses, we found that overall, participants tended to have a quicker approach tendency towards healthy food items compared to unhealthy food items. This happened independent of the effect of color. This is interesting as it defends one of the two current opposing views, whether people have an approach tendency towards healthy food or unhealthy foods. We believe people
approaching healthy food more is due to the increasing amount of awareness people have towards what they eat and how it affects their health. This finding is also in line with Rohr and colleagues’ (2015) finding when they carried out their study.
Another finding that was not related to a hypothesis was that participants
approached earthy cool colors and popping red more, independently of whether red was presented alongside green or blue (traffic light context). Although this did not yield information for which type of food (healthy and unhealthy) this finding pertained to, it still provides some useful knowledge to keep in mind when designing food packages. Consumers do have certain preferences in hue, saturation and value used in relation to food as was found by Guilford and Smith (1959) and further demonstrated by Baik and Suk (2010) who found earthy, paler shades to be more prevalent in healthy food
packaging.
Limitations and Future Research
One main reason why the participants’ approach-avoidance tendencies did not differ significantly when manipulating the Traffic-Light context or the shades of color (earthy or popping) may be due to the lack of trials. In this experiment, participants had 10 trials for each combination of food type, hue, saturation and value. For optimal results, this should have been 20 trials for each combination. However, this would have been too
taxing for the participants as they have to hold the phone up during the entire length of the experiment, and it could have created lack of interest and involvement. For future research, researchers should administer 20 trials for each combination. This would increase the statistical power of the experiment and could lead to more reliable results.
There were also some drawbacks with the material used. Firstly, the stimuli represented drawings of food and not photographs of the food. This drawback could have made the food items less recognizable or tangible to the participants. Additionally, using the colors green, blue and red meant that some of these colors displayed with the food item were congruent (for example, red with a strawberry) and some were incongruent (blue with a banana). This may have led participants to have innate approach tendencies towards congruent food-color item stimuli and avoidance tendencies towards incongruent food-color item stimuli. We advise future researchers to think about replicating this study using photographs of real food items and their packaging and manipulating the colors of these directly. Additionally, future research could look into testing a greater range of colors (both cool and warm) in order to give marketers a larger variety of colors to choose from when designing packaging.
Using the mobile AAT and an online recruitment platform like Prolific led to some additional drawbacks. As participants carried out the experiment from their own personal devices, we were unable to monitor whether the participants were not faced with distractions or whether they carried out the task seriously. For example, they could have held the phone the wrong way, not sideways as instructed to do so (see Figure 1). It would be favorable to carry out this experiment in a more controlled setting such as a lab to control for this.
Finally, the mobile app itself had limitations as it detects false movement as well. For example, a slight movement of the phone without fully pushing it away or pulling it towards the participant could be counted as an approach or an avoidance movement. This could also be controlled for by performing the experiment in a lab setting for the
experimenter to be able to correct any wrong movement or positioning of the phone.
Theoretical Implications
One of the aims of this study was to replicate the one carried out by Rohr and colleagues (2015), however, not all the same findings replicated. We did not obtain significant findings on the interaction between food type and color whereas Rohr et al. (2015) did. They found that the color green led to more approach, that healthy food items and green led to more approach and that unhealthy food items led to avoidance especially when surrounded by red. Despite being unable to differentiate between the colors green and blue in the analysis, it did however provide useful information that both earthy green and blue led to greater approach tendencies than earthy red. This extended Rohr et al.’s (2015) finding that green led to more approach tendencies compared to the color red which led to more avoidance, independently of food type. This could therefore be the effect of the traffic-light context and the association people have with red, being stop, presented alongside green. This study adds to the scientific evidence that other cool colors, like blue, may also lead to more approach tendencies compared to the color red, which fosters mostly avoidance tendencies (Bruno, Martani, Corsini & Oleari, 2013; Genschow, Reuntner & Wänke, 2012). Additionally, this validates both Baik and Suk’s (2010) findings of the effect of earthy colors being symbolic of nature and well-being but
it also validates that some brighter and more saturated colors do lead to pleasure which can be likened to approach behaviors (Guilford & Smith, 1959).
Practical Implications
This study also has some practical implications. As mentioned above, one finding of this study is that people approach healthy food faster than unhealthy food, showing an approach tendency towards healthy food. This is important information for marketers as it shows that people are more and more drawn to healthy foods compared to unhealthy foods and therefore this latter industry can be further developed. With this information, marketers in healthy food industries can work towards gaining consumer loyalty by, for example, pursuing more research on what it is that attracts consumers towards healthy foods. In contrast, marketers in unhealthy food industries can use some of our insights to improve the appeal of their unhealthy food packaging or try to understand why it is that consumers nowadays are less attracted to unhealthy food. It can additionally be used as an incentive to make unhealthy food industries reconsider the products they sell and to provide evidence that developing into healthier food industries is the way forward.
Marketers can also gain some important insights from our study on what colors and shades to use in the conception of packaging to attract the consumer. This study showed that people are more likely to approach earthy green or blue compared to earthy red and popping red compared to popping green or blue. Although this study was unable to differentiate between the specific effects of using earthy green or earthy blue, it still demonstrates that these two lead to more approach tendencies than popping colors. We also extend Rohr et al.’s (2015) finding that green led to more approach tendencies by showing that this holds for the color blue as well, independently of food type. Marketers
can retain this information and use it when creating package designs to increase their consumers’ likelihood to select their food product over another brand’s. For example, marketers can use earthy cool colors and popping warm colors in their packaging design (Baik & Suk, 2010; Guildford & Smith, 1959). However, it would be interesting to further research whether this only applies to the food industry or whether it has an effect in other industries.
The use of the mobile AAT as an approach-avoidance tendency measure in the field of marketing also poses a practical implication. The advantage of using a phone app to measure approach and avoidance tendencies of food items is that it recreates the real-life movements consumers make when shopping in a supermarket. We grab a product we would like to buy or are interested in and put back or physically walk away from those we do not wish to purchase. Therefore, the results of this study, although mostly not significant, could have a real-life application for consumers shopping in supermarkets. The mobile AAT could then be used for more research on consumption of products in the food industry but also in other industries (beauty, home decoration, etc.). As mentioned before, other industries may benefit from the mobile AAT, for example the fashion industry. It could be used to assess which new clothing designs please the consumer before releasing new products in order to maximize profit made from them.
Finally, the AAT is a task which has been used to retrain unhealthy behaviors such as hazardous alcohol consumption or drug use (Wiers, Eberl, Rinck, Becker & Lindenmeyer, 2011). As significant findings were obtained on participants’ approach bias towards healthy foods, future research on approach-avoidance tendencies could be
AAT could be used as a mobile behavioral retraining tool. Using the mobile AAT to retrain these approach tendencies towards unhealthy foods could be a crucial for people suffering from obesity or who need to follow a healthy diet due to health complications. This behavioral intervention could be a good tool to use especially before exposure to the food items they are more likely to want to approach (e.g: a trip to the supermarket).
Conclusion
In conclusion, although the results of this study are not as promising as expected due to the limitations faced, it has led to one replication and one new finding. Firstly, people have an approach bias towards healthy foods compared to unhealthy foods. This finding is most likely due to the rising awareness of the importance of healthy diets. It may also be due to the fact that Raghunathan, Naylor, and Hoyer’s (2018) unhealthy-tasty intuition is likely to be limited to people with low self-control or who suffer from obesity. Secondly, earthy cool colors (blue or green) and popping red increased approach tendencies to food stimuli more than popping cool colors or earthy red independently of food type. This finding supports Goldstein’s (1942) warm vs. cold colors theory and the effect these two categories of colors have on people’s interpretation of them. The paler the cool colors, the more they reflect serenity and colors the way we see them in nature. However, the more saturated the color red is, the more exciting it is in the context of food. Therefore, it can be said that the saturation and value of a color reinforce its warm or cool aspect and leads to congruent effects towards a consumer’s approach tendencies. This leads to the general conclusion that, although there is more research to be done, this study provides valuable information to marketers on what their consumers approach in terms of food type and in terms of hue, saturation and value of the packaging. Replicating
this study in the future, in a controlled environment and using more sophisticated analyses could lead to more precise findings on the effect that hue, saturation and value has on the approach or avoidance of healthy and unhealthy foods.
References
Adamson, J. C. (2018). Hue, Saturation & Value - The characteristics of color. Baik, M., & Suk, H. J. (2010, August). Environmentally Friendly Packaging
Appealing to Consumers’ Emotions-Focusing on Organic Food Packaging. In The
2010 Asian Conference on Design and Digital Engineering, Jeju (pp. 25-28).
Bellizzi, J. A., Crowley, A. E., & Hasty, R. W. (1983). The effects of color in store design. Journal of Retailing, 59, 21-45.
Bellizzi, J. A., & Hite, R. E. (1992). Environmental color, consumer feelings, and purchase likelihood. Psychology and Marketing, 9(5), 347-363.
doi:10.1002/mar.4220090502
Bruno, N., Martani, M., Corsini, C., & Oleari, C. (2013). The effect of the color red on consuming food does not depend on achromatic (Michelson) contrast and extends to rubbing cream on the skin. Appetite, 71, 307-313.
doi:10.1016/j.appet.2013.08.012
Carter, C. S., Mintun, M., & Cohen, J. D. (1995). Interference and facilitation effects during selective attention: an H215O PET study of Stroop task
performance. Neuroimage, 2(4), 264-272.
Cheung, T. T., Gillebaart, M., Kroese, F., & De Ridder, D. (2014). Why are people with high self-control happier? The effect of trait self-control on happiness as mediated by regulatory focus. Frontiers in psychology, 5, 722.
Devaux, M., Graf, S., Goryakin, Y., Cecchini, M., Huber, H., & Colombo, F. (2017).
Eberl, C., Wiers, R. W., Pawelczack, S., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2013). Approach bias modification in alcohol dependence: do clinical effects replicate and for whom does it work best?. Developmental cognitive
neuroscience, 4, 38-51.
Genschow, O., Reutner, L., & Wänke, M. (2012). The color red reduces snack food and soft drink intake. Appetite, 58(2), 699-702. doi:10.1016/j.appet.2011.12.023 Goldstein, K. (1942). Some experimental observations concerning the influence of
colors on the function of the organism. Occupational Therapy Rehabilitation, 21, 147-151. doi:10.1097/00002060-194206000-00002
Guilford, J. P., & Smith, P. C. (1959). A system of color-preferences. The American Journal of Psychology, 72(4), 487-502.
Kakoschke, N., Kemps, E., & Tiggemann, M. (2017). Approach bias modification training and consumption: A review of the literature. Addictive behaviors, 64, 21-28.
Klein, A. M., Becker, E. S., & Rinck, M. (2011). Approach and avoidance tendencies in spider fearful children: The approach-avoidance task. Journal of Child and Family Studies, 20(2), 224-231.
Kliger, D., & Gilad, D. (2012). Red light, green light: color priming in financial decisions. The Journal of Socio-Economics, 41(5), 738-745.
doi:10.1016/j.socec.2012.07.003
Lichtenfeld, S., Elliot, A. J., Maier, M. A., & Pekrun, R. (2012). Fertile Green: Green Facilitates Creative Performance. Personality and Social Psychology
Mehta, R., & Zhu, R. J. (2009). Blue or red? Exploring the effect of color on cognitive task performances. Science, 323(5918), 1226-1229.
Moller, A. C., Elliot, A. J., & Maier, M. A. (2009). Basic hue-meaning associations. Emotion, 9(6), 898.
Nakshian, J. (1964). The Effects of Red and Green Surroundings on Behavior. The Journal of General Psychology, 70(1), 143-161.
Pravossoudovitch, K., Cury, F., Young, S. G., & Elliot, A. J. (2014). Is red the color of danger? Testing an implicit red–danger association. Ergonomics, 57(4), 503-510. Raghunathan, R., Naylor, R. W., & Hoyer, W. (2006). The Influence of the
Unhealthy = Tasty Intuition on Food Consumption Decisions. PsycEXTRA Dataset, 70(4), 170-184. doi:10.1037/e621442012050
Rohr, M., Kamm, F., Koenigstorfer, J., Groeppel-Klein, A., & Wentura, D. (2015). The color Red Supports Avoidance Reactions to Unhealthy Food. Experimental Psychology, 62(5), 335-345. doi:10.1027/1618-3169/a000299
Schaie, K. W., & Heiss, R. (1964). Color and personality.
Turnwald, B. P., Jurafsky, D., Conner, A., & Crum, A. J. (2017). Reading between the menu lines: are restaurants’ descriptions of “healthy” foods
unappealing?. Health Psychology, 36(11), 1034.
Valdez, P., & Mehrabian, A. (1994). Effects of color on emotions. Journal of Experimental Psychology: General, 123(4), 394-409. doi:10.1037//0096 3445.123.4.394
Watson, P., De Wit, S., Cousijn, J., Hommel, B., & Wiers, R. W. (2013). Motivational mechanisms underlying the approach bias to cigarettes. Journal of Experimental Psychopathology, 4(3), jep-030512.
Wiers, R. W., Eberl, C., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2011). Retraining automatic action tendencies changes alcoholic patients’ approach bias for alcohol and improves treatment outcome. Psychological science, 22(4), 490-497.
Appendix A
Examples of stimuli as presented would be presented in the mobile AAT
popping red: hue = 0, saturation = 100, value = 100;
popping green: hue = 120, saturation = 100, value = 87
popping blue: hue = 242, saturation = 100, value = 87
earthy red: hue = 0, saturation = 95, value = 40
earthy green: hue = 120, saturation = 70, value = 30
Appendix B
Appendix C
Informed consent given to participants on the mobile AAT.
Dear participants,
Welcome and thank you for participating in our study! In this study “Food Game” we ask you to answer some demographic questions and play a little game on your phone or tablet. We will show you some food items and tell you to push or pull your phone/tablet.
Participation is completely voluntary. You can decide at any moment not to participate or to stop without this having negative consequences. You do not need to provide a reason for quitting. In total, playing the game and filling in the questionnaires won’t take longer than 10 minutes. As a compensation for your effort, we will pay the standard hour rate or you will receive course credits. All the data that will be collected during the study will be anonymized, that is, it will be processed and stored in a coded way and will be decoupled from any personal information. All the collected data will only be used for research purposes and reports will only involve anonymized results.
If everything is clear we would like to ask you to read and sign the informed consent. If you have further questions, you can contact us via
louise.dillies13@gmail.com or v.f.m.rosenboom@umail.leidenuniv.nl
We would like to thank you again for participating in our study.
Valerie Rosenboom and Louise Dillies
By signing this form you agree to the following:
- I have read the above information letter. I could ask additional questions. My questions have been answered adequately. I have had sufficient time to decide whether or not I participate.
- I am aware that participation is completely voluntary. I know that I can decide at any moment not to participate or to stop without negative consequences and without having to provide a reason.
- My responses are processed and stored in a coded way.
- I give consent to use my data for the purposes that are mentioned in the information letter. I consent to participating in this study.
Appendix D
Debrief given to participants at the end of the AAT.
Thank you very much for your participation!
In this study you played a game on your phone or tablet. With that we measured automatic approach and avoidance tendency by comparing how quickly you push and pull the drawings of healthy and unhealthy food items. For example, if you pull healthy foods quicker towards yourself than unhealthy foods, you will show a stronger automatic approach tendency towards healthy food.
Besides that, the food items had red, green or blue circles around them. This was to test whether a certain color affected your automatic approach or avoidance tendency towards the food items. Some of you were in the Traffic Light context and therefore saw red and green circles. In this condition we also tested whether manipulating the saturation and value of the colors could lead to quicker approach tendencies towards healthy foods. Some of you were taken out of the Traffic Light context and therefore saw red and blue circles. This was to investigate whether dissociating red from the traffic light association (seeing it with green or yellow) would lead to less avoidance tendencies.
We expect that:
1. People will have an approach bias towards healthy compared to unhealthy food. 2. In unhealthy foods but not in healthy foods, people have an approach bias towards
bright green compared to bright red (i.e. an avoidance bias towards red compared to green).
3. In healthy foods but not in unhealthy foods, people have an approach bias towards earthy green compared to earthy red.
4. People will have an approach bias to unhealthy foods presented with bright red compared to bright blue (taking away the Traffic Light context).
Because some people might still participate in the experiment over the coming weeks, we want to ask you not to talk about the research with other potential participants. Like this we can ensure that each participant starts the experiment with the same amount of knowledge.
Again, thank you very much for your cooperation. We appreciate it very much!
If you have any questions, comments, or complaints you can contact:
Hilmar Zech, Institute of Psychology, Leiden University, email: h.g.zech@fsw.leidenuniv.nl
