• No results found

Just pink for the profit? Examining the effect of direct and indirect pinkwashing attributions on consumers’ perceptions

N/A
N/A
Protected

Academic year: 2021

Share "Just pink for the profit? Examining the effect of direct and indirect pinkwashing attributions on consumers’ perceptions"

Copied!
48
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Just pink for the profit?

Examining the effect of direct and indirect pinkwashing attributions on consumers’ perceptions

Sjoerd Tomassen Student number: 10878807

Master’s Thesis

Graduate School of Communication Master’s programme Communication Science

Corporate Communication track University of Amsterdam Supervisor: dr. A.C. Kroon

Word count: 7.495 words June 20, 2020

(2)

Abstract

Although greenwashing and its effects on consumers have become well-researched in the last few decades, research on greenwashing and “washing” practices is confined only to the context of environmental responsibility, virtually ignoring other CSR dimensions such as social and economic responsibility. This study aims to address this research gap by extending research on “washing” practices to the domain of social responsibility, specifically, by looking at a “washing” practice related to LGBTQ+-friendliness called pinkwashing. The study examines the effect of directly and indirectly attributing pinkwashing to an organization on consumer perceptions of that company, and the extent to which this relationship is

moderated by the tolerance the consumer has towards LGBTQ+ individuals. An online experiment (N = 203) was executed to examine this relationship, using a 2x2 between-subjects factorial design, with factors direct attribution of pinkwashing (present/not present) and indirect attribution of pinkwashing (present/not present). Analyses of the results could not confirm any effect of attributing pinkwashing to a company, neither directly or indirectly, to the consumer perception of that company. In addition, LGBTQ+ tolerance was not found to moderate this relationship, although a significant positive relationship between LGBTQ+ tolerance and consumer perceptions was found separate from the experimental conditions. This result indicates that the effects of pinkwashing messages are not equal to the effects of greenwashing messages, and therefore not all “washing” practices have equal effects, as is often assumed. Further research is necessary to examine what factors make pinkwashing different than greenwashing, and whether these different factors also impact “washing” practices in other domains of corporate responsibility.

(3)

Introduction

Since a few decades, many large organizations have moved away from the idea of existing just to create profits and shareholder value, and have started to incorporate a meaningful contribution to society in their bottom line. In the last decade, as the idea of a Corporate Social Responsibility (CSR) and the accompanying responsibilities have become commonplace in larger organizations, there has also been an increase in companies who attempt to use the CSR principles to better their reputation, without actually changing their business processes, especially in regards to sustainability. Such a practice, of companies misleading consumers regarding their environmental consciousness, is referred to as

greenwashing. Greenwashing can be defined as “the act of misleading consumers regarding the environmental practices of a company (firm-level greenwashing) or the environmental benefits of a product or service (product-level greenwashing)” (Delmas & Burbano, 2011, p.66). When companies are accused of greenwashing, it suggests that their CSR

communication is just a token effort (Laufer, 2003).

As the interest in greenwashing and its effects has increased, the idea that (perceived) greenwashing negatively influences consumer perceptions has become well-established. Most evidently, it has been found that when companies are accused of greenwashing, consumers’ cynicism and mistrust in the company increases (Jahdi & Acikdilli, 2009; Wood, 2015; Chen & Chang, 2013). Also, consumer purchase intentions are lower for “greenwashed” products than for regular products (De Jong, Harkink & Barth, 2018). The influence of greenwashing does not seem to be limited to the companies directly accused of greenwashing themselves; Du (2015) found that when companies are accused of greenwashing, the market will also respond negatively to other companies within the same industry.

(4)

encapsulates the environmental dimension of the CSR field. Van Marrewijk (2003) has noted that Corporate Social Responsibility (CSR) and the related field of Corporate Sustainability (CS) traditionally have a bias towards environmental policies. Similarly, research on “washing” practices seems to be confined to the context of environmental responsibility, without looking at the other dimensions of CSR, such as social and economic responsibility (Ginder, 2016). Often, greenwashing is presented as a direct antithesis to CSR (e.g. De Jong et al., 2018). This is based on an unproven assumption, namely, the idea that the effects of a company presenting itself as more environmentally responsible than it really is will be the same as the effects of a company presenting itself as more socially or economically responsible than it really is.

Within the subfield of social responsibility, one term has recently emerged to describe companies presenting themselves as more LGBTQ+-friendly than they are, namely,

pinkwashing. Companies that have faced accusations of pinkwashing in the media include Primark, H&M, Uber, Coca Cola, and Shell (Appels, 2019; De Bruijn, 2020). However, although the media might have already embraced pinkwashing as a concept, there is close to no academic literature available about the subject. With media coverage of pinkwashing continually increasing, it is important to provide theoretical insights into the workings of pinkwashing and its effects. Ginder and Byun (2015) claim that there is currently a research gap regarding the effects of pinkwashing messages on consumers. This study aims to tackle this research gap and find whether the already existing findings on the effects of

greenwashing messages on consumers can be extended to pinkwashing messages as well. It is currently unclear whether attributing pinkwashing to a company will influence consumers in the same way that attributing greenwashing to the company would, i.e.,

increase cynicism and mistrust and lower brand attitudes and purchase intentions. In addition, it is also unknown whether attributing pinkwashing to a company will have the same effect

(5)

on other companies in the same industry that greenwashing attributions are likely to have. It is worthwhile to check whether such forms of indirect attribution of pinkwashing to a company will similarly lead to negative consumer perceptions. Finally, although various studies have been published regarding the effects of companies presenting themselves as LGBTQ+-friendly on LGBTQ+ and LGBTQ+-friendly consumers (e.g., Tuten, 2006), these studies have also not examined whether this process is influenced by pinkwashing

attributions. This makes it unclear whether consumers are influenced by whether a company is actually LGBTQ+-friendly, or just presenting as such. In this study, these knowledge problems will be examined using an experimental design looking at the effects of media messages in which pinkwashing is attributed to companies, both directly (where the company is being accused) and indirectly (where the company’s industry is being accused) on

consumers’ perceptions. Using this experiment, the various workings of attributions of pinkwashing can be examined, and it can be established whether attributing pinkwashing to a company leads to comparable effects on consumers’ perceptions as greenwashing has been established to do. These study aims lead to the following research questions:

RQ1 - To what extent do media messages in which pinkwashing is directly and/or indirectly attributed to a company influence consumer perceptions of that company?

RQ2 – To what extent is the effect mentioned in RQ1 moderated by consumers’ tolerance towards LGBTQ+ individuals?

Answering these two research questions will have both scientific and societal relevance. In terms of scientific relevance, answering these research questions will help to address the research gap on other forms of “washing” than greenwashing. Determining the

(6)

strengthen existing knowledge on greenwashing and help extend it to other fields.

Furthermore, it will prove to what extent “washing” is a process with a universal reaction, or whether it is dependent on the topic in regards to which the washing takes place. In addition, the outcomes of this study can be used to help organizations who aim to present themselves as LGBTQ+-friendly to better predict the consequences of their policies and help clarify the consequences of inconsistencies within these policies. Finally, the results of the study will help LGBTQ+ interest groups who aim to bring attention to organizational pinkwashing practices; gaining better insight into how consumers react to these practices will help in formulating an efficient response in order to prevent them.

Theoretical background Attribution theory and greenwashing

Much of the multidisciplinary research on the effects of (perceived) greenwashing messages on consumers grounds itself in attribution theory (e.g. Parguel, Benoît-Moreau & Larceneux, 2011, Nyilasi, Gangadharbatla & Paladino, 2014, De Jong et al., 2018). Rooted in the field of psychology, attribution theory poses that people always aim to understand why observed behaviors are happening by making attributions regarding which causal factors the behaviors can be ascribed to (Lin-Hi & Blumberg, 2018). Attribution theory can be used as a framework to understand how consumers process messages related to greenwashing,

especially when looking at negative attributes and skepticism (Nyilasy et al., 2014). Parguel et al. (2011) distinguish two types of causal attributions people can make based on CSR communications: attributions to the dispositions of the company and attributions to environmental factors, i.e., intrinsic or extrinsic motives. If a person makes a causal

attribution of extrinsic motives to a company, it can be seen as a distrust of the true intentions of the company. Skarmeas and Leonidou (2013) found that when consumers perceive CSR

(7)

efforts to be disingenuous or opportunistic, they will assign extrinsic motives to the CSR efforts, which leads to increased doubt and skepticism.

Attribution theory can similarly explain how greenwashing affects consumers. When a consumer is confronted with information that shows them that the actual environmental efforts by the company are not aligned with how they communicate about sustainability, this will lead the consumer to question the motives of the company. When these motives are questioned, this will lead to an attribution of extrinsic motives. This, in turn, leads to more negative consumer perceptions.

The idea that greenwashing leads to an attribution of extrinsic motives to the company, which leads to more negative consumer perceptions, has been tested and

demonstrated in various experimental studies. For example, Parguel et al. (2011) found that, for communicated CSR efforts to have a positive effect on how consumers evaluate a brand, the consumer would need to ascribe intrinsic motives to the company in question. They found that verified sustainability ratings had a significant effect on whether the consumer perceived intrinsic motives, which in turn affected how they evaluated brand performance.

Many of the studies on the effect of greenwashing on consumer perceptions focus on purchase intention. In general, these studies show that purchase intentions decrease when consumers are given proof that a company is greenwashing. For example, Nyilasy et al. (2014) ran an experiment in which participants within a “low environmental performance, but with green advertising” condition (i.e., greenwashing) had more negative purchase intentions and brand attitudes than in similar conditions where a company had low environmental performance but did not advertise greenly, either. Their study shows that, when the actual environmental performance is the same, a message which mentions false information

(8)

Finally, De Jong et al. (2018) performed an experiment which showed that messages about companies that make environmental claims while having a low environmental

performance lead to a lower purchase interest for consumers than messages about

organizations that have a high environmental performance do. Based on this finding, De Jong et al. (2018) also emphasize that the only way that communicating about green efforts as an organization actually has a positive effect on the consumer is when the organization is not being disingenuous; honesty and sincerity are fundamental.

The difference between greenwashing and pinkwashing

Although there exists quite some empirical evidence regarding the effect of greenwashing messages on consumers, the existing literature on pinkwashing is limited. Ginder and Byun (2015) have published an extensive literature review of most literature on the characteristics of the gay and lesbian consumer market. One of the existing gaps in the research they identify is that pinkwashing, and its effect on consumers, had not been

researched at all at their time of writing. Since only one such article has been published since (namely, Ginder, 2016), the research gap on pinkwashing remains. As no formal definition of pinkwashing exists, it can be created by looking at the definition of greenwashing of Delmas and Burbano (2011). Using this definition, pinkwashing can be defined as the act of a

company misleading consumers regarding the extent to which the company is actually supportive of the LGBTQ+ community. Tuten (2006) sees the “gay-friendliness” of

companies as a dynamic interplay between an internal dimension (e.g., progressive employee policies) and external dimension (e.g., using gay imagery in marketing and supporting gay causes). The author notes here that simply showing “gay-friendliness” in the external dimension is not enough for the average gay consumer: companies need to make

contributions in both the internal and external dimensions and be behaviorally consistent to achieve moral legitimacy in the gay consumers’ eyes. If a company would contribute to the

(9)

external dimension, but not the internal dimension, it might be viewed as an insincere attempt to appeal to the target audience without giving the proper support expected. This process, like greenwashing, again comes back to the importance of being honest and consistent.

As noted, there has been little to no research which examines the effects of

pinkwashing messages on consumers’ perceptions. However, it can be noted that honesty and sincerity play a similarly crucial role for both consumers evaluating companies with green messaging (Nyilasy et al, 2014), as well as gay consumers evaluating the “gay-friendliness” of a company (Tuten, 2006). Therefore, it can be expected that the mechanism underlying both greenwashing and pinkwashing is likely similar. Thus, the hypotheses posed for this study also mirror the effects of greenwashing mentioned above, and are in line with existing assumptions about washing practices.

First, it is expected that attributing pinkwashing to a company will lead to a more negative perception of the company. Specifically, it is expected that attributing pinkwashing will lead to more consumer skepticism, a lower purchase intention and a more negative brand attitude than not attributing pinkwashing. Within the existing literature on greenwashing, greenwashing is not always operationalized consistently. Some studies operationalize greenwashing by attributing it directly to a company (i.e., a company is accused of

pinkwashing, used by Parguel et al., 2011, amongst others), whilst others operationalize it by attributing it indirectly (i.e., by mentioning how greenwashing exists within the industry in which the company operates whilst not directly mentioning a link with the company, used by Rahman et al., 2013, amongst others). To test whether these differences in operationalization impact the underlying internal mechanism, both were included in the study. Thus, for this study, attributed pinkwashing is conceptualized in two ways: it can either be attributed to an organization directly or indirectly. Since the effects of greenwashing are similar in most

(10)

both attributional methods will lead to similar effects. Furthermore, it is expected that they will be more pronounced when both types of attribution are made together, as this will likely reinforce the message. This leads to hypotheses 1, 2 and 3.

H1 – Directly attributing pinkwashing to a company causes more negative consumer

perceptions (i.e. a. higher skepticism, b. lower purchase intention and c. lower brand attitude) of that company than not directly attributing pinkwashing to a company.

H2 – Indirectly attributing pinkwashing to a company causes more negative consumer

perceptions (i.e., a. higher skepticism, b. lower purchase intention and c. lower brand attitude) of a company than not indirectly attributing pinkwashing to a company.

H3 – The negative effect of direct pinkwashing attribution to a company on a. higher skepticism, b. lower purchase intention and c. brand attitude will be more pronounced when pinkwashing is also indirectly attributed to that company than when pinkwashing is not also indirectly attributed to the company.

The moderating role of LGBTQ+ tolerance on pinkwashing effects

Although the basic effect of attributing pinkwashing to a company on consumers’ perceptions can be conceptualized similarly to greenwashing, an important difference

between the two concepts lies in the idea of LGBTQ+ tolerance. As Ginder and Byun (2015) note, within the research field of advertising, many studies support the idea that whether consumers are supportive of advertising featuring gay imagery is moderated by tolerance towards LGBTQ+ individuals; those who are less tolerant towards homosexuality will also be less likely to be positive towards advertising featuring gay imagery. Although they also note that advertising featuring implicit gay imagery (such as iconography) can evoke a positive

(11)

response from both gay and straight consumers, they still pose that tolerance of homosexuality is an important marker of how a message is received.

One possible explanation for the moderating role of LGBTQ+ tolerance in advertising lies in the idea of attitude strength. Howe and Krosnick (2017) note that one of the key factors in determining attitude strength for an individual is social identification. For an individual, attitudes can become more important through social identification with reference groups or reference individuals: if they know these groups or individuals have taken a strong stance on an issue, they are likely to take a similarly strong stance. Traditionally, the

LGBTQ+ community is a community with a strong social membership (Cox & Gallois, 1996), which would implicate that belonging to or personally knowing individuals in the LGBTQ+ community would lead to a strong attitude towards a certain issue if the LGBTQ+ community (and their allies) generally took a similar stance.

Although Ginder and Byun (2015) only discuss the effect of LGBTQ+ tolerance in the context of advertising, likely, this moderating effect will also extend to the effect of pinkwashing on consumers’ perceptions. As was mentioned previously, Tuten (2006) noted that companies would need to align their “gay-friendliness” in both their internal and external in order to achieve moral legitimacy in the eyes of LGBTQ+ consumers. Therefore, it can be posed that companies being inconsistent in their policies regarding LGBTQ+-friendliness will lead to those affected by these policies having more negative perceptions. In addition,

because social identification is strong for LGBTQ+ individuals and their allies, likely, this will positively affect the strength of the attitude. Individuals who are in or who are connected to the LGBTQ+ community will likely care more about a company being inconsistent in their policies towards LGBTQ+-friendliness than those who are not. Therefore, it can be expected that the effect of pinkwashing attribution on consumer perceptions is dependent on LGBTQ+

(12)

they will care about the company’s inconsistencies, and the more negative their perceptions of the company will be. This is reflected in hypotheses 4 and 5.

H4 – The negative effect of directly attributing pinkwashing on consumer perceptions (i.e., a. higher skepticism, b. lower purchase intention and c. lower brand attitude) will be more pronounced the more the consumer is tolerant towards LGBTQ+ individuals, as compared to being less tolerant towards LGBTQ+ individuals.

H5 – The negative effect of indirectly attributing pinkwashing on consumer perceptions (i.e., a. higher skepticism, b. lower purchase intention and c. lower brand attitude) will be more pronounced the more the consumer is tolerant towards LGBTQ+ individuals, as compared to being less tolerant towards LGBTQ+ individuals.

Methods Participants

To test the hypotheses formulated above, an experiment was undertaken using an online survey. Before recruitment for participants began, the study was preregistered on the Open Science Framework (OSF). Participants were recruited over a period of 27 days (April 18, 2020 to and including May 14, 2020). In total, 260 participants agreed to participate in the study. The participants were mainly gathered through convenience sampling using three methods. First, the primary source of participants was the personal network of the author. A link to the survey was distributed through various online platforms, social networking sites and email. Second, participants were also gathered through SurveySwap, an online

participant retrieval tool. Third, the survey was also sent to multiple LGBTQ+ interest groups to increase the number of LGBTQ+ participants in the study. Only one of the approached

(13)

interest groups (an LGBTQ+-oriented association with over 300 members) responded positively to the request and distributed the survey.

Of the 260 people who agreed to participate in the study, 57 needed to be removed. 47 participants were removed because they did not complete the survey. In addition, 5 more people were removed because they had also participated in the pre-test, and had therefore already been exposed to the stimulus materials. Finally, 5 participants who took more than half an hour to complete the survey were also removed. The final dataset consisted of 203 participants. The distribution of these participants over the different experimental conditions is given in Appendix C, Table 1. The participants’ median age was 23 (M = 25.03, SD = 8.76), with participants’ ages ranging between 18 and 80. Most of the participants identified as female (70.9%), with most of the other participants identifying as male (28.6%) – one respondent identified as non-binary. There was some variance between the education level of the participants: most had finished a form of higher education (82.3%), with the other

participants having all finished a form of lower education (17.7%). The final dataset contained 18 participants (8.9%) who self-identified as being LGBTQ+.

For this study, participants needed to be randomly assigned to the different

experimental conditions. To check whether this aim was successful, multiple randomization checks were performed for the measured demographics. Three Chi-square tests showed that there were no significant differences between the experimental conditions on gender, χ2 (6, N = 203) = 8.54, p = .201, education level, χ2 (15, N = 203) = 14.22, p = .509, and LGBTQ+ identity, χ2 (3, N = 203) = 1.57, p = .667. In addition, using a one-way ANOVA with the experiment condition as an independent variable and age as a dependent variable, it was found that there were no significant differences between the conditions regarding the

(14)

for all measured demographic variables, no demographic variables were added as covariates in the analyses for testing the hypotheses.

Design

The research design for this study was primarily based on previous research done on the effects of greenwashing on consumers. Specifically, the research done by Rahman et al. (2015) on the effects of connecting greenwashing to a company has on consumers’

perceptions of that company was used as a primary reference point. In their study, the authors conducted an experiment with two conditions (a condition where greenwashing was visible and a control condition). However, a limitation of their research is that they do not directly relate greenwashing to the fictitious company they present in their stimulus material (i.e., greenwashing is not directly attributed to the company). Since this study aims to broaden the operationalization of pinkwashing to account for both direct and indirect attributions of pinkwashing, the research design was modified by adding a second factor to the experiment, namely, whether pinkwashing is directly attributed to the company itself.

This leads to a 2x2 between-subjects factorial design. The first factor corresponds to whether pinkwashing is directly attributed to a specific company (with levels present/not present), the second factor corresponds to whether pinkwashing is indirectly attributed to a specific company (with levels present/not present). This design is graphically represented in Appendix A, Table 1.

Procedure

The experiment was distributed using an online survey, written in Dutch. In the survey, participants were given a small introduction with a summary of the project (not revealing the actual manipulations within the experiment) and a participation consent form. After this, they were asked a few questions regarding a few possible confounding variables, after which they were randomly placed in one of the four experimental conditions. In all of

(15)

the experiment conditions, the participants first saw the stimulus material for the study, which was a fabricated news article from a politically balanced Dutch news outlet, nu.nl. The news article concerned the launch of a new soft drink from a fictitious company. A news article was chosen as a stimulus material because it allows for giving more background information about the company, which also makes it possible to include both the direct and the indirect pinkwashing attribution. In addition, since most of the attention paid to pinkwashing so far has come from popular media, it is fitting to also test the effects of pinkwashing using a message from a news outlet. A politically balanced news source was chosen to prevent participants to think that the news organization would be overly biased or either left- or right-wing. To reduce artificiality and increase external validity, it was decided to use a real news organization, nu.nl, which is the largest digital news outlet in the Netherlands, reaching more than 7 million readers each month (Nu.nl, 2020). In the news article, it is written that the company aims to give half of the profits of sales of the soft drink to LGBTQ+ interest groups. A soft drink was chosen as a product because it is a neutral product for which most

participants are possible consumers. Furthermore, it was decided to make the soft drink and the company launching it both fictitious, to prevent previous brand attitudes about existing brands to influence the experiment. This also increases the internal validity of the experiment.

After participants were exposed to the experiment condition, the participants were asked to respond to measures of their perception towards the company and product they just saw. After this, they were asked a few questions as a manipulation check (see the section Measures). Finally, they were asked a few questions related to demographics, including questions about their sexual and gender identity, after which they were debriefed on the aim of the experiment and were informed that all of the stimulus materials were fictitious and manipulated for this study. The complete survey (in Dutch) is included in Appendix B.

(16)

Measures

Independent variables

The independent variables ‘direct pinkwashing attribution’ and ‘indirect pinkwashing attribution’ were operationalized by modifying the stimulus materials for each experimental condition.

Direct pinkwashing attribution. In the conditions where pinkwashing was directly attributed to the company, the second paragraph was modified to mention that the company has been directly linked to pinkwashing, along with a small definition of pinkwashing. In this paragraph, the direct link to pinkwashing is represented by a spokesperson of a fictitious LGBTQ+ interest group being quoted as saying that the accused company “does not

contribute meaningfully to the LGBTQ+ community other than with the limited visibility that the advertising campaign offers”.

Indirect pinkwashing attribution. In the conditions where pinkwashing was indirectly attributed to the company, the third paragraph was slightly modified to mention pinkwashing accusations within the industry, along with a small definition of pinkwashing in the condition where pinkwashing was not also directly attributed. In this paragraph, the pinkwashing accusations within the industry are represented by mentioning the real-life case of Coca-Cola, who received accusations of pinkwashing after their sponsorship of the Sochi Olympics (Gunther, 2013).

Manipulation check. To test the effectiveness of the manipulation, a pre-test was registered among 50 participants using a within-subjects design, with all participants being asked to what extent they perceived pinkwashing to directly attributed to the company or indirectly attributed to the company (using a 5-Point Likert scale). Out of the 50 participants, 38 participants completed the survey. Most of these respondents were female (71.1%). Except for 2 respondents, all respondents were 30 years or younger of age. Almost all of the

(17)

participants (92.1%) had completed a form of higher education. Using the participant dataset, two univariate ANOVAs were executed, with the manipulation check as a dependent variable and each of the factors as separate independent variables. The first analysis showed that exposure to articles where pinkwashing was directly attributed (M = 3.89, SD = 1.15) had a significant positive effect on whether the participant believed pinkwashing was directly attributed or not compared to exposure to articles where pinkwashing was not directly

attributed (M = 2.13, SD = 1.33), F(1,150) = 76.43, p = .000, ηp2 = .338. Similarly, the second analysis showed that exposure to articles where pinkwashing was indirectly attributed (M = 3.74, SD = 1.16) had a significant positive effect on whether the participant believed pinkwashing was indirectly attributed or not compared to exposure to articles where

pinkwashing was not indirectly attributed (M = 3.04, SD = 1.52), F(1,150) = 10.13, p = .002, ηp2 = .063. This indicated that the manipulation check was successful for both manipulated factors. However, the results did show that, in the condition in which pinkwashing was directly, but not indirectly attributed, respondents generally believed that pinkwashing was indirectly attributed. Based on these pre-test findings, the stimulus material for this condition was modified to further emphasize the direct attribution. Finally, the pre-test also indicated that participants generally thought the articles were both realistic (M = 3.63, SD = 0.80) and believable (M = 3.59, SD = 0.80), although they generally did not think they were as

neutrally written (M = 3.04, SD = 1.10).

In the final experiment, the manipulation check was also successful, with both a significant result for an ANOVA test using indirect attribution as an independent variable and the manipulation check as a dependent variable, F(1, 201) = 21.13, p = .000, ηp2 = .095, as well as a significant result for an ANOVA test using direct attribution as an independent variable, F(1, 201) = 6.21, p = 0.013, ηp2 = 0.030.

(18)

LGBTQ+ tolerance. The moderating variable ‘LGBTQ+ tolerance’ was measured using the operationalization of Collier et al. (2015), namely, by asking the participants about the extend to which they thought having an LGBTQ+ identity is acceptable, using a 5-point Likert scale (1 = completely unacceptable, 5 = completely acceptable).

Dependent variables

The study contains one dependent variable, namely ‘consumer perceptions’. Consumer perceptions were measured using three dimensions: skepticism towards the company, attitude towards the product and purchase intentions.

Skepticism. The skepticism construct was measured using a scale also used by Rahman et al. (2015), slightly modified to fit the context of this study. The scale includes four statements, which could be answered using a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). The scale includes statements like “The company’s claim about it

striving towards equality is true” and “The company’s claim about it striving towards equality is exaggerated”. An exploratory factor analysis using the principal axis factoring approach with a direct oblim rotation strategy indicated that the scale was one-dimensional, with the latent component explaining 57.36% of the variance in the four items. The factor analysis only looked at factor loadings of above .40. A reliability analysis showed that the scale was reliable, indicated by a Cronbach’s Alpha of .75, which could not be improved by removing any items from the scale. The final scale was computed by taking the mean of the four items (M = 4.14, SD = 1.09).

Product attitude. The product attitude was measured using the Ab scale as provided by Spears and Singh (2004). The Ab scale is a 7-Point Likert scale, in which respondents are asked to note what word describes their attitudes towards the product best, using 5

combinations of words. Example word combinations include “Unappealing/appealing”, “Bad/good” and “Unlikeable/likable”. A similar exploratory factor analysis using the

(19)

principal axis factoring approach with a direct oblim rotation strategy indicated that the scale was one-dimensional, with the latent component explaining 80.35% of the variance in the five items. A reliability analysis showed that the scale was very reliable, indicated by a Cronbach’s Alpha of .94 (M = 4.54, SD = 1.23), which could not be improved by removing any items from the scale. The final scale was computed by taking the mean of the five items.

Purchase intention. Similar to the product attitude construct, the purchase intention was also measured by using a scale provided by Spears and Singh (2004). The PI scale is also a 7-Point Likert scale that asks participants which word describes their purchasing intentions best, using 5 combinations of words. Examples of word combinations include “Very low purchase interest/very high purchase interest” and “Definitely do not intend to buy/Definitely intend to buy”. A final similar exploratory factor analysis using the principal axis factoring approach with a direct oblim rotation strategy indicated that the scale was one-dimensional, with the latent component explaining 89.98% of the variance in the five items. A reliability analysis showed that the scale was very reliable, indicated by a Cronbach’s Alpha of .97 (M = 2.87, SD = 1.45), which could not be improved by removing any items from the scale. The final scale was computed by taking the mean of the four items.

Analysis

In order to test the proposed hypotheses, one two-way MANOVA and two one-way MANCOVA’s were executed. MANOVA was deemed as the most suitable test to test hypotheses 1abc, 2abc and 3abc, because both independent variables used (the experimental conditions) are categorical, and there are multiple dependent variables. Testing the dependent variables together in the same model reduces the chance of Type 1 errors. Since the

moderating variable ‘LGBTQ+ tolerance’ is not categorical, but continuous, a MANOVA test could not be used here. Therefore, a MANCOVA was chosen as a method of analysis, with

(20)

pinkwashing attribution were tested together, along with their interaction effect (in order to answer hypotheses 1abc, 2abc and 3abc). Then, two separate MANCOVA’s were executed to test the effect of LGBTQ+ tolerance on both direct and indirect attribution (in order to answer hypotheses 4abc and 5abc).

Results

To get a general impression of the relationships between the measured variables, correlations were calculated between all variables measured in the study. Table 1 shows descriptive statistics and correlations among these variables.

Table 1

Descriptive statistics and correlations among the study variables

Variable Mean SD Min Max 1. 2. 3. 4. 5. 6.

1. Pinkwashing directly attributed - - - - - 2. Pinkwashing indirectly attributed - - - - 3. LGBTQ+ tolerance 4.60 0.80 1 5 - - - 4. Attitude towards the brand 4.54 1.23 1 7 -.11 .13 .33** - 5. Product purchase intention 2.87 1.45 1 7 -.03 .06 .22** .59** - 6. Skepticism towards the company 4.14 1.10 1 7 .10 .02 -.14* -.62** -.43** - *p < .05. **p < .01.

(21)

The effects of direct and indirect attribution on consumer perceptions

In order to test the extent to which attributing pinkwashing to a company, either directly or indirectly, causes more negative consumer perceptions of that company than not attributing pinkwashing to a company (H1abc, H2abc), and to test whether there is an interaction effect between these two attributions (H3abc), a two-way MANOVA was

executed. The MANOVA used direct attribution and indirect attribution as indirect variables and purchase intention, brand attitude and skepticism as direct variables. A Shapiro-Wilk test showed a significant departure from normality for brand attitude, W(203) = 0.96, p = .000 and for purchase intention, W(203) = 0.94, p = .000. Although all other assumptions of the

ANOVA were met, the violation of the multivariate normality assumption led to the use of Pillai’s trace as the analyzed measure for the MANOVA.

The MANOVA indicated the effect of direct attribution on the combined dimensions of the dependent variable was nonsignificant, Pillai’s Trace = .02, F(3, 197) = 1.20, p = .312. Further inspection of the individual between-subject effects also showed no significant effect of direct attribution on purchase intention, F(1, 199) = 0.13, p = .722, brand attitude, F(1, 199) = 2.59, p = .109, or skepticism, F(1, 199) = 2.13, p = .146. Therefore, there is no evidence to support that direct attribution of pinkwashing influences consumer perceptions, thus, hypothesis 1abc cannot be supported.

Second, the MANOVA did indicate a significant effect of indirect attribution on the combined dimensions of the dependent variable, Pillai’s Trace = .05, F(3, 197) = 3.28, p = .022. However, further inspection of the individual between-subjects effects did not show any significant effects of indirect attribution on the individual dimensions of purchase intention, F(1, 199) = 0.79, p = .375, brand attitude, F(1, 199) = 3.26, p = .072, or skepticism, F(1, 199) = 0.15, p = .146. Therefore, although there is an effect of indirect attribution on the combined

(22)

dimensions of consumer perceptions visible, the significance of this effect does not extend to the individual dimensions. Thus, hypothesis 2abc also cannot be supported.

Finally, the interaction effect of direct and indirect attribution on the combined dimensions of the dependent variable also was nonsignificant, Pillai’s Trace = .03, F(3, 197) = 1.79, p = .150. Further inspection of the individual between-subject effects also showed no significant effect of direct attribution on purchase intention, F(1, 199) = 2.42, p = .122, brand attitude, F(1, 199) = 0.54, p = .463, or skepticism, F(1, 199) = 0.59, p = .444. Therefore, there is no evidence to support that the effect of direct attribution of pinkwashing on consumer perceptions is stronger when pinkwashing is also indirectly attributed. Thus, hypothesis 3abc cannot be supported. All of the reported multivariate and univariate effects are summarized in Appendix C, Tables 2 and 3.

The (moderating) effect of LGBTQ+ tolerance

To test whether the effect of direct pinkwashing attribution on consumer perceptions is moderated by tolerance towards LGBTQ+ individuals, a one-way MANCOVA was conducted with direct attribution as an independent variable, purchase intention, brand attitude and skepticism as dependent variables and LGBTQ+ tolerance as a covariate. In order to measure the moderating effect proposed in hypothesis 4abc, the interaction effect between LGBTQ+ tolerance and the direct attribution was also included in the design of the MANCOVA. Because a Shapiro-Wilk test showed a significant departure from normality for LGBTQ+ tolerance, W(203) = 0.56, p = .000, along with the previously established

departures from normality for brand attitude and purchase analysis, Pillai’s trace was used as the analyzed measure. The MANCOVA showed that the interaction effect of direct

attribution and LGBTQ+ tolerance on the combined dimensions of the dependent variable was nonsignificant, Pillai’s Trace = .01, F(3, 197) = 0.51, p = .675. Thus, there is no evidence to support an interaction effect between LGBTQ+ tolerance and direct attribution, which

(23)

means that hypothesis 4abc cannot be supported. The complete reported multivariate tests are given in Appendix C, Table 4.

Whether the effect of indirect pinkwashing attribution on consumer perceptions is moderated by tolerance towards LGBTQ+ individuals was similarly tested using a one-way MANCOVA as the MANCOVA used for testing hypothesis 4abc, only using indirect attribution as an independent variable instead of direct attribution. The MANCOVA showed that the interaction effect of indirect attribution and LGBTQ+ tolerance on the combined dimensions of the dependent variable was nonsignificant, Pillai’s Trace = .01, F(3, 197) = 0.79, p = .500. Thus, there is no evidence to support an interaction effect between LGBTQ+ tolerance and indirect attribution, which means that hypothesis 5abc cannot be supported. The complete reported multivariate tests are given in Appendix C, Table 5.

Although both hypothesis 4 and 5 cannot be supported, the MANCOVA analyses did show that there is a significant direct effect of LGBTQ+ tolerance on the dependent variables in both the MANCOVA with direct attribution as an independent variable, Pillai’s Trace = .12, F(3, 197) = 8.52, p = .000, as well as in the MANCOVA with indirect attribution as an independent variable, Pillai’s Trace = .11, F(3, 197) = 7.82, p = .000. This finding seems to suggest that, although LGBTQ+ tolerance does not moderate the effects of pinkwashing attribution on consumer perceptions, it might directly influence consumer perceptions, regardless of whether pinkwashing was attributed or not. Three individual exploratory regression analyses using LGBTQ+ tolerance as an independent variable and the three dimensions of consumer perceptions as dependent variables individually seem to support this idea. When not controlling for direct and indirect attribution, LGBTQ+ tolerance explains a significant portion of variance in brand attitude, F(1, 201) = 24.50, p = .000, R2 = .109,

(24)

the more positive their brand attitudes and purchase intentions will be, and the lower their skepticism will be of a company presenting itself as LGBTQ+-friendly.

Conclusion and discussion

This study focused on the extent to which directly and indirectly attributing pinkwashing to a company influences the consumer perceptions of that company, and the extent to which this effect is moderated by consumers’ tolerance towards LGBTQ+ individuals. The results given above show that there is no support for the claim that either directly or indirectly attributing pinkwashing to a company influences the consumer perceptions of that company, nor that this effect is moderated by consumers’ tolerance towards LGBTQ+ individuals. However, tolerance towards LGTBQ+ individuals was found to have a significant positive effect on consumer perceptions of the (LGBTQ+-oriented) company presented in the study, regardless of whether pinkwashing was attributed or not.

Based on the results presented in this study, it becomes evident that not all forms of “washing” have the same effect on consumers. Although the negative effect of greenwashing messages on consumer perceptions has been examined and confirmed often (e.g. Parguel et al., 2011, Nyilasi et al., 2014, De Jong et al., 2018, Rahman et al., 2015), the findings

presented here show that for pinkwashing, the same effect cannot be established. This finding differs from what most current research assumes, namely, that “washing” has a universal reaction, and implicates that it rather is dependent on the subfield in which it takes place. Further research on greenwashing and other forms of “washing” practices should avoid making this unconfirmed assumption.

Besides these scientific implications, the preliminary findings of this study also have social implications. The fact that the preliminary results of this study suggest that

(25)

organizations to present themselves as being LGBTQ+-friendly. Even with the risk of being accused of pinkwashing for not doing enough, it seems that this has little influence on consumers, meaning that for organizations, there seems to be less risk in presenting

themselves as LGBTQ+-friendly as compared to presenting themselves as environmentally conscious. However, it is important to also emphasize the importance of honesty and sincerity as noted by Tuten (2006). Even if pinkwashing attributions seem to have a limited impact, it has been shown that consumers expect organizations to be honest in their approach in order to assign them moral legitimacy; organizations should behave accordingly. In

addition, the results indicate that LGBTQ+ interest groups who aim to bring attention to organizational pinkwashing practices are likely better off trying to resolve this with the organization themselves, without relying too heavily on involving the general public, as they seem to be mostly unaffected by attributed pinkwashing.

A possible explanation on the difference between the effects of greenwashing and pinkwashing on consumers lies in the idea of ambiguity. In this study, pinkwashing was attributed to a company, but the claim of pinkwashing was never truly substantiated – only little argumentation was given in the stimulus materials on why critics thought the company was pinkwashing, leaving some ambiguity. Within the field of attribution theory, ambiguity has been shown to have an impact on how attributions are made. Ajzen, Dalto and Blyth (1979) show that, when someone is asked to make an attribution about an actor, and they are given ambiguous information regarding the actor, they are more inclined to interpret the ambiguous information as being consistent and in line with any previous dispositions they had regarding the actor. This described process possibly explains the lack of impact of the pinkwashing attributions here, as well: if the organization says they are LGBTQ+-friendly, and some critics say they are not (though it is unclear whether this criticism is actually valid),

(26)

the previously established disposition and see the company as sincere in their actions. However, further research should be conducted to further examine this process and the impact of ambiguity on “washing” claims.

Ambiguity does not just have a likely impact on this study, but on the general concept of pinkwashing as well. Within the field of sustainability, lots of certifications, rankings and labels exist which measure the environmental practices of a company or the environmental benefits of their products, with more than 400 of such consumer-facing eco-labels operational around the world (Potts et al., 2014). However, in contrast, there are currently no generalized definitions or (consumer-facing) labels regarding companies’ diversity policies. As a

consumer, there exists more ambiguity in situations where a company is presenting itself as LGBTQ+-friendly than in situations where it is presenting itself as environmentally

conscious, in regards to the extent to which the claim is actually correct. Accusing a company of pinkwashing is not as clear-cut and able to be substantiated as most accusations of

greenwashing are. Parguel et al. (2011) show the significant impact of sustainability rankings on consumers’ perceptions of the company. Future research should be conducted regarding the conceptualization and operationalization of similar rankings for diversity and LGBTQ+-friendliness, and the impact of such rankings on consumers.

Limitations and future research

Besides the factor of ambiguity in the stimulus materials, the study presented here has some noteworthy limitations. First, only a small number of the participants self-identified as being LGBTQ+ (8.9%). Although LGBTQ+ identity was expected to be an important

moderator in the extent to which pinkwashing has an effect on the consumer, this claim could not be properly tested due to the limited amount of LGBTQ+ participants. This problem is not limited to this study; as Ginder and Byun (2015) note, it is difficult to get a representative sample of LGBTQ+ individuals, which is why many studies focused on LGBTQ+ issues rely

(27)

on convenience sampling, purposive sampling and snowball sampling. It is recommended to further examine the effects of pinkwashing on LGBTQ+ consumers specifically, to see whether LGBTQ+ identity influences the relationship.

Second, since all of the included participants were Dutch and most of the included participants were under 30, the used sample is not completely representative, limiting

external validity. It is currently uncertain to what extent age and cultural background have on LGBTQ+-related consumer issues; Ginder and Byun (2015) note this as another research gap. Therefore, as they say, looking into the differences between different age groups and groups of different cultural backgrounds warrants future attention in a study with a more diverse sample.

Finally, as was noted in the introduction, only one study was found which looks at other types of “washing” than greenwashing (namely, Ginder, 2016), which limited this research in the theoretical foundations available for the concept of pinkwashing. This is not as much a limitation to this study as it is evidence underlining the importance of future research into the concept of pinkwashing. As this study gives a first indication that not all “washing” practices are the same and have the same effect, it suggests that the CSR subfield of greenwashing should be looked at through a new lens. This way, “washing” practices can stop being defined by just their “green” counterpart, but rather, can be extended to represent the entire color spectrum.

Reference list

Ajzen, I., Dalto, C. A., & Blyth, D. P. (1979). Consistency and bias in the attribution of attitudes. Journal of Personality and Social Psychology, 37(10), 1871.

(28)

Parool. Retrieved from https://www.parool.nl/amsterdam/bedrijven-die-meeliften-met-pride-oprecht-of-pinkwashing~be82e83e/

Chen, Y. S., & Chang, C. H. (2013). Greenwash and green trust: The mediation effects of green consumer confusion and green perceived risk. Journal of Business Ethics, 114(3), 489-500.

Collier, K. L., Horn, S. S., Bos, H. M., & Sandfort, T. G. (2015). Attitudes toward lesbians and gays among American and Dutch adolescents. The Journal of Sex Research, 52(2), 140-150.

Cox, S., & Gallois, C. (1996). Gay and lesbian identity development: A social identity perspective. Journal of Homosexuality, 30(4), 1-30.

De Bruijn, M. (2020, January 17). Bedrijven, stop met lhtbi’ers gebruiken om je imago op te

vijzelen. De Volkskrant. Retrieved from

https://www.volkskrant.nl/columns-opinie/bedrijven-stop-met-lhbti-ers-gebruiken-om-je-imago-op-te-vijzelen~b4b302f5/ De Jong, M. D., Harkink, K. M., & Barth, S. (2018). Making green stuff? Effects of corporate

greenwashing on consumers. Journal of Business and Technical Communication, 32(1), 77-112.

Delmas, M. A., & Burbano, V. C. (2011). The drivers of greenwashing. California

Management Review, 54(1), 64-87.

Du, X. (2015). How the market values greenwashing? Evidence from China. Journal of

Business Ethics, 128(3), 547-574.

Ginder, W. (2016). An Attribution Theory Approach to Corporate Social Responsibility (CSR) Communication: Consumer Response to Consistency-Based CSR Positioning. Ginder, W., & Byun, S. E. (2015). Past, present, and future of gay and lesbian consumer

research: Critical review of the quest for the queer dollar. Psychology &

(29)

Gunther, M. (2013, September 17). Coke's Olympics backlash: The rising bar for corporate

action. The Guardian. Retrieved from http://www.theguardian.com/sustainable-business/coke-olympics-gay-rights

Howe, L. C., & Krosnick, J. A. (2017). Attitude strength. Annual Review of Psychology, 68, 327-351.

Jahdi, K. S., & Acikdilli, G. (2009). Marketing communications and corporate social responsibility (CSR): marriage of convenience or shotgun wedding?. Journal of

Business Ethics, 88(1), 103-113.

Laufer, W. S. (2003). Social accountability and corporate greenwashing. Journal of Business

Ethics, 43(3), 253-261.

Lin-Hi, N., & Blumberg, I. (2018). The link between (not) practicing CSR and corporate reputation: Psychological foundations and managerial implications. Journal of Business

Ethics, 150(1), 185-198.

Nyilasy, G., Gangadharbatla, H., & Paladino, A. (2014). Perceived greenwashing: The interactive effects of green advertising and corporate environmental performance on consumer reactions. Journal of Business Ethics, 125(4), 693-707.

NU.nl. (2020, April 21). Retrieved June 7, 2020, from https://www.dpgmediamagazines.nl/brand/nu-nl/

Parguel, B., Benoît-Moreau, F., & Larceneux, F. (2011). How sustainability ratings might deter ‘greenwashing’: A closer look at ethical corporate communication. Journal of

Business Ethics, 102(1), 15.

Potts, J., Lynch, M., Wilkings, A., Huppe, G., Cunningham, M., & Voora, V. (2014). The State of Sustainability Initiatives Review 2014: Standards and the Green

(30)

Rahman, I., Park, J., & Chi, C. G. Q. (2015). Consequences of “greenwashing” Consumers’ reactions to hotels’ green initiatives. International Journal of Contemporary Hospitality

Management, 27(6), 1054-1081.

Skarmeas, D., & Leonidou, C. N. (2013). When consumers doubt, watch out! The role of CSR skepticism. Journal of Business Research, 66(10), 1831-1838.

Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues & Research in Advertising, 26(2), 53-66. Tuten, T. L. (2006). Exploring the importance of gay-friendliness and its socialization

influences. Journal of Marketing Communications, 12(2), 79-94.

Van Marrewijk, M. (2003). Concepts and definitions of CSR and corporate sustainability: Between agency and communion. Journal of Business Ethics, 44(2-3), 95-105. Wood, M. E. (2015). Exploring consumers' experiences with corporate greenwashing.

(31)

Appendix A

Graphical representations of study framework

Figure 1. Graphical representation of the conceptual framework.

Table 1

Graphical Representation of the Factorial Design

Direct pinkwashing attribution Not attributed Attributed

Indirect pinkwashing

attribution

Not attributed No direct/no indirect attribution

Direct attribution, but no indirect attribution

Attributed

Indirect attribution, but no direct attribution

Both direct and indirect attribution Direct pinkwashing attribution Consumer perceptions LGBTQ+ tolerance Indirect pinkwashing attribution + + + - -

(32)

Appendix B

Complete distributed survey (in Dutch)

BLOCK 1 - FACTSHEET

T1 - Beste deelnemer,

Deze vragenlijst is onderdeel van een master afstudeeronderzoek aan de Universiteit van Amsterdam. In het onderzoek zult u een nieuwsartikel te zien krijgen over een nieuw frisdrankproduct. Na het zien van het nieuwsartikel zal u een aantal vragen worden gesteld over het nieuwsartikel en over het product wat daarin wordt besproken. Iedereen boven een leeftijd van 18 kan deelnemen aan deze vragenlijst. Het invullen van de vragenlijst duurt ongeveer 6 minuten.

Omdat deze research wordt uitgevoerd onder de verantwoordelijkheid van de ASCoR, Universiteit van Amsterdam, kunnen we u garanderen dat:

1. Uw anonimiteit wordt beschermd, en uw persoonlijke gegevens niet doorgegeven zullen worden aan derde partijen, tenzij u daar eerst expliciet toestemming voor geeft. 2. U kunt weigeren om deel te nemen in de studie of uw participatie onderbreken zonder

een reden hiervoor te geven. U kunt ook tot 24 uur na deelname in de studie uw toestemming om uw antwoorden en data te gebruiken in de studie intrekken.

3. Deelnemen aan de studie zal niet betekenen dat u wordt onderworpen aan risico’s of ongemak, de onderzoekers zullen u niet opzettelijk misleiden, en u zult niet worden blootgesteld aan expliciet beledigende materialen.

(33)

4. We zullen u, tot 5 maanden na het einde van de studie, kunnen voorzien van een researchverslag dat de algemene resultaten van de studie uitlegt. Voor meer informatie over de studie kunt u ten alle tijden terecht bij de projectleider Sjoerd Tomassen, sjoerd.tomassen@student.uva.nl.

Mocht u klachten of opmerkingen hebben over de gang van zaken van deze studie na deelname kunt u de aangewezen persoon van de Ethics Committee die ASCoR

vertegenwoordigt bereiken op het volgende adres: ASCoR Secretariaat, Ethics Committee, Universiteit van Amsterdam, Postbus 15792, 1001 NG Amsterdam; 020-525 3680; ascor-secr-fmg@uva.nl. Alle klachten en opmerkingen zullen met volledige vertrouwelijkheid worden behandeld.

Alvast bedankt voor het invullen van deze vragenlijst!

BLOCK 2 - CONSENT

IO1 - Bij dezen verklaar ik dat ik op een duidelijke manier ben geïnformeerd over de aard en methode van dit onderzoek, zoals verklaard op de vorige pagina van deze vragenlijst. Ik stem volledig en vrijwillig in met deelname aan dit onderzoek. Hiermee hou ik het recht om mijn toestemming terug te trekken, zonder hier een reden voor te hoeven geven. Ik ben me ervan bewust dat ik mijn deelname in dit experiment ten alle tijden mag stopzetten. Als mijn onderzoeksresultaten in wetenschappelijke publicaties zullen worden gebruikt of op een andere manier publiek worden gemaakt, dan zal dit op zo een manier gebeuren dat mijn anonimiteit beschermd blijft. Mijn persoonlijke data zal niet aan derde partijen worden gegeven tenzij ik hier expliciet toestemming voor geef. Als ik meer informatie zou willen

(34)

(sjoerd.tomassen@student.uva.nl) hiervoor benaderen. Mocht ik klachten of opmerkingen hebben over dit onderzoek dan kan ik de aangewezen persoon van de Ethics Committee die ASCoR vertegenwoordigt hiervoor bereiken op het volgende adres: ASCoR Secretariaat, Ethics Committee, Universiteit van Amsterdam, Postbus 15792, 1001 NG Amsterdam; 020-525 3680; ascor-secr-fmg@uva.nl.

o Ik begrijp de tekst hierboven en stem in met deelname aan dit onderzoek.

BLOCK 3 – POSSIBLE CONFOUNDING VARIABLES

MOD1 - Voordat u het artikel te zien krijgt willen we u eerst een aantal vragen stellen over uw meningen en koopgedrag. De eerstvolgende vragen gaan over frisdrank.

Hoe vaak drinkt u gemiddeld frisdrank? o Nooit

o Een keer per maand of minder o 2-3 keer per maandag

o Een keer per week

o Meer dan 1 keer per week

MOD2 - Geef alstublieft aan in hoeverre u in het algemeen geneigd bent om frisdrank te kopen (bijvoorbeeld wanneer u boodschappen doet).

Nooit ◯ ◯ ◯ ◯ ◯ ◯ ◯ Vast en zeker

Zeker niet van plan te kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Zeker van plan te kopen Lage interesse om te kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Hoge interesse om te kopen

(35)

Zal het zeker niet kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Zal het zeker kopen Zal het waarschijnlijk niet

kopen

◯ ◯ ◯ ◯ ◯ ◯ ◯ Zal het waarschijnlijk kopen

MOD3 - Geef voor de volgende stellingen aan in hoeverre u het er mee eens bent, op basis van uw eigen (koop)gedrag. Ik vind frisdrank...

Onbelangrijk ◯ ◯ ◯ ◯ ◯ ◯ ◯ Belangrijk

Niet relevant voor mij ◯ ◯ ◯ ◯ ◯ ◯ ◯ Relevant voor mij

Niet van betekenis ◯ ◯ ◯ ◯ ◯ ◯ ◯ Van betekenis

Iets wat ik niet nodig heb ◯ ◯ ◯ ◯ ◯ ◯ ◯ Iets wat ik nodig heb

MOD4 - Voor de laatste vraag zijn we benieuwd naar uw mening. Er is geen goed of fout antwoord. In de vraag wordt de afkorting LHBTQ+ gebruikt. Dit is een afkorting voor lesbisch, homoseksueel, biseksueel, transgender en queer.

In mijn mening is het hebben van een LHBTQ+-identiteit... o Helemaal onacceptabel

o Vrij onacceptabel

o Niet acceptabel, noch onacceptabel o Vrij acceptabel

(36)

CONDITION 1 – No direct, no indirect attribution (NDNI)

T3 - Hieronder staat een artikel van nu.nl. Lees dit artikel alstublieft zorgvuldig door. Als u deze vragenlijst op uw telefoon invult, kunt u het beste uw telefoon horizontaal houden om het artikel goed te kunnen lezen. Nadat u het artikel hebt gelezen zal u een aantal vragen worden gesteld over de inhoud van het artikel. De knop om naar de volgende pagina te navigeren wordt na 15 seconden getoond.

(37)

CONDITION 2 – Has direct attribution, but no indirect attribution (YDNI)

T4 - Hieronder staat een artikel van nu.nl. Lees dit artikel alstublieft zorgvuldig door. Als u deze vragenlijst op uw telefoon invult, kunt u het beste uw telefoon horizontaal houden om het artikel goed te kunnen lezen. Nadat u het artikel hebt gelezen zal u een aantal vragen worden gesteld over de inhoud van het artikel. De knop om naar de volgende pagina te navigeren wordt na 15 seconden getoond.

(38)

CONDITION 3 – Has no direct attribution, but does have indirect attribution (NDYI)

T5 - Hieronder staat een artikel van nu.nl. Lees dit artikel alstublieft zorgvuldig door. Als u deze vragenlijst op uw telefoon invult, kunt u het beste uw telefoon horizontaal houden om het artikel goed te kunnen lezen. Nadat u het artikel hebt gelezen zal u een aantal vragen worden gesteld over de inhoud van het artikel. De knop om naar de volgende pagina te navigeren wordt na 15 seconden getoond.

(39)

CONDITION 4 – Has both direct and indirect attribution (YDYI)

T6 - Hieronder staat een artikel van nu.nl. Lees dit artikel alstublieft zorgvuldig door. Als u deze vragenlijst op uw telefoon invult, kunt u het beste uw telefoon horizontaal houden om het artikel goed te kunnen lezen. Nadat u het artikel hebt gelezen zal u een aantal vragen worden gesteld over de inhoud van het artikel. De knop om naar de volgende pagina te navigeren wordt na 15 seconden getoond.

(40)

BLOCK 4 – DEPENDENT VARIABLES

T7 - U heeft zojuist een artikel gelezen dat afkomstig is van nu.nl. De volgende vragen gaan over het bedrijf wat in het artikel wordt besproken, SoGood, en het nieuwe product wat zij hebben uitgebracht, SoGood Proud. In de vragen wordt de term LHBTQ+ gebruikt. Dit is een afkorting die staat voor lesbisch, homoseksueel, biseksueel, transgender en queer.

DV1 - Geef alstublieft uw gevoel tegenover het bedrijf SoGood aan, gebaseerd op het artikel wat u zojuist gelezen heeft.

Onaantrekkelijk ◯ ◯ ◯ ◯ ◯ ◯ ◯ Aantrekkelijk

Slecht ◯ ◯ ◯ ◯ ◯ ◯ ◯ Goed

Onaangenaam ◯ ◯ ◯ ◯ ◯ ◯ ◯ Aangenaam

Ongunstig ◯ ◯ ◯ ◯ ◯ ◯ ◯ Gunstig

Onsympathiek ◯ ◯ ◯ ◯ ◯ ◯ ◯ Sympathiek

DV2 - Geef alstublieft aan hoe waarschijnlijk het is dat u het product SoGood Proud zou willen kopen, gebaseerd op het artikel wat u zojuist gelezen heeft.

Nooit ◯ ◯ ◯ ◯ ◯ ◯ ◯ Vast en zeker

Zeker niet van plan te kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Zeker van plan te kopen Lage interesse om te kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Hoge interesse om te kopen Zal het zeker niet kopen ◯ ◯ ◯ ◯ ◯ ◯ ◯ Zal het zeker kopen

Zal het waarschijnlijk niet kopen

◯ ◯ ◯ ◯ ◯ ◯ ◯ Zal het waarschijnlijk kopen

(41)

DV3 - Geef alstublieft per stelling hieronder aan in welke mate u het met de stelling eens bent.

Sterk mee oneens Sterk mee eens De claim van SoGood dat zij streven naar

LHBTQ+-gelijkwaardigheid is waar.

◯ ◯ ◯ ◯ ◯ ◯ ◯

De claim van SoGood dat zij streven naar LHBTQ+-gelijkwaardigheid is misleidend.

◯ ◯ ◯ ◯ ◯ ◯ ◯

De claim van SoGood dat zij streven naar LHBTQ+-gelijkwaardigheid is overdreven.

◯ ◯ ◯ ◯ ◯ ◯ ◯

Ik geloof niet dat SoGood zo veel geeft om LHBTQ+-gelijkwaardigheid als dat het claimt.

◯ ◯ ◯ ◯ ◯ ◯ ◯

BLOCK 5 – MANIPULATION CHECK

MC1 - De volgende vragen gaan over het nu.nl-artikel zelf. In de vragen wordt het begrip "pinkwashing" gebruikt. Pinkwashing refereert naar het idee dat bedrijven zich willen profileren als homovriendelijk (bijvoorbeeld in hun marketingcampagnes) zonder dat ze daadwerkelijk betrokken zijn of hun beleid hierop aanpassen. Geef alstublieft per vraag aan in welke mate u denkt dat dit klopt, gebaseerd op wat u heeft gelezen.

(42)

Volledig niet Een beetje niet Neutraal Een beetje wel Volledig wel

In welke mate wordt in het artikel wat u zojuist gelezen heeft het bedrijf SoGood direct beschuldigd van de praktijk “pinkwashing”?

◯ ◯ ◯ ◯ ◯

In welke mate wordt in het artikel wat je zojuist gelezen hebt het bedrijf SoGood indirect in verband gebracht met de praktijk “pinkwashing”?

◯ ◯ ◯ ◯ ◯

MC2 - Wordt “pinkwashing” in het artikel direct toegeschreven aan SoGood? o Ja

o Nee

o Weet ik niet

MC3 - Wordt "pinkwashing" in het artikel indirect toegeschreven aan SoGood? o Ja

o Nee

(43)

MC4 - Geef aan in hoeverre u het eens bent met de volgende stellingen. Helemaal

oneens Oneens Neutraal Eens

Helemaal eens

Het artikel is realistisch. ◯ ◯ ◯ ◯ ◯

Het artikel is geloofwaardig. ◯ ◯ ◯ ◯ ◯

Het artikel is neutraal geschreven.

◯ ◯ ◯ ◯ ◯

MC5 - Heeft u het artikel wat u zojuist heeft gelezen al eens eerder gezien, bijvoorbeeld in een andere vragenlijst die u eerder heeft ingevuld?

o Ja o Nee

BLOCK 6 – DEMOGRAPHICS

T8 - Afsluitend willen we u een aantal demografische vragen stellen.

DEM1 - Wat is uw geslacht? o Man

o Vrouw

o Anders, namelijk _______________________________________________

DEM2 - Wat is uw leeftijd?

(44)

DEM3 - Wat is uw hoogst genoten opleiding? o WO doctoraal of master

o HBO of WO bachelor of kandidaats

o HAVO en VWO bovenbouw / WO en HBO propedeuse o MBO

o Eerste drie jaar HAVO en VWO / MAVO / VMBO (theoretisch en gemengde leerweg)

o LBO / VBO / VBO (kader- en beroepsgerichte leerweg) o Geen onderwijs / basisonderwijs

o Weet ik niet / zeg ik liever niet

DEM4 - Met welke geaardheid identificeert u uzelf het meest? o Heteroseksueel o Homoseksueel / lesbisch o Biseksueel o Queer o Aseksueel o Anders, namelijk: ________________________________________________ o Zeg ik liever niet

DEM5 - Identificeert u uzelf als transgender? o Ja

o Nee

(45)

DEM6 - Heeft u deze vragenlijst gevonden via SurveySwap? o Ja

o Nee

BLOCK 7 – DEBRIEFING

T9 - Bedankt voor uw deelname aan dit onderzoek. Voordat u deze vragenlijst afrondt willen we u graag extra informatie bieden over de opzet van het onderzoek. Het nieuwsartikel wat u zojuist gezien heeft van nu.nl is speciaal gemaakt voor dit onderzoek en is dus niet echt. Ook de frisdrankfabrikant SoGood die in het artikel genoemd wordt bestaat niet, eveneens als belangenorganisaties Equality Together en OneUnited. Het artikel is gefabriceerd om te onderzoeken of de meningen van consumenten over een bedrijf veranderen naar mate “pinkwashing” wordt toegekend aan een bedrijf. Pinkwashing refereert naar het idee dat bedrijven zich willen profileren als homovriendelijk (bijvoorbeeld in hun

marketingcampagnes) zonder dat ze daadwerkelijk betrokken zijn of hun beleid hierop aanpassen.

Mocht u naar aanleiding van dit onderzoek nog vragen hebben kunt u zich richten tot Sjoerd Tomassen, sjoerd.tomassen@student.uva.nl. Vergeet niet om hieronder verder te klikken, daarmee rondt u de vragenlijst af.

(46)

Appendix C

Extra tables with results from statistical testing Table 1

Distribution of respondents per condition

IVs No direct attribution Pinkwashing directly attributed Total No indirect attribution 55 52 107 Pinkwashing indirectly attributed 50 46 96 Total 105 98 203 Table 2

Multivariate effects for direct and indirect attribution on consumer perceptions

Predictor Pillai’s Trace F Df p ηp2

(Intercept) .987 4974.935 (3, 197) .000 .987

Has direct attribution .018 1.198 (3, 197) .312 .018 Has indirect attribution .048 3.280 (3, 197) .022 .048 Has direct attribution*has indirect

attribution

(47)

Table 3

Univariate between-subjects effects for direct and indirect attribution on consumer perceptions Predictor Dependent variable Sum of squares Df Mean square F p ηp2 Intercept Skepticism 3483.957 3 3483.957 2892.936 .000 .936 Brand attitude 4148.748 3 4148.748 2773.839 .000 .933 Purchase intention 1669.519 3 1669.519 792.079 .000 .799 Has direct attribution Skepticism 2.564 1 2.564 2.129 .146 .011 Brand attitude 3.867 1 3.867 2.586 .109 .013 Purchase intention 0.268 1 0.268 0.127 .722 .001 Has indirect attribution Skepticism 0.181 1 0.181 0.150 .699 .001 Brand attitude 4.876 1 4.876 3.260 .072 .016 Purchase intention 1.666 1 1.666 0.790 .375 .004 Has direct attribution*has indirect attribution Skepticism 0.709 1 0.709 0.588 .444 .003 Brand attitude 0.810 1 0.810 0.542 .463 .003 Purchase intention 5.092 1 5.092 2.416 .122 .012 Error Skepticism 239.655 199 1.204 Brand attitude 297.638 199 1.496 Purchase 419.446 199 2.108

Referenties

GERELATEERDE DOCUMENTEN

real-time train operations. In addition, we wanted to determine whether, and how, we can measure workload WRS at a rail control post and demonstrate how it can be utilized. A

Keywords: HEART score, Chest pain, Clinical prediction rule, Risk score implementation, Impact, Stepped wedge design, Cluster randomised trial1. *

This paper discusses four tools that fit in the engineers’ toolkit to approach these multidisciplinary problems: TRIZ, Systematic Inventive Thinking, Quality Function Deployment

Nuclear security is defined by the International Atomic Energy Agency (IAEA) as the measures put in place for the prevention, detection of and response to theft, sabotage,

The study established the ensuing variables as critical in auditing challenges in the department: the participants were always informed about the actual commencement of

It is secondly postulated that with the addition of drought as co-stress, partial stomatal closure will occur in both Zea mays and Brassica napus crop plants thus mitigating the

The terms Dano-Frisian and North Sea Viking need to be coined for those Viking-Age groups of people that were closely related to both Frisia and Denmark on the one

The goal of this paper is to analyse the occurrences of self-citations with re- spect to different characteristics of papers (publication year, number of authors,