The Color Naming Effect: How Fancy Color Names Influence Product Preference

(1)

The Color Naming Effect: How Fancy Color

Names Influence Product Preference

Master Thesis

Ilona Idserda

(2)

The Color Naming Effect: How Fancy Color

Names Influence Product Preference

(3)

Management Summary

In all sorts of product categories, from wall paint to cosmetics, fancy color names such as Glamour Green or Vivid Plum are emerging. Clearly, as companies that are dealing in colors spend enormous amounts of time and money into selecting the most appealing names, this color naming practice suggests that color names have an influence on product preference. This is confirmed by previous research, which revealed that color names significantly influence consumers’ preferences, and that fancy color names lead to significantly higher product evaluations than generic color names do.

By conducting empirical research, this study examines how fancy color names influence product preference. Three explanations for how fancy color names affect product preference are discussed, namely the Attention Theory, the Association Theory, and the Attributes Theory.

By completing an experiment, the participants of the study rated towels with positively associated unique, negatively associated unique, and neutrally associated common color names attached to them. Where the results of the main study showed no difference in participants’ rating of product preference for positively associated unique, negatively associated unique, and neutrally associated common color names, the results of the follow-up study in which exposure of the color names was increased did in fact give some more insights into the color naming phenomenon. More specifically, the results provide an indication of how the color naming practice influences consumers’ product preference, and so could act as a guidance of how to use the color naming practice beneficially. First of all, results indicated that the extent of exposure of color names matters; the better the color name is noticed, the more likely it is that the color name influences product preference. Second, the uniqueness factor of fancy color names does not seem to influence product preference for low-involvement products, and negatively associated color names should be avoided, since they have the least significant impact on product preference. Third and finally, positively and neutrally associated color names activate positive categories, which might lead to product preference for low-involvement products.

(4)

Preface

You are holding my master thesis on the influence of color names on product preference. I have written this thesis as a final project in order to complete the master Business Administration, in the direction of Marketing Management. Personally, I believe the topic of this thesis is an interesting and entertaining topic, and by writing this thesis I have developed my knowledge on the subject, but also on the process of doing empirical research.

Hereby I would like to take the opportunity to thank a couple of people who supported me during the process.

First of all I would like to thank my first supervisor, dr. Mirjam Tuk. During the whole process she gave me useful feedback, and interesting ideas in order to improve my thesis. Also, I would like to thank dr. Debra Trampe for examining my thesis.

Second, I would like to thank my friends and family for their support. My special thanks go out to my parents Pieter and Hanneke, for giving me the opportunity to study at the university and for their continuous love and support. Furthermore, I would like to thank Koen for always being there for me. Also, I would like to thank Carla, Monique, and Henrieke for their support. Finally, I would like to thank all the respondents for completing the questionnaires. Without their input, the completion of this study would have been impossible.

The completion of this thesis is at the same time the completion of my study period in Groningen. A period I will never forget and will always look back on with great pleasure.

(5)

An examination of the color names attached to many product categories reveals an interesting phenomenon. For example, it is remarkable that hardly any lipstick carries its generic name, such as pink or red, nor the name that specifically describes its color, such as light pink or dark red. Instead, most shades of lipstick carry a fancy name, which links a positive emotion to a shade, such as Red Passion or Golden Toffee or using a positive emotion without even linking it to a shade, such as Cute Pie or Star.1 These types of fancy color names are more popular than ever,

and are appearing in all sorts of product categories, from wall paint (e.g., Flexa and the color ‘Room Service’) to cosmetics (e.g., Maybelline and the nail polish Metallic Rose) to household equipment (V&D and the Apple Green towel).

Clearly, as companies that are dealing in colors spend enormous amounts of time and money into selecting the most appealing names for their products, this naming practice suggests that the color names attached to products are an important influence on consumers’ decisions. There are various suggested reasons why color naming has a positive effect. Miller and Kahn (2005) investigated whether color and flavor names have an effect on consumer’ choice, and they found that color names indeed influence consumers’ choice, and that consumers prefer atypical and unspecific color names (e.g., friendly green) to more typical and more specific color names (e.g., lemon yellow). Their explanation for this finding is that consumers react favorably to unusual color names because consumers assume that all information offered to them by the marketer is meant to be relevant and informative. If the message is not informative and does not conform to expectations (as is an unusual color name), consumers search for the reason for deviation. This search results in consumers having more attention for the product with the unusual color name, and thus noticing additional (positive) attributions about the product, leading to a more favorable response. Besides Miller and Kahn, also Skorinko et. al., (2006) studied the color naming effect. More specifically, they investigated whether or not fancy names are indeed more appealing than generic names, and whether the names associated with colors really make a difference in consumers’ behavior. Also the results of their study showed that names indeed influence how colors are perceived, and that the fancy-name effect also resulted in the fancy color named product to be the preferred product, to be the more likely to be purchased product, and also the

(8)

participants of the study were willing to pay more money for the fancy color named products than they were for exactly the same products carrying the generic color name. In order to explain their findings, Skorinko et. al. refer to Tversky and Kahneman (1981), who found that people reliably opted for those scenarios framed in terms of gains instead of losses. In the study of Skorinko et. al. the fancy color names can be seen as positive frames, whereas the generic names were just that – generic. Furthermore, Skorinko et. al. refer to Fiske and Neuberg (1990), who argued that people use categorization when forming impressions of others as a defaults. In the case of color naming, the fancy name activates a positive category, and this positive category in turn positively influences the impression forming of the product.

Even though both above mentioned studies found the preference for unusual, fancy color names, it is remarkable that their suggested explanations are deviating. Whereas Miller and Kahn (2005) suggest the incongruency between the message of the color name and the knowledge and expectation of the consumers leads consumers to search for additional (positive) attributions, resulting in a more favorable response, Skorinko et. al. (2006) suggest that the positive association of the fancy name leads to the preference.

Besides the explanations mentioned in previous research on color naming, this research suggests another explanation for the preference of fancy color names and products carrying fancy color names. This suggested explanation is based on the product attribute uniqueness, and more specifically on the concept of individual’s need for uniqueness. As consumers feel a need to be different from other individuals, they acquire, utilize and dispose products in order to develop and enhance one’s personal social identity (Tian, Bearden, and Hunter 2001). Color naming is possibly one of a marketer’s strategies to appeal to this need for uniqueness, by giving the product in question a unique color name. This in turn conforms to the individuals need for uniqueness, and as a result, the individual is likely to prefer the unique color name and the product carrying the unique color name over the generic color name and product carrying the generic color name.

(9)

How do fancy color names influence product preference?

In particular, this research will address the influence of fancy color names on consumers’ color preference, purchase intentions and price willingness to pay. In order to get a deeper understanding of the explanation for the fancy color naming effect, this research will introduce two new variables: negatively associated color names and uniqueness.

Negatively associated color names - Besides fancy color names with a positive association, such

as Glamour Green or Vivid Plum, there are also fancy color names that do not carry a positive element. In fact, some fancy color names even carry a negative element, such as Blood Red (lipstick by Sephora) or Storm (wall paint by VT Wonen). As it has already become clear from previous research, fancy color names are preferred over generic color names. However, when examining how fancy color names influence product preference, the question is whether or not both positively as well as negatively associated fancy color names are preferred over generic color names, or that it is just the positively associated fancy color names.

Uniqueness – Individuals feel the need to be different from other individuals, which is also

known as the consumers’ need for uniqueness. As a result, individuals acquire, utilize and dispose goods in order to develop and enhance one’s personal identity (Tian, Bearden, and Hunter 2001). The color naming practice might be a method of marketers to attach a uniqueness appeal to a product, in order for the product to conform to the consumer’s need for uniqueness. By introducing the variable uniqueness in this research, it will become clear whether or not a fancy color name is indeed a uniqueness appeal, and consequently whether or not it is an explanation for the color naming effect.

The introduction of the two variables negatively associated color names and uniqueness results in the following research questions:

What is the influence of negatively versus neutrally and positively associated color names on product preference?

(10)

By answering these research questions and consequently providing more insight into the above mentioned problem statement, a new dimension to the already existing literature on color naming effects will be contributed, namely how fancy color names influence product preference. More specifically, the results of this research will give more insights into the suggested explanations of the finding that fancy color names are preferred over generic color names.

When results will show that there is a difference between positively and negatively associated color names on product preference, this supports the Association Theory of Skorinko et. al. (2006), as they claim it is the positive association of a fancy color name that leads consumers to prefer the fancy color name. When however the results will show that uniqueness is the main explanation for the color naming effect, this will support the newly suggested Attributes – Uniqueness Theory. When results show that there is in fact no significant difference between positively and negatively associated color names on decision making and also uniqueness is not significant in explaining the color naming effect, both theories can be rejected while the Attention theory of Miller and Kahn (2005) cannot be rejected. As it is not the association in the name nor the uniqueness that leads to preferring the fancy color name, it can still be the additional search for attributions. Finally, it is also possible that both positively and negatively associated color names are preferred over generic color names, however additionally the positively associated color names are preferred over the negative ones. In this case, a combination of the Association theory of Skorinko et. al. (2006) with the Attention theory of Miller and Kahn (2005) or the newly suggested Attributes – Uniqueness theory is possible.

In practice, the results of this study will provide marketers with the empirical evidence of why fancy color naming is advantageous. As a result, knowing whether is the attention, association, or uniqueness that leads to product preference, this makes it more comprehensible for marketers to select an appealing color name to accompany their products.

(11)

Theoretical Framework

In this theoretical framework, the explanations suggested in previous research in the field of color naming will be discussed. Besides the explanations suggested in previous research, this theoretical framework offers and discusses an alternative explanation based on previous consumer behavior research. Moreover, hypotheses on how fancy color names influence product preference are stated.

The Attention Theory

A first explanation why consumers might prefer atypical, fancy color names, is that due to the ambiguity of the color name, the product carrying the atypical fancy color name receives more attention from the consumer as the consumer engages in an extra thought process in order to explain for the ambiguous color name. Due to this extra thought process consumers might notice additional (positive) attributions about the product that they would not have noticed otherwise (Theory of Conversational Implicature), or the consumer might discover the connection of how the adjective of the atypical color name describes the color, leading to a feeling of proud and resulting in a positive affect and evaluation of the product (Incongruency Theory). It should be noted that these two theories are not mutually exclusive, however that the Theory of Conversational Implicature applies for names that are uninformative in a literal sense (ambiguous color names) and the Incongruency Theory applies for names which are unfamiliar due to their atypicality (unexpected descriptive color names). (Miller and Kahn 2005). Both theories are discussed in more detail below.

Theory of Conversational Implicature

(12)

question, however rather than assuming the respondent has changed the topic, the listener assumes that the information is relevant to the goal of the ongoing conversation and the respondent is going to the theatre that night to see a great play. These same assumptions hold for marketing messages as well. The consumer may assume that all information offered by the marketer is meant to be relevant and/or informative, and they will consequently try to make sense of it. If however a color name is uninformative in the literal or semantic sense, consumers will search for the meaning of the communication by paying more attention to the product, and consequently noticing additional (positive) attributions about the product that they would not have noticed otherwise. As a result, ambiguous color names will yield more positive attributions about the product than will a common name, which does not require any additional search for attributions. (Miller and Kahn 2005)

Incongruency Theory

(13)

helplessness, resulting in a negative affect and negative product valuations. To conclude, a color name that is mildly incongruent, will lead consumers to pay more attention to the color name in order to resolve the incongruity. This in turn might make the item in question more interesting, and when the consumer is able to resolve the incongruity, this is likely to lead to a feeling of proud and a positive affect, and in the end a more positive evaluation of the product.

Hypotheses Based on the Attention Theory

According to the theories discussed above, it is the extra attention the customer gives to the atypical fancy color name and the product with the atypical fancy color name, what leads consumers to prefer products with fancy color names. Therefore, according to the Attention Theory, it is likely to assume that it makes no difference whether the fancy name encompasses a positive or a negative association, as it is the atypicality of a color name which leads to an extra thought process and consequently to a more positive product evaluation. As a consequence, the following hypotheses are proposed:

H1: Positively associated fancy color names have a significantly more positive influence on

product preference than neutrally associated generic color names have

H2: Negatively associated fancy color names have a significantly more positive influence on

product preference than neutrally associated generic color names have

H3: There is no significant difference between the influence of positively and negatively

associated fancy color names on product preference

(14)

Miller and Kahn, 2005). Therefore, as these theories make a distinction between different types of fancy color names, namely ambiguous and unexpected descriptive color names, and also the timing effect is important, investigating which theory better accounts for the overall Attention Theory is beyond the scope of this research.

Furthermore, with respect to the Incongruency Theory, for this research it should be considered that there are a variety of fancy color names, of which some are in general more or less often and to a smaller or larger extent incongruent than others. And in this line of reasoning, it is reasonable to assume that negatively associated color names are more often, and perhaps also to a larger extent, incongruent with consumers’ expectations, as negatively associated color names are less common and therefore less likely to be anticipated. Even though it is beyond the scope of this research to investigate whether or not negatively associated fancy color names indeed are more often and to a larger extent incongruent than positively associated color names, current research will make sure that the extent of incongruency of both positively and negatively associated selected color names are in balance in order to ensure that the extent of incongruency does not play a role in explaining why fancy color names are preferred over generic color names.

The Association Theory

A second explanation why consumers prefer fancy names over generic names, might be due to the fact that most fancy names used by marketers are positively framed. Products with these positively associated color names might lead consumers to experience a positive affect, activate a positive category which influences impression making and/or identify this positively associated product with their desired end states, and consequently to prefer this product. This Association Theory is further explored in the following theories.

The Framing Theory

(15)

other in negative or loss terms (Donovan and Jalleh 1999). A considerable body of research has already investigated whether or not a positively or gain presented outcome, or a negatively or loss presented outcome affects judgment and decision making. Levin et al. (1998) suggested that attribute framing effects occur when evaluations of an object or event are more favorable if a key attribute within any given context is framed in positive rather than negative terms. Furthermore, Levin et. al. (1998) argue that these attribute framing effects indeed do occur, because information is encoded relative to its descriptive valence. Positive labeling of an attribute will thus lead to an encoding of the information that tends to evoke favorable associations in memory, whereas negative labeling of the same attribute will cause an encoding that evokes unfavorable associations. In other words, positive framing supports more favorable evaluations and negative framing supports less favorable evaluations. Translating this to the color naming effect and assuming that fancy color names are in general more often positive than negative, the preference of fancy names over generic names may reflect the fact that the fancy names can be seen as positive frames, where the generic names were just that – generic. The positive frame of the fancy color name may result in encoding of the information that tends to evoke favorable associations in memory, resulting in more favorable evaluations of the color name and the product in question. On the contrary however, the negative frame of a negatively associated fancy color name may result in encoding that tends to evoke unfavorable associations in memory, resulting in more negative evaluations of the color name and the product in question.

The Categorization Theory

(16)

Self-image Congruence Theory

(17)

It should be noted that there are exceptions to the above-mentioned translation of the undesired self and color naming. For example, tough guys would not like to be associated to the color ‘sweet pink’ as this is incongruent with their self-image, even though the association of the color name is a positive association. However, albeit the exceptions, it can be assumed that in general people prefer to identify themselves with positive associations instead of negative associations.

Hypotheses Based on the Association Theory

According to the theories discussed above, the positive associations used in most fancy color names are the reason why consumers prefer the fancy color names and products carrying the fancy color names over generic color names and products carrying a generic color name. Therefore, according to the Association Theory, it is likely to assume that it does in fact make a difference whether or not the fancy name encompasses a positive or a negative association. As a consequence, the following hypotheses are proposed:

H4: Positively associated fancy color names have a significantly more positive influence on

product preference than neutrally associated generic color names have

H5: Generic color names have a significantly more positive influence on product preference

than negatively associated fancy color names have

However, as explained in the theories above, it is suggested that associations can influence product preference in multiple ways. In order to develop a deeper understanding of how fancy color names influence product preference, the following hypotheses are stated:

H6: As suggested by the Categorization Theory, positively associated fancy names activate a

positive category which in turn leads consumers to prefer the product with the positively associated color name

H7: As suggested by the Self-image Congruence Theory, when people can better identify

(18)

The Attributes Theory

A number of previously published studies have identified a wide range of product attributes relevant to individual consumers. The attribute that is expected to be the most influenced by the color naming practice is ‘uniqueness’. As marketers strive to come up with original, differentiating color names in order for their products to stand out, this results in an enormous pile of color names, all unique in its kind. Besides the fact that individuals pursue to be different from other people – the need for uniqueness –, this uniqueness attribute also could spill over to other products attributes. The concept of the need for uniqueness and its spillover effect will be discussed in more detail below.

The Need for Uniqueness

As people seek to be different from other individuals, the products that they purchase and their uses and displays may serve as recognizable symbols of uniqueness or specialty. For example, the clothing that people purchase and wear contribute to the style that person wants to portray in order to enhance the person’s unique identity. More specifically, individuals pursue differentiation relative to others by acquisition, utilization, and disposition of consumer goods in order to develop and enhance one’s personal social identity, also known as the consumers’ need for uniqueness (Tian, Bearden, and Hunter 2001). This need to be different from other people is driven by the so called ‘counter conformity motivation’. According to this motivation individuals feel a threat to their identity, as occurs when they perceive that they are highly similar to others (Snyder and Fromkin 1977).

(19)

Hypotheses Based on the Attributes - Uniqueness Theory

According to the theory discussed above, a fancy color name leads consumers to perceive the product carrying this fancy color name to be unique, and this unique image which conforms with the consumers need to be different is what leads consumers to prefer products with fancy color names. Therefore, according to the Attributes Theory, it is likely to assume that it makes no difference whether the fancy name encompasses a positive or a negative association, as long as the color name employs a product-scarcity and uniqueness appeal. As a consequence, the following hypotheses are proposed:

H8: Unique fancy color names have a significantly more positive influence on product

preference than common generic color names have

H9: There is no significant difference between the influence of positively and negatively

associated unique fancy color names on product preference

The Uniqueness Spillover-Effect

Besides the fact that unique fancy color names might conform more to an individual’s desire to be unique than common generic color names do, the unique image a product gets from a fancy color name is also likely to influence consumer’s perceptions of other product attributes.

(20)

consumers to believe that the quality of the product is higher. Last, as a unique product might be perceived as more exclusive, beautiful, and having a better quality, the price consumers are willing to pay is probably higher as well. In fact, this has also been confirmed by Skorinko et. al. (2006), who showed that consumers were willing to pay a higher price for products carrying a fancy, and thus original, color name compared to products carrying a generic color name.

Hypothesis Based on the Attributes - Uniqueness – Spillover Theory

According to the Spillover Theory suggested above, fancy color names give a product a unique image, and this uniqueness might spill over to other product attributes and lead consumers to believe the product is more exclusive, more beautiful, more expensive, and of a better quality. As a consequence, the following hypothesis is proposed:

H10: A color name that results in the product carrying the name to be perceived as unique

significantly influences other product attributes and in turn product preference indirectly

Conceptual Model

(21)

Summary

As this research suggests different theories which could explain the preference of fancy color names over generic color names, this theoretical framework ends with a summary of the discussed theories and their effects.

According to the Attention Theory, due to their atypicality, fancy color names lead to an extra thought process about the product and the color name in question. This results in the noticing of additional (positive) attributions and/or solving the discrepancy between the color name and consumers’ expectations, which in the end results in product preference. So for the Attention Theory to hold, as it is the atypicality in the fancy color name which leads to product preference, this research should show that both positively and negatively associated color names equally have a significant stronger positive influence on product preference than generic color names have.

According to the Association Theory, it is the positive association of fancy color names that leads to product preference. This can be explained by three theories, namely the Framing Theory, the Categorization Theory, and the Self-image Congruence Theory. Besides investigating the general effect of whether positively associated color names are indeed preferred over neutrally and negatively associated color names and that neutral color names are preferred over negatively associated color names, current research will also investigate whether the Categorization or the Self-image Congruence Theory explains for the preference of fancy over generic color names (testing whether or not the Framing Theory explains for the effect is beyond the scope of this research). In order for the Categorization Theory to hold, current research should show that positively associated color names indeed activate a positive category, resulting in product preference. In order for the Self-image Congruence Theory to hold, current research should show that when people can identify themselves with a product carrying a specific color name, they will prefer the product carrying this color name.

(22)

(23)

Research Design

The current research is interested in how fancy color names influence product preference. In order to investigate the explanation for this effect, the influence of negatively versus positively associated color names, and the influence of the product attribute uniqueness are investigated in more detail.

Terminology

As in previous research various definitions for fancy and generic names were used, first it will be explained what in this research is meant by fancy and generic color names. Following Miller and Kahn (2005), there are four categories of color names, namely common (typical, unspecific; e.g. dark green, light yellow), common descriptive (typical, specific; e.g. pine green, lemon yellow), unexpected descriptive (atypical, specific; e.g. Kermit green, rainslicker yellow), and ambiguous color names (atypical, unspecific; e.g. friendly green, party yellow). In this research, common and common descriptive color names are categorized as generic color names, whereas unexpected descriptive and ambiguous names are categorized as fancy color names. As the following research is a within- and between participants design, and it should be avoided that consumers see through the purpose of the research and consequently answer in line of expectations, for generic color names only common descriptive color names will be used, as they blend in more with the fancy color names. Furthermore, the generic color names are naturally less unique compared to the fancy color names. Consequently, for the purpose of this research, generic color names can be seen as common color names, whereas fancy color names can be seen as more unique color names.

Pretest

The purpose of the pretest was to determine which color names to use in the main study. More specifically, the intention was to determine which color names people feel match the different color name categories used in the main study; positively associated unique color names;

negatively associated unique color names; and neutrally associated common color names. For

(24)

name is indeed labeled as positively associated and unique by the participants. Furthermore, with the pretest the extent of incongruency with people’s expectations of the color and the actual color was measured. The purpose of testing the incongruency for positively associated unique, negatively associated unique, and neutrally associated common color names was to guarantee that all color names are more or less equally incongruent, so that it could be excluded that the extent of incongruency as suggested by the Incongruency Theory would influence product preference. The questionnaire the participants received contained pictures of the three colors and its three shades, and also a variety of color names and statements to go together with these pictures. An example of such a statement is ‘I believe the color name ‘Zandstorm’ is common / unique. These color name statements were given so that the participants could rate the extent to which they considered each color name to be positively or negatively associated and unique or common on a semantic differential scale (1 = common, positive.; 9 = unique, negative). Furthermore, in order to measure the extent of incongruency, respondents were asked to rate the extent to which they believed the color names match the displayed colors. See appendix 1.1 for the pretest questionnaire.

Pre-test Results

A total of 26 people completed the questionnaire. Out of these 26 questionnaires, three participants’ data were not used (due to not fully completing the questionnaire), so that the results are based on 23 participants. Of these 23 respondents, 6 respondents were male and 17 respondents were female. The average age of the respondents was 40.

(25)

Limitation

Besides analyzing the data by conducting T-tests in order to see whether or not a color name is significantly positively associated unique, negatively associated unique or neutrally associated common, it is also constructive to see whether or not the color names significantly differ from each other on these dimensions. The reason why this is of interest is that when respondents do not experience a significant difference between for example a positively associated color name and a negatively associated color name, it can be expected that there will not be a significant difference between their ratings of product preference.

By conducting Oneway ANOVA tests en Post hoc tests, results showed that the selected unique names indeed significantly differ from the selected common color names. However, on the other hand not all positively associated color names differ significantly from the neutral color names. More specifically, ‘Belangrijk’ (brown; important) does not differ significantly from ‘Hout’ (brown; wood) (p = .751), and ‘Rust’ (green; serenity) does not significantly differ from ‘Appel’ (green; apple) (p = .148). Even though all the other positively, negatively and neutrally associated color names do differ significantly from each other, it should be kept in mind that the difference between the brown and the green positively and neutrally associated color names is not significant, which may result in the fact that the evaluations of the respondents will not differ significantly between these color names. For more extensive results of the Oneway ANOVA tests see appendices 1.4, 1.5 and 1.6.

Brown Blue Green

Positiveness Uniqueness Positiveness Uniqueness Positiveness Uniqueness Positive Unique Color Name 3.78a 5.96a 3.70a 6.09a 3.78a 6.83a Negative Unique Color Name 6.13b 6.30b 6.70b 7.04b 6.78b 6.30b Neutral Common Color Name 3.96a 3.96c 5.22c 3.91c 4.61a 4.09c

(26)

Research Method

First, in preparation for the current research, a product that is available in multiple colors and shades was selected. Besides the availability of multiple colors and shades, the product should also be unisex – interesting to both sexes –, due to the random selection of the sample population. Consequently, the product chosen for this research was a towel, which is in the same style available in three colors and three shades. In order to avoid color preferences, the method for color selection suggested by Skorinko et. al. (2006) was applied. Accordingly, the colors brown and green were chosen because they are less standard, and the color blue was chosen because past research has shown the preference for non-primary colors and a preference for the color blue in particular. Three colors instead of one color were chosen in order to be able to determine the effect on product preference for multiple colors.

Second, for each color towel three different but closely resembling shades were selected. For all shades names were selected, where one shade was given a positively associated unique fancy color name, one a negatively associated unique fancy color name, and one a neutrally associated common generic color name (see appendix 1.2 for the selected color names).

(27)

Figure 2, The Uniqueness Spillover Effect

Current research can be described as a within- and between-participants design, as each respondent only views one color of towel (either brown, blue, or green) but views three shades (light, medium, dark). The first towel was given a positively associated unique color name, the second towel was given a negatively associated unique color name, and the third towel was given a neutrally associated common generic color name. The names are counterbalanced for all three shades of towels, so that 1/3 of the participants will see the light shaded towel with a positively associated unique color name, 1/3 will see the medium shaded towel with a positively associated unique color name, and 1/3 will see the dark shaded towel with a positively associated unique color name.

Data Collection

(28)

Procedure

(29)

Results

As suggested by previous research, fancy names influence product preference in a positive manner. The theoretical framework of this research has suggested some explanations on how fancy color names influence product preference, and consequently this chapter discusses the tests used to analyze the data and its resulting outcomes, in order to attempt to explain how fancy color names influence product preference.

Associations, Uniqueness, and Product Preference

It is suggested by the Attention Theory, the Association Theory, and the Attributes Theory in the theoretical framework, that there is a difference between the influence of positively associated unique color names, negatively associated unique color names, and neutrally associated common color names on product preference. In order to study this deviation, and thus contribute to the explanation of how fancy color names influence product preference, first the overall effect of positively associated unique, negatively associated unique, and neutrally associated common color names on product preference will be studied. In other words, first, no difference will be made between the effect of brown color names, blue color names, and green color names, but all color names will be taken together.

Even though differences between the influence of positively associated unique, negatively associated unique, and neutrally associated common color names on product preference were hypothesized, current research revealed no significant difference between the different types of color names on color preference. More specifically, current research performed three Oneway ANOVA tests, which showed that there is no significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common color names on color preference F(2,267) = .456, p = .634, purchase intentions F(2,267) = .075,

p = .928, and price willingness to pay F(2,267) = .230, p = .795. Consequently, as there is no

significant difference between positively associated unique, negatively associated unique, and neutrally associated common color names product preference, this results in the rejection of the hypotheses H1, H2, H4, H5, H8 and H9. Hypothesis H3 can be accepted on the basis of these

(30)

effect of positively and negatively associated color names on product preference, these positively and negatively associated color names also do not significantly differ from neutrally associated color names. For more extensive results, see Appendix 2.2.

Furthermore, current research investigated the effect of different types of color names on the activation of positive categories and in turn on product preference. In order to measure the effect of the activation of positive categories on product preference, two different questions were asked in the survey. The first question would measure the extent the color name made the respondent think of fun things, the second question would measure the extent the color name made the respondent think of positive products. In order to study the effect of the activation of positive categories on product preference, first it should be explored whether or not the two questions from the survey statistically could be combined so that together they could account for the extent the color name made the respondent activate positive categories. In order to test whether these two variables could be combined, a Cronbach’s Alpha was computed, as the Cronbach’s Alpha measures how well a set of items (or variables) measures a single unidimensional latent construct. In the case of color naming, the reliability of thinking of fun things and of thinking about positive products is very high (0.938), and also the inter-item correlation is high (0.884). As a consequence, the two variables thinking of fun things and thinking of positive products can be combined to create one scale; the extent the color name leads to the activation of positive categories.

Next, to see whether there is a significant difference between the effect of positively, negatively and neutrally associated color names on the activation of positive categories, a Oneway ANOVA was performed. This test however illustrated that on average neither positively as well as negatively as well as neutrally associated color names activate a positive category (respectively

M = 3.59, M = 3.33, and M = 3.51) and that they do not significantly differ F(2,267) = .521, p =

.594. As a result, this indicates that H6 can be rejected, since positively associated color names do

not activate positive categories which result in product preference more than negatively and neutrally associated color names do. For more extensive results see Appendices 2.2 and 2.3.

(31)

In order to test that people prefer products with which they can identify themselves, a Correlation Analysis was performed. This Correlation Analysis indicated that the extent to which people can identify themselves with the product and the evaluation of the product is indeed significantly positively correlated. More specifically, the extent to which people can identify themselves with the product and color preference is positively correlated, (Pearson Correlation .566, p = .000), and the extent to which people can identify themselves with the product and their purchase intentions are positively correlated (Pearson Correlation .678, p = .000), and finally the extent to which people can identify themselves with the product and the price they are willing to pay is positively correlated (Pearson Correlation .518, p = .000). As a result, H7 is accepted. Also see

Appendix 2.4 for these results.

Finally, the Attributes Theory mentioned in the theoretical framework also suggested an indirect effect of uniqueness on product preference. More specifically, it was suggested that uniqueness could also spill over to other product attributes (availability, image, appearance, quality, and price) and so indirectly influence product preference. In order to test this spillover effect, mediation of the variable uniqueness should be analyzed. However, in order to test for mediation, 4 requirements have to be fulfilled (Baron and Kenny, 1986). The first requirement is that the independent variable should significantly influence the dependent variable. Second, the independent variable also should have a significant influence on the mediator. Third, this mediator should have a significant effect on the dependent variable when being controlled for the effect of the independent variable. Fourth, the remaining effect of the independent variable should decrease or disappear when the mediator is included in the analysis. Unfortunately, as already shown in current research, the independent variable uniqueness does not have a significant influence on the dependent variable product preference. As a result, it is unfeasible to test the mediator effect, and hypothesis H10 can be rejected.

The Color Brown, Associations, Uniqueness, and Product Preference

(32)

preference, this part of the research will discuss the effect of the different types of color names of

one color on product preference. By doing so, the variance becomes smaller and consequently, it

is more likely to find a significant difference between the different types of color names.

First of all, three Repeated Measures ANOVA’s were performed, in order to find out whether or not there is a significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common brown color names on product preference. However, despite the smaller variance, again no significant difference was found between the effect of the different types of brown color names on color preference F(2,87) = 1.544, p = .222, purchase intentions F(2,87) = 1.321, p = .275, and price willingness to pay F(2,87) = 0.568, p = .570. Consequently, again H1, H2, H4, H5, H8, and H9 can be rejected, and H3 can be accepted. For

more extensive results, see Appendix 2.5.

Second, the effect of the different kinds of brown color names on the activation of positive categories was studied. To see whether or not there is a significant difference between the effect of positively, negatively and neutrally associated brown color names on the activation of positive categories, a Repeated Measures ANOVA test was performed. However, again this test illustrated that on average neither positively, negatively, nor neutrally associated color names activate a positive category (respectively M = 3.133, M = 2.850, and M = 2.967) and also that they do not significantly differ F(2,87) = .389, p = .680. As a result, again H6 can be rejected. (Also see

Appendix 2.5)

(33)

they are willing to pay is positively correlated (Pearson Correlation .347, p = .000). As a result, again H7 can again be accepted. Also see Appendix 2.5 for these results.

Finally, the indirect effect of uniqueness on product preference as hypothesized by the Attributes Theory could not be tested for mediation, as the independent variable uniqueness does not have a significant influence on the dependent variable product preference, resulting in the rejection of H10.

The Color Blue, Associations, Uniqueness, and Product Preference

In order to test whether there is a significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common blue color names on product preference, three Repeated Measures ANOVA tests were performed. However despite the suggested differences by the Attention Theory, the Association Theory, and the Attributes Theory, again no significant differences were found between the effect of the different types of blue color names on color preference F(2,87) = .501, p = .608, purchase intentions F(2,87) = .095, p = .909, and price willingness to pay F(2,87) = .379 , p = .686. As a result, again H1, H2,

H4, H5, H8 and H9 can be rejected, and again H3 can be accepted. (Also see Appendix 2.6).

Furthermore, the effect of the different types of blue color names, –positively associated unique, negatively associated unique, and neutrally associated common-, on the activation of positive categories and in turn on product preference was studied. By conducting a Repeated Measures ANOVA test, it became clear that also the different types of blue color names on average do not activate positive categories (respectively M = 3.600, M = 3.267, and M = 3.317), and in addition, the effect of the different types of blue color names also do not significantly differ from each other F(2,87) = .480, p = .621. Consequently, it can be concluded that also just for the blue color names, compared to the negatively associated unique and the neutrally associated common color names, the positively associated unique color names do not lead to the activation of positive categories and thus product preference more than the other color names do. As a result, again H6

(34)

Additionally, in order to investigate whether the extent people can identify themselves with a blue color name and its influence on product preference, a Correlation Analysis has been performed. Again a significant positive relation has been found, which means that the more a consumer can identify him or herself with a color name, the higher the color preference (Pearson .405, p = .000), purchase intentions (Pearson .516, p = .000), and price willingness to pay (Pearson .455, p = .000). Consequently, again H7 can be accepted. (Also see Appendix 2.6)

Finally, again the indirect effect of uniqueness on product preference as hypothesized by the Attributes Theory could not be tested for mediation, because the independent variable uniqueness does not have a significant influence on the dependent variable product preference, resulting in the rejection of H10.

The Color Green, Associations, Uniqueness, and Product Preference

Finally, also for the green color names three Repeated Measures ANOVA’s were performed in order to evaluate whether or not there is a significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common green color names on product preference. Again the results showed that also for the green color names there is no significant difference between the different types of color names on color preference

F(2,87) = .775, p = .465, purchase intentions F(2,87) = .145, p = .866, and price willingness to

pay F(2,87) = .286, p = .752. As a result, again H1, H2, H4, H5, H8, and H9 can be rejected, and H3

can again be accepted. For more extensive results see Appendix 2.7.

(35)

activation of positive categories. F(2,87) = .677, p = .512. Therefore, again H6 can be rejected.

(Also see Appendix 2.7)

Furthermore, in order to investigate whether the extent people can identify themselves with a green color name affects product preference, a Correlation Analysis has been performed. Also for the green color names a significant positive relation has been found. In other words, the more a consumer can identify him or herself with a color name, the higher the color preference (Pearson .767, p = .000), purchase intentions (Pearson .856, p = .000), and price willingness to pay (Pearson .698, p = .000), and therefore again H7 can be accepted. (Also see Appendix 2.7)

(36)

Post Study to Rule Out the influence of Survey Phrasing

As in this research no significant difference has been found between the effect of positively associated unique, negatively associated unique, and neutrally associated common color names on product preference, the results of this study deviate from previous studies. One of the possible explanations why the results of the current study deviate from previous studies might be the phrasing used in the questionnaires. The questionnaire used for current research displays a picture of a towel accompanied with the color name. Subsequently, in the questions below, the color name is not mentioned again, however the phrasing ‘the depicted towel above’ is used. This is done on purpose, as in real life when purchasing a towel the color name is also stated just on for example the price tag. In other words, for current study it was deliberately chosen not to put more focus on the color name than would happen in real life. However, the risk of the used phrasing is that respondents base their answers more on the picture, and more or less overlook the color name.

In order to rule out the phrasing used in the questionnaires as the explanation for the deviation of current results with results from previous research, a post study will be done. For this post study, the questionnaires used in the main study will be rephrased, so that the wording ‘the depicted towel above’ is replaced by the color name of the towel. For an example of the questionnaire, see Appendix 3.1.

Post Study Results

Associations, Uniqueness, and Product Preference

Just as in the main study, again all 9 versions of the survey, so all three colors brown, blue and green, were distributed. A total of 28 respondents completed the questionnaires, of which 27 proved to be useful (1 survey was not useful due to not fully completing the questionnaire). Of these 27 respondents, 10 were male and 17 were female. The average age of the sample is 46.

(37)

First of all, just as in the main study, the overall effects of positively associated unique, negatively associated unique, and neutrally associated common color names on product preference will be discussed. In other words, no distinction will be made between the effect of brown, blue, and green color names on product preference.

In order to study the effect of the different types of color names on product preference, three Oneway ANOVA tests were performed, and the results showed again that there is no significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common color names on purchase intentions F(2,78) = 2.181, p = .120, and price willingness to pay F(2,78) = 1.080, p = .345. However, when employing an alpha level of .10, the results also showed that there is in fact a marginally significant difference between the effect of the different types of color names on color preference F(2,78) = 2.631, p = .078,. More specifically, by conducting a Post hoc test it became clear that the color of the product with the positively associated color name (M = 4.70) is preferred more than the color of the product with the negatively associated color name (M = 3.48), p = .027. Consequently, this could imply that -when investigated on a larger scale and with the right extent of color name exposure- associations influence color preference and thus in turn product preference.

As a result, again H1, H2, H4, H5, and H8 can be rejected and now H3 and H9 can be partially

rejected. For more extensive results about the ANOVA tests, see Appendix 3.2.

Second, the effect of the different types of color names, -positively associated unique, negatively associated unique, and neutrally associated common-, on the activation of positive categories and in turn on product preference was evaluated. In order to do so, first just as in the main study, a Cronbach’s Alpha was calculated in order to asses whether or not the two questions from the survey about the activation of positive categories could be combined. As the Cronbach’s Alpha turned out to be .902, the two variables could indeed be combined, and by conducting a Oneway ANOVA test, it became clear that in contrary to the main study, a significant difference between the different types of color names on the activation of positive categories exists, F(2,78) = 3.501,

p = .035. Next, in order to reveal which types of color names differed from each other, a Posthoc

(38)

with negatively associated color names (M = 3.185) are, and products with neutrally associated color names (M = 3.907) did not differ from products with positively nor negatively associated color names on the activation of positive categories. As a result, it can be concluded that positively associated color names indeed slightly activate positive categories, however, not significantly more than neutrally associated common color names do. Also, from the ANOVA test in the previous paragraph it can be derived that even though the positively associated color names slightly activate a positive category, this does not lead to a significantly different effect on product preference compared to negatively and neutrally associated color names. As a result, again H6 cannot be accepted. (See Appendices 3.2 and 3.3 for more extensive results).

Third, in order to investigate whether or not the extent people can identify themselves with a color name and its influence on product preference, a Correlation Analysis has been performed. From this analysis, it became clear that just as in the main study, a significant positive relationship exists. In other words, the better people can identify themselves with a certain color name, the higher the color preference (Pearson Correlation .762, p = .000), purchase intentions (Pearson Correlation .783, p = .000) and price willingness to pay (Pearson Correlation .513, p = .000). As a result, also H7 can again be accepted. For more extensive results of the Correlation

Analysis, see Appendix 3.4.

Finally, as in the main study, the indirect effect of uniqueness on product preference as hypothesized by the Attributes Theory could not be tested for mediation, as the independent variable uniqueness does not have a significant influence on the dependent variable product preference.

The Color Brown, Associations, Uniqueness, and Product Preference

(39)

First, three Repeated Measures ANOVA’s were performed to see whether or not there is a significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common brown color names on product preference. However, as in the main study, again no significant difference was found between the effect of the different types of brown color names on color preference F(2,24) = .370, p = .696, and purchase intentions

F(2,24) = .288, p = .753. On the other hand, when employing an alpha level of .10, current results

did show a marginally significant difference between the effect of the different types of color names on price willingness to pay F(2,24) = 2.890, p = 0.085. More specifically, by conducting a Post hoc test it became clear that respondents were willing to pay significantly more money for the product with the neutrally associated common color name (M = 5.77) than they were for exactly the same product with the negatively associated unique color name (M = 4.77). No significant difference was found between the effect of positively associated unique color names (M = 5.44) and negatively or neutrally associated color names on price willingness to pay.

As a result, again H1, H2, H4, and H8 can be rejected, while H5 can be partially accepted and H3, ,

and H9 can be fully accepted. (Also see Appendix 3.5).

Second, the effect of the different types of brown color names on the activation of positive categories and consequently on product preference was studied by conducting a Repeated Measures ANOVA. This test showed that just as in the main study, there is no significant difference between positively associated unique brown color names (M = 5.17), negatively associated unique brown color names (M = 3.78) and neutrally associated brown common color names (M = 4.56) on the activation of positive categories F(2,24), p = .337, resulting again in the rejection of H6.

(40)

price willingness to pay was found, this positive relation was not significant (Pearson Correlation .361, p = .064). As a result, H7 can only be partially accepted. (Also see Appendix 3.5)

Finally, also for the brown color names, the indirect effect of uniqueness on product preference could not be tested for mediation, as the independent variable uniqueness does not have a significant influence on the dependent variable product preference.

The Color Blue, Associations, Uniqueness, and Product Preference

Just as for the brown color names, also the effect of the different types of blue color names on product preference has been studied in this post study. More specifically, three Repeated Measures ANOVA’s have been performed to find out whether or not there is a significant difference between the effect of positively associated unique, negatively associated unique, and neutrally associated common blue color names on product preference. Just as in the main study, this post study shows that there is no significant difference between the effect of the different types of blue color names on color preference F(2,24) = 1.396, p = .276, and purchase intentions

F(2,24) = 2.141, p = .150. However, this post study did find a marginally significant difference

between the effect of the different types of color names on price willingness to pay F(2,24) = 3.173, p = .069. More specifically, a Post hoc test showed that respondents were willing to pay significantly more for products with a positively associated color name (M = 6.27) than they were for exactly the same product with a negatively associated color name (M = 4.11), p = .085. As a result, H1, H2, H4, H5, H8, and H9 can be rejected and H3 can be partially rejected. (Also see

Appendix 3.6).

(41)

Also, the relationship between the extent people can identify themselves with a blue color name and product preference has been studied. By conducting a Correlation Analysis it became clear that, just as in the main study, the more people can identify themselves with a blue color name, the higher the color preference (Pearson Correlation .759, p = .000), the higher the purchase intentions (Pearson Correlation .728, p = .000), and the higher the price willingness to pay (Pearson Correlation .612, p = .001). As a result, again H7 can be accepted. (Appendix 3.6)

Finally to conclude the analysis for the blue color names, again the indirect effect of uniqueness on product preference as hypothesized by the Attributes Theory could not be tested for mediation, as the independent variable uniqueness does not have a significant influence on the dependent variable product preference.

The Color Green, Associations, Uniqueness, and Product Preference

Last, the effect of the different types of green color names, -positively associated unique, negatively associated unique, and neutrally associated common-, on product preference was studied. First of all, the influence of associations and uniqueness of green color names on product preference was examined, by conducting three Repeated Measures ANOVA’s. Just as in the main study, no significant difference was measured between the different types of green color names on color preference, F(2,24) = 2.266, p = .136, purchase intentions F(2,24) = 1.773, p = .202, and price willingness to pay F(2,24) = .582, p = .570, leading again to the rejection of H1, H2, H4, H5,

H8, and H9, and the acceptance of H3 (Also see Appendix 3.7). However, it should be noted that

even though the differences between the different types of green color names are not significant, these differences are fairly larger than the differences measured in the main study.

(42)

there was a significant difference found during the post study, F(2,24) = 5.305, p = .017. More specifically, by conducting a Post hoc test it became evident that products with positively associated color names (M = 4.44) and products with neutrally associated common color names (M = 4.17) are more likely to activate positive categories than products with negatively associated color names (M = 2.44) (Appendix 3.7). However, as positively associated color names only slightly activate positive categories, and this research also already has shown that positively associated color names do not lead to a more positive product evaluation than negatively and neutrally associated color names do, still H6 cannot be accepted.

Furthermore, the relation between the extent people can identify themselves with a green color name and product preference was evaluated. By conducting a Correlation Analysis, it became clear that, just as in the main study, there was a significant positive relation. More specifically, the more people could identify themselves with a green color name, the higher the color preference (Pearson Correlation .734, p = .000), the higher the purchase intentions (Pearson Correlation .676, p = .000), and the higher the price willingness to pay (Pearson Correlation .607,

p = .001), resulting in the acceptance of H7. For more extensive results of this Correlation

Analysis, see Appendix 3.7.

The Color Naming Effect: How Fancy Color Names Influence Product Preference

The Color Naming Effect: How Fancy Color

Names Influence Product Preference

Master Thesis

Ilona Idserda

The Color Naming Effect: How Fancy Color

Names Influence Product Preference

Management Summary

Preface

Table of Contents