Personalized environmental advertising – a
double-edged sword?
A study on the effectiveness of personalization levels and the role of privacy concern, message relevance, and environmental concern
Tabea Valeska Filipski Student ID-card number: 12828866
Master Thesis
Graduate School of Communication Master’s programme Communication Science
Supervisor: Prof. Dr. Peter Neijens Date of completion: 26th of June 2020
Abstract
Personalization is often praised as an effective persuasion strategy that can be applied to a wide variety of areas – a premise that makes personalization attractive for environmental communication. Empirical studies dealing with the effectiveness of personalization, however, find both validating and invalidating evidence. This study conceptualizes personalization on three levels, generic, individual- and group-level, and investigates whether they produce different persuasive effects for environmental messages by activating contradictory mechanisms, namely privacy concern and perceived message relevance. In addition, consumers’ environmental concern is examined to identify a possible moderator of
personalization effectiveness. In an online experiment, 194 Dutch and German participants were exposed to either a generic, individual-level, or group-level personalized advertisement. The results showed that those exposed to an individual-level personalized message believed the organization to be more credible and had higher intentions to support the environmental campaign, compared to a group-level personalized or a generic message. Also, it appeared that neither participants’ privacy concern nor their perceived message relevance mediated the effect of personalization on persuasive outcomes for the message, the organization, or the environment. The findings imply that environmental concern did not moderate the relation between personalization and the proposed mechanisms but directly increased the attitudes and credibility of the organization, the credibility of the message and intentions to support the campaign. The possible reasons for the small or non-significant results are discussed together with suggestions for future research.
Introduction
Sustainability is a highly relevant and trending topic in our society and offers new opportunities for environmental organizations and corporations to foster pro-environmental behavior (McDonagh & Prothero, 2014). A key challenge for sustainability advocates is then to encourage the adoption of more environmentally beneficial behavior through persuasive communication (Lauren, Smith, Louis, & Dean, 2017; Kollmuss & Agyeman, 2002). Due to the rise of media channels and the diversity in available content, environmental messages are competing for a limited resource: individuals’ attention and willingness to process a message (Bartsch & Kloß, 2019). Therefore, it is more difficult than ever to reach consumers and to get them involved, as this is not an automatic byproduct of mere message exposure (Petty & Cacioppo, 1986).
A promising strategy to approach this challenge is personalized advertising, which is particularly effective in attracting attention and involvement (Boerman, Kruikmeier, & Zuiderveen Borgesius, 2017; Bang & Wojdynski, 2016). Personalization refers to the technique of targeting advertisements to individuals based on personally identifying
information (Boerman et al., 2017; Bang, Wojdynski, & Lee, 2019; Maslowska, Smit, & van den Putte, 2016). Marketers incorporate a wide range of information about the consumer, such as their name, age, gender, location, and more recently data about search engine queries, watched videos, or read articles (Boerman et al., 2017). The latter is facilitated by the rapid development of database technology and behavior tracking, which provides marketers with a thorough insight into the interests, preferences, lifestyle, and needs of consumers, allowing them to tailor their advertisements to this information (Aguirre, Mahr, Grewal, de Ruyter, & Wetzels, 2015).
Naturally, the growth and increased usage of personalized advertising has been accompanied by research into its effectiveness (Maslowska et al., 2016). A number of studies
have found personalization to be effective, especially when consumers perceived the message content as relevant and useful (Van Noort, Smit, & Voorveld, 2013), and engaged in deeper message processing (Bartsch & Kloß, 2019; Maslowska et al., 2016). However, other scholars have found relevant downside effects when the technique was perceived as too personal, such as an increase of privacy concerns (Aguirre et al., 2015), feelings of intrusiveness (Van Doorn & Hoekstra, 2013), and reactance (Baek & Morimoto, 2012). This discrepancy emphasizes that, in certain situations, personalization can be a “double-edged sword” (Bartsch & Kloß, 2019, p. 357).
Recently, research has started recognizing the underlying mechanisms of
personalization as responsible for the inconsistency of these effects, namely privacy concern and message relevance (Bleier & Eisenbeiss, 2015; Zhu & Chang, 2016). Moreover,
Maslowska et al. (2016) have indicated that certain levels of personalization differ in their ability to trigger mechanisms and thus, yield different persuasive outcomes. This emphasizes the need to compare both the absence and presence of personalization, and the distinctive degrees of personalization on their mechanisms and persuasive effects. Additionally, studies have found individuals’ predispositions to be influential for the relative effect of different personalization levels (Bang et al., 2019; Stiglbauer & Kovacs, 2019; Baek & Morimoto, 2012). In the context of environmental advertising, the extent to which consumers are already concerned with an environmental issue may enhance the effects of personalization
(Wonneberger, 2018).
Although personalization has already been investigated in the context of social marketing (Bartsch & Kloß, 2019), its application in the environmental context is yet unexplored. Being the first of its kind, this study aims to expand the research on environmental marketing by focusing on the effects of personalized messages on the outcomes of an environmental campaign. Moreover, it aims at examining the contradictory
effects of personalization by testing a conceptual model that includes the underlying
mechanisms, and environmental concern as an additional factor. On the one hand, this study responds to the call for more in-depth research and fills the gap in the current literature (Bartsch & Kloß, 2019; Boerman et al., 2017). On the other hand, this study can have significant implications for environmental marketers looking for the right "fit" between their organization and personalization strategy. Given the above, this study will be guided by the following research question:
How does the level of personalization affect consumers’ responses to an environmental persuasive message through the activation of underlying mechanisms? And are the effects enhanced by consumers’ environmental concern?
Theory
To answer the research question, this thesis develops a conceptual model (Figure 1) that aims at exploring the mechanisms and effects of personalization with regard to the environmental message, the organization, and environmentalism. The relevant concepts and relationships are defined in the following.
Personalization and personalization levels
Personalization is not an entirely new concept in advertising (Baek & Morimoto, 2012). The constant development of technology is extending the areas of application and gives rise to strategies that can tailor persuasive messages specifically to individuals (Boerman et al., 2017). The broad scope of the concept includes personalized e-mail marketing, telemarketing, and online behavioral advertising, which is based on consumers’ demographics as well as their online and offline behavior (Boerman et al. 2017, Baek & Morimoto, 2012). Similarly, the definitions in studies dealing with personalization are just as diverse and use varying terms, such as customization (Boerman et al., 2017), tailoring
(Maslowska, 2013), or micro-targeting (Dobber, 2020). This study will follow Dijkstra and Ballasts’ (2012) definition of personalized communication, in which "personalization refers to incorporating elements in a text that refer to the recipient" (p. 61).
Scholars have indicated that the extent of the persuasive effects depends on the level of personalization, which is based on the type and the amount of information used in the message that identifies or characterizes its receiver (Boerman et al., 2017; Bang et al., 2019). A common conceptualization of personalization in multiple dimensions distinguishes between individual appeals and social identity appeals (White, Zahay, Thorbjørnsen, & Shavitt, 2008; Perez & Steinhart, 2014; Maslowska et al., 2016; Bang et al. 2019). Individual appeals target receivers as unique individuals, for instance, by personalizing a message according to issues that are congruent with the receivers’ interests, beliefs, or opinions (Dobber, 2020). Herewith, it is assumed that a message that provides information about a congruent issue increases the chances of message scrutiny and, ultimately, leads to higher degrees of persuasion (Dobber, 2020). Social identity appeals, in contrast, target receivers as members of a group by using information that is associated with their social identity (e.g., by including their workplace, gender, location, or ethnicity) (Bang et al. 2019). Therefore, current literature often considers individual appeals as highly personalized, and social identity appeals as moderately
personalized in comparison to generic appeals, that do not include personalization. Given the above, this research will conceptualize personalization on three levels: generic, group-level, and individual-level.
A variety of studies has examined the presence or absence of personalization, and found that personalized messages attract more attention from the receiver and promote positive attitudes and message relevance, as well as favorable attitudes and increased
credibility of the advertising company (Maslowska et al., 2016; Zhu & Chang, 2016; Aguirre et al., 2015). Also, in terms of behavioral effects, recent research has suggested that
personalization reinforces behavioral intentions and behavior in conformity with the persuasive message (Bartsch & Kloß, 2019; Dijkstra & Ballast, 2012). Moreover, issue-congruent messages have been found to lead to a higher voting likelihood (Dobber, 2020). In comparison to social identity appeals, research has found mixed patterns for the effectiveness of individual appeals. Perez and Steinhardt (2014) have found that social identity appeals outperform individual appeals that refer to the receivers’ name and generally led to more positive evaluations. In contrast, the results of two recent studies have ascribed superiority to individual appeals (Maslowska et al., 2016; Bang et al., 2019). Following the line of evidence that indicates that individual-level personalization provokes stronger positive persuasive outcomes compared to group-level or generic personalization, the first hypothesis is:
H1. Individual-level personalized messages have stronger positive effects for the message, the organization, and the environment than group-level personalized messages or generic messages.1
Mechanisms of Personalization
Increasingly, scholars emphasize the importance of examining the underlying
mechanisms that influence the relationship between personalized messages and outcomes of interest, such as the receivers’ attitudes, intentions, and behavior (Maslowska et al., 2016; Kalyamaran & Sundar, 2006; Bartsch & Kloß, 2019). Therefore, this study aims to investigate how personalization works and which mechanisms may explain contradictory effects. Two possible mediators of personalization are proposed in the following.
1The effects for a) the personalized message contain attitudes and credibility, for b) the advertising
Personalization and privacy concern
Personalized messages have the potential to raise privacy concern due to the activities related to collecting, analyzing, and leveraging the receivers’ private information (Bleier & Eisenbeiss, 2015). According to Westin (1967), privacy is “the ability of the individual to control the terms under which personal information is acquired and used” (p.7). Moreover, privacy concern most likely occurs when receivers find the information used in
advertisements to be too close or personal (Aguirre et al., 2015). The level of personalization serves then as an indicator of the extent to which the organization has collected personal information and determines whether the receiver experiences a violation of his privacy (Bleier & Eisenbeiss, 2015). Prior research has shown that privacy concern negatively influences the effects of personalization in terms of click-through intentions (Aguirre et al., 2015; Bleier & Eisenbeiss, 2015). Drawing from this, a high level of personalization should increase privacy concern and lead to more negative effects regarding the message, the organization, and the environment. The second hypothesis is:
H2. Individual-level personalized messages increase privacy concern and therefore elicit more negative effects regarding the personalized message, the advertising organization, and the environment compared to group-level personalized messages or generic messages.2
Personalization and perceived message relevance
In contrast to the above, perceived message relevance is recognized as one of the main drivers of the effectiveness of personalized messages (Pavlou & Stewart, 2000; Rimer & Kreuter, 2006; Zhu & Chang, 2016; Kim & Huh, 2017). According to Zhu and Chang (2016),
2 The effects for a) the personalized message contain attitudes and credibility, for b) the advertising
organization, they contain attitudes and credibility, and c) for the environment, they contain behavioral intentions.
message relevance refers to the degree to which receivers perceive a personalized message to be self-related or in some way instrumental in achieving personal goals. The influence of message relevance can be explained by rational choice theory, which suggests that the evaluation of a message depends on its perceived benefits or utility and costs (Zhu & Chang, 2016). Based on this, receivers are more likely to accept personalization when the message provides them with great benefit, such as content that is relevant and useful for them (Boerman et al., 2017). Message relevance can be enhanced by incorporating even minimal personal information of the receiver (Dijkstra, 2005). Yet, especially highly personalized messages were found to be perceived as relevant and to lead to higher levels of attention, elaboration, central message processing, and ultimately to persuasion (Zhu & Chang, 2016; Rimer & Kreuter, 2006). In line with previous evidence, the third hypothesis is:
H3. Individual-level personalized messages increase perceived message relevance and therefore elicit more positive effects regarding the personalized message, the
advertising organization, and the environment compared to group-level personalized messages or generic messages.
Personalization, environmentalism and environmental concern
Previous research has indicated that the receivers’ predispositions influence the effectiveness of personalization (Hirsh, Kang, & Bodenhausen, 2012; Stiglbauer & Kovacs, 2019; Bang et al., 2019). In the particular context of environmentalism, this research
examines the role of receivers’ environmental concern. Environmental concern refers to values or perceptions regarding “the balance between humans and nature, the impact of humans on the environment, and humans living in harmony with nature versus dominating it” (Wonneberger, 2018, p.171). In environmental marketing research, environmental concern has been found to strengthen the effects of green product advertising (Lee, Choi, Youn, Lee, Ha-Brookshire, & Wilson, 2012), to foster favorable campaign attitudes and to lead to higher
intentions to support environmental campaigns (Wonneberger, 2018). These effects can be explained by the Elaboration Likelihood Model (ELM) (Petty & Cacioppo, 1986), which predicts that involvement facilitates the motivation to engage with a persuasive message and leads to thorough processing (Petty & Cacioppo, 1986). Moreover, messages are considered to have higher chances of influencing receivers who are highly involved (McKeen,
Guimaraes, & Wetherbe, 1994), or are ‘concerned’ with the environment (Wonneberger, 2018). Given the implications of the ELM, environmental concern can be assumed to increase the motivation to engage with an environmental message and subsequently affect the
mechanisms of personalization: On the one hand, environmental concern could lead to a dispersion of privacy concern and on the other hand, emphasize the perceived message relevance. Accordingly, the next hypotheses describe the influence of environmental concern on the effects of personalization on privacy concern and perceived message relevance.
H4. The effect of personalization on privacy concern is moderated by environmental concern, such that for receivers who are highly concerned about the environment the effect of personalized messages on privacy concern is lower than for receivers who are less concerned about the environment.
H5. The effect of personalization on perceived message relevance is moderated by environmental concern, such that for receivers who are highly concerned about the environment the effect of personalized messages on perceived message relevance is higher than for receivers who are less concerned about the environment.
For a clear understanding of environmental concern, this study also explores the direct effects on the outcomes for the message, the organization, and the environment. In line with previous evidence that has found a positive impact of environmental concern on sustainable behavior (Lee et al., 2012), the last hypothesis is:
H6. Environmental concern increases receivers’ positive responses related to the personalized message, the advertising organization, and the environment.
Figure 1 Conceptual Model Method
To examine the research question and the hypotheses based on it, an online experiment was conducted with message personalization as the experimental factor. Personalization was operationalized using generic, group-level personalized, and individual-level personalized versions of an environmental advertisement. Participants were randomly allocated to one version and were asked about message relevance, privacy concern, attitudes, and perceptions towards the advertisement and the advertising organization, and their intentions to support the campaign. The following sections describe the experimental design, the stimuli, the pretest,
Personalization Generic Group-level Individual-level Persuasive outcomes Message relevance Environmental concern
the recruitment and composition of the sample, the procedure and the measures of the final experiment.
Design
The experiment applied a non-factorial, between-subject design with three
personalization conditions: generic, group-level, and individual-level personalization. The experimental variable personalization was considered as a nominal factor. In addition, the participants’ environmental concern and the dependent variables were measured as continuous variables.
Stimuli
Different versions of an advertisement were created based on existing examples of the non-profit organization World Wide Fund for Nature (WWF) that aims to promote public awareness for environmental issues, and solicits engagement and financial support for their initiatives (WWF, 2020). WWF was chosen as the advertising organization because it engages in environmental issues that suited the purpose of this study. All versions of the advertisement shared a similar appearance and the same layout. They included the WWF-logo as well as an activating claim (“Support WWF and sign up now for our new initiative!”). Moreover, the advertisements were embedded in a fictional Instagram-post to create a more realistic online context.
The three levels of personalization were manipulated by varying the type of personal information included in the advertisement (Perez & Steinhart, 2014; Bang et al., 2019). In the generic condition, the advertisement included a general statement about WWFs’ conservation work. In the group-level personalized condition, the advertisement showed the same general statement but included a reference to the participants’ nationality, which was realized by varying the language of the ad between German and Dutch. The individual-level
personalization condition integrated personal information that referred to the participants’ scores on environmental issue relevance, which was collected before the stimuli exposure (Dobber, 2020). Hereby, participants were asked to rate five environmental issues (freshwater systems, forests, endangered animals, renewable energy, and sustainable food) on their
perceived importance. If participants indicated “sustainable food” as their most relevant issue, for instance, they received a version of the advertisement including an informative statement about the specific issue (“Sustainable food systems conserve nature and feed humanity.”). All versions of the stimuli can be seen in Appendix A.
Pretest
To determine whether participants perceived the intended levels of personalization as such, a pretest was conducted with the tool Qualtrics. The short questionnaire included demographic items (age, gender, and nationality), a measure of environmental issue
relevance, the stimulus material, and finally a manipulation check of two items that measured perceived personalization based on the respondents’ nationality and issue importance.
Twenty-five participants, who were randomly allocated to one of the conditions, participated in the pretest. The results indicated that the group-level personalized
advertisement was perceived as more personalized on the basis of the respondents’ nationality (Mgroup= 4.88, SD = 2.10, n = 8), compared to the generic (Mgeneric= 2.00, SD = .943, n = 10)
and individual-level personalized advertisement (Mindividual= 1.71, SD = 1.11; n = 7). Further,
participants perceived that the individual-level personalized advertisement was strongly linked with their indicated most important issue (Mindividual= 4.86, SD = 1.77; n = 7), compared
to the generic (Mgeneric= 3.5, SD = 1.58, n = 10) or the group-level personalized advertisement
(Mgroup= 3.13, SD = 1.64, n = 8). Since the stimuli showed the desired trends, they were
Participants
The online experiment took place between the 7th and 17th of May 2020. Participants
were recruited through personal communication as well as through invitations on social media platforms, such as Facebook, LinkedIn, Instagram, and SurveyCircle. A total of 222
participants agreed to participate in the study, whereby 24 participants were removed from the dataset because they either did not complete the experiment (n = 20) or they took an
exceptionally short time of less than three minutes (n = 4). Thus, the analysis was conducted over a final dataset of 194 participants, which were evenly distributed among the three conditions (generic condition, n = 61; group-level personalized condition, n = 67; individual-level personalized condition, n = 65). The adjusted sample consisted of 120 female and 73 male participants and one participant who preferred not to indicate a gender. Their age varied between 18 and 65 years (Mage = 25.92, SD = 7.57, n = 194). Furthermore, 144 (72.2%)
participants indicated that they have a German nationality and 50 (25.8%) participants stated to have a Dutch nationality. Overall, the respondents were highly educated: 29 (14.9%) participants completed high school, three (1.5%) participants completed post-secondary education (MBO), four (2.1%) participants have an associate degree, 94 (48.5%) participants have a WO or HBO Bachelor’s degree, 56 (28.9%) a WO or HBO Master’s degree, three (1.5%) participants have a professional degree and five (2.6%) participants have a doctoral degree.
Procedure
The questionnaire was presented in English and was described as a study about
consumers’ perceptions of environmental campaigns. Participants signed an informed consent and were randomly assigned to one of the three conditions. The questionnaire started with questions about the participants’ age, gender, nationality, and the highest completed educational level. Then, the participants indicated the extent to which they are concerned
about the environment and rated the importance of five environmental issues. Once the participants finished the question about their daily actions for the environment, they were exposed to the advertisement belonging to their condition. Participants could look at the advertisement as long as they wanted before continuing with the questionnaire. Thereafter, they indicated their attitudes and perceptions, such as message relevance and credibility of the advertisement. These questions were followed by a manipulation check including questions about perceived personalization, participants’ attitudes and perceptions towards the
organization, and their intentions to support the campaign. The questionnaire ended with questions about the participants’ privacy concern and a debriefing which thanked the participants, informed that the viewed advertisements were especially designed for research purposes, and briefly explained the objective of the study.
Measures
Eight latent constructs were each measured with multiple items, which can be found in Appendix B. To ensure that these items formed single unidimensional scales, a principal axis analysis (PAF) with direct oblimin rotation was performed for each measure. The analyses showed that the items loaded on one component and positively correlated with the first factor. Furthermore, all scales showed a good reliability with Cronbachs’ alpha larger than .80. The construction and values for each scale are described in the following.
Environmental concern. Environmental concern was measured by asking participants to indicate on a 7-point Likert scale ranged from 1 (strongly disagree) to 7 (strongly agree) to what extent they agreed with statements, e.g., “I am concerned about the environment” (Schuhwerk & Lekoff-Hagius, 1995). The mean score of the four items was used as a
measure of environmental concern (Eigenvalue =2.75, explained variance = 68.67%, = .844, M = 5.65, SD = .95).
Privacy concern. To measure privacy concern, participants indicated the extent to which they agreed with four statements, e.g., “I am concerned that WWF has too much information about me” (Smith, Milberg, & Burke, 1996; Diney & Hart, 2004). The 7-point Likert scale ranged from 1 (strongly disagree) to 7 (strongly agree). The mean score was used as a measure of privacy concern (Eigenvalue= 3.18, explained variance = 79.64%, = .913, M = 5.65, SD = 1.46).
Perceived relevance of the message. Perceived relevance was measured with seven items that were adapted from Laczniak and Muehling (1993). Participants rated the items, e.g., “The advertisement was meaningful to me”, on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). The mean score was used as the measure for
perceived message relevance (Eigenvalue = 5.31, explained variance = 75.86%, = .946, M = 4.04, SD = 1.37).
Attitude towards the message. Attitude towards the message was measured on a semantic differential scale with three contradictory items, that had to be rated with a score from 1 to 7: “bad” (1) … “good” (7)”; “unfavorable” (1) … “favorable” (7); “negative” (1) … “positive” (7) (Lee & Aaker, 2004). The mean score of these items was used as the measure for attitudes (Eigenvalue = 2.17 explained variance = 72.38%, = .807, M = 4.84, SD = 1.09).
Credibility of the message. Credibility of the message was measured by asking participants to indicate on a scale ranging from 1 (strongly disagree) to 7 (strongly agree) to what extent they agreed with six statements, e.g., “The advertisement was credible” (Chang, 2011). The mean score of the items was used as the measure for credibility (Eigenvalue = 3.79, explained variance = 63.23%, = .866, M = 4.98, SD = 1.05).
Attitudes towards the organization. Attitudes towards the organization WWF were measured on a 7-point semantic differential scale with three items (Lee & Aaker, 2004), e.g.,
the organization WWF is “bad” (1) … “good” (7). The mean score of these items was used as a measure of attitudes towards the organization (Eigenvalue = 2.82, explained variance = 94.09%, = .968, M = 5.46, SD = 1.34).
Credibility of the organization. To measure credibility of the organization, the scale of McCroskey and Teven (1999) was used, which includes the three elements of source
credibility; trustworthiness, expertise and goodwill. Each of these elements was measured with one contradictory pair, that had to be rated with a score from 1 to 7. An example for trustworthiness is the contradiction between “honest” (1) … “dishonest” (7), for expertise the contradiction between “informed” (1) … “uninformed” (7), and for goodwill the contradiction between “not self-centred” (1) … “self-centred” (7). The mean score of these items was used as the measurement (Eigenvalue = 2.20, explained variance = 73.39%, = .821, M = 5.37, SD = 1.13).
Behavioral intentions. Behavioral intentions were measured by asking participants to what extent they would be likely to engage in behaviors concerning the environmental
campaign. Two items were modified from Venable, Rose, Bush, and Gilbert (2005): “I would make a donation to WWF” and “I would contribute personal time for WWF’s initiatives”. Five additional items were developed that asked for intentions, such as “I would comment below the advertisement on Instagram”. The items had to be rated on a scale from 1 (extremely unlikely) to 7 (extremely likely). The mean value was used as a final measure for intentions (Eigenvalue = 4.08, explained variance = 58.37%, = .876, M = 2.79, SD = 1.29).
Issue relevance. In addition to the scales, one item on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree) asked the participants whether they are concerned about environmental issues, namely freshwater, renewable energy, sustainable food, forests, and endangered animals. An index was created to examine the average number of issues the participants were highly concerned about (issues with a score of 5 = ‘somewhat
agree’ or higher). The participants indicated that they were concerned about 4.25 issues (SD = 1.25). Moreover, they were most often concerned about forests (28.86%), followed by
sustainable food (21.13%), endangered animals (19.07%), renewable energy (17.52%), and freshwater (13.40%).
Results
As with the preparation of the data, the statistical software program IBM SPSS Statistics 26 was used to calculate the variables and the subsequent tests. The significance level was set at or below .05 before conducting the main analysis for the manipulation check and the test of the hypotheses, which are described in the following.
Manipulation check
Two items measured whether the manipulation of personalization was successful. The first item asked to indicate the extent to which the participants noticed whether the
advertisement was targeted at their nationality on a scale from 1 (not at all) to 5 (a great deal). The one-way analysis of variance with this item as dependent and ‘personalization’ as
independent variable was significant (F (2,190) = 59.84, p<.001, 2 = .38). Post-hoc comparisons with Bonferroni correction indicated that participants in the group-level personalized condition perceived the personalization significantly different compared to the generic condition (Mdifference= 1.80, p<.001) and the individual-level condition (Mdifference=
1.76, p<.001).
The second item asked participants to rate their perceptions of the question “Did you notice whether or not the advertisement was targeted at your most important environmental concern?”, on a scale ranging from 1 (not at all) to 5 (a great deal). The analysis of variance was significant (F (2,190) = 37.91, p<.001, 2 = .28). The post-hoc comparisons showed that participants in the individual-level condition indicated a higher rating compared to the generic
condition (Mdifference= 1.84, p<.001) and the group-level condition (Mdifference= 1.40, p<.001).
Thus, the manipulations of personalization were successfully recognized by the participants.
Randomization
To ensure that the sampling was randomized and participants were evenly allocated over the conditions, one-way analyses of variance were conducted for the independent variable ‘personalization’ and the participants’ background variables. The results showed no significant differences between the conditions regarding age (F (2,189) = .082, p = .921), gender (F (2,190) = 2.43, p = .091), or nationality (F (2,190) = .09, p = .911). For the variable education, however, a significant difference was found (F (2,190) = 3.15, p = .045, 2 = .03).
Post-hoc comparisons with Bonferroni correction showed that only the group-level condition differed significantly from the generic condition (Mdifference= .59, p = .048). As this test was
almost not significant and education did not correlate with any of the variables in the model (Table 2 in Appendix C), it was decided to continue without further corrections for the sample composition. The mean scores and standard deviations of the analyses can be found in Table 1 in the Appendix C.
Main analysis
According to Hypothesis 1, the expectation is that individual-level personalized messages have a stronger positive effect on persuasive outcomes for the message, the
organization, and the environment than group-level personalized or generic messages. To test this hypothesis, one-way analyses of variance were performed with ‘personalization’ as independent variable and persuasive outcomes as dependent variables.
First, as can be seen in Table 3, the analyses showed that participants assigned to the individual-level personalization condition scored highest on message credibility, but the effect was not significant (F (2,190) = 2.90, p = .057). The effect of individual-level personalization
on message attitudes was also not significant (F (2,190) = .92, p = .399). Second, concerning the outcomes for the organization, individual-level personalization led to credibility that was significantly more positive (F (2,190) = 3.58, p = .030). The eta-squared value of 2 = .03, however, indicated a very weak effect. The post-hoc comparisons with Bonferroni correction found a significant difference between individual-level and group-level personalization (Mdifference= .52, p = .025). The effect of individual-level personalization on organization
attitudes was not significant (Welchs’ F (2,216) = 1.17, p = .3143). Third, individual-level personalization led to significantly higher and positive behavioral intentions (F (2,190) = 3.72, p = .026), but the effect was very small (2=.03). In addition, the post-hoc comparisons found no significant difference between individual-level and generic personalization
(Mdifference = .53, p = .060) or group-level personalization (Mdifference = .52, p = .056). Although
the effects were very small, Hypothesis 1 is partly supported for the credibility of the organization and behavioral intentions.
Table 3
One-way analysis of variance for the effects of personalization on the persuasive outcomes
Message Organization Behavioral
intentions Credibility Attitude Credibility Attitude
Personalization
generic 4.98(.99) 4.95 (1.10) 5.39 (1.09)ab 5.49(1.13) 2.60 (1.12)
group-level 4.76(1.18) 4.69 (1.12) 5.09 (1.26)a 5.25 (1.54) 2.61 (1.37)
individual-level 5.20(.94) 4.88 (1.07) 5.62 (.98)b 5.63 (1.28) 3.14 (1.35)
Note. N = 194. Mean scores with standard deviations between parentheses. a b Means with a different superscript in the same column differ significantly at p < .05.
Mediation analyses
According to Hypothesis 2 and 3, the expectation is that individual-level personalization elicits greater persuasive outcomes because it influences underlying
mechanisms, such as privacy concern and perceived relevance, differently from group-level and generic personalization. To test the simple mediation effects, Model 4 of the PROCESS version 3.5 with the extension of multicategorical independent variables was run in SPSS (Hayes, 2018). The dummy variable created in PROCESS coded individual-level
personalization as 1, and group-level and generic personalization as 0. All analyses used 5.000 bootstrap samples to estimate ‘bias-corrected bootstrap confidence intervals’. A detailed report of the results of Hypotheses 2 and 3 can be found in Appendix D.
To test Hypothesis 2, the model was run with ‘personalization’ as independent variable, ‘privacy concern’ as mediator and each persuasive outcome as dependent variable. Table 4 represents the unstandardized beta-coefficients of the relative indirect effects and direct effects of the analyses. The results in this overview show that individual-level personalization increased privacy concern, but not significantly. Privacy concern did not significantly influence the persuasion outcomes and there were no significant relative indirect effects of individual-level personalization that led to more negative persuasion outcomes. As a consequence, Hypothesis 2 is rejected.
Table 4
Regression coefficients (unstandardized) with individual-level personalization and privacy concern as independent variables and persuasion outcomes as dependent variables
Note. * p<.05. ** p<.01. *** p<.001. ILP= individual-level personalization, PC= privacy concern.
To test Hypothesis 3, the same procedure was followed with ‘message relevance’ as mediator. Table 5 gives an overview of the unstandardized beta-coefficients of the relative indirect effects and direct effects of individual-level personalization through message
relevance in comparison to group-level and generic personalization. The results demonstrate that the effect of individual-level personalization on perceived message relevance was positive, but not significant. Therefore, there were no significant relative indirect effects of individual-level personalization on the persuasion outcomes. However, message relevance directly increased the persuasive outcomes, and led to significantly more favorable attitudes and higher credibility of the message and the organization, as well as to higher behavioral intentions. As a consequence of the non-significant indirect effects of individual-level personalization, Hypothesis 3 is not supported.
Privacy concern
Message Organization Behavioral
intentions Credibility Attitude Credibility Attitude
Relative indirect effect of ILP through PC -.00 .00 -.02 -.02 -.00 Direct effect of ILP .11 .21 -.07 .24 .16 .54* Direct effect of PC . -.02 -.00 -.21 -.18 -.04 R2 .03 .01 .11 .05 .03 F 2.00 .612 8.02*** 3.53* 2.60
Table 5
Regression coefficients (unstandardized) with individual-level personalization and perceived message relevance as independent variables and persuasion outcomes as dependent variables
Note. * p<.05. ** p<.01. *** p<.001. ILP= individual-level personalization, MR= message relevance.
Moderation analyses
To test the interrelationship between personalization and the degree of environmental concern influencing the underlying mechanisms proposed in Hypothesis 4 and 5, two-factor analyses of variance were conducted. As a preparatory step, a median split was carried out with the continuous variable ‘environmental concern’ to distinguish between a high and a low degree. The split showed a median value of 5.75, so that participants who scored 5.75 or higher on the variable were considered as highly concerned (n = 107) and participants who scored below 5.75 were regarded as low concerned (n = 87). Based on this, environmental concern was transformed into a categorical variable ‘environmental concern categorical’ (low concern = 0; high concern = 1) to proceed with the further analysis.
Message relevance
Message Organization Behavioral
intentions Credibility Attitude Credibility Attitude
Relative indirect effect of ILP through MR .12 .12 .10 .14 .15 Direct effect of ILP .27 .09 -.19 .12 -.00 .37* Direct effect of MR .45*** .46*** .37*** .53*** .58*** R2 .35 .33 .24 .29 .40 F 35.41*** 30.82*** 19.61*** 26.66** 42.77***
Following Hypothesis 4 the expectation is that the effect of personalized messages on privacy concern is moderated by environmental concern, so that participants who are highly concerned about the environment have less privacy concerns than participants who are not concerned about the environment. Thus, an analysis of variance was performed with ‘personalization’ and ‘environmental concern categorical’ as independent variables, and ‘privacy concern’ as dependent variable. The group means and standard deviations are presented in Table 6. As can be seen in Table 7, the analysis found no significant interaction effect between personalization and environmental concern (F (2,193) = .24, p = .785), indicating that environmental concern did not influence participants’ privacy concern when they were exposed to personalization. Consequently, Hypotheses 4 is rejected.
Table 6
Group means and standard deviations of the two-factor analyses of variance for privacy concern Environmental concern low high Personalization generic 3.80 (1.35) 3.33 (1.77) group-level 3.84 (1.41) 3.61 (1.13) individual-level 3.75 (1.71) 3.65 (1.50)
Table 7
Two-factor analysis of variance for privacy concern
SS df MS F p Eta-squared Pers .927 3 .31 .14 . 935 .00 EC 3.35 1 3.35 1.53 .217 .00 Pers* EC 1.06 2 .53 .24 .785 .00 Error 407.86 187 Total 3016.81 194
Note. N = 194. SS = Sum of squares; df = degrees of freedom; MS = Mean squares; Pers = personalization; EC = environmental concern categorical.
The same procedure was performed to test Hypothesis 5, which entails that
environmental concern strengthens the effect of personalization on the perceived relevance of the message. As can be seen in Table 9, the analysis of variance found no significant
interaction effect between personalization and environmental concern on perceived message relevance (F (2,193) = .29, p = .745). Therefore, participants did not perceive the personalized messages as more relevant when they were highly concerned about the environment,
compared to participants who were not concerned about the environment. Thus, Hypotheses 5 is rejected. The group means and standard deviations are presented in Table 8.
Table 8
Group means and standard deviations of the two-factor analyses of variance for perceived message relevance Environmental concern low high Personalization generic 3.82 (1.12) 4.33 (1.34) group-level 3.57 (1.49) 3.81 (1.61) individual-level 4.00 (1.12) 4.59 (1.22)
Table 9
Two-factor analysis of variance for perceived message relevance
SS df MS F p Eta-squared Pers 11.63 2 .5.82 3.21 .042 .00 EC 9.49 1 9.49 5.25 .023 .00 Pers* EC 1.06 2 .53 .29 .745 .00 Error 338.58 187 Total 3505.40 193
Note. N = 194. SS = Sum of squares; df = degrees of freedom; MS = Mean squares; Pers = personalization; EC= environmental concern categorical.
To test the proposed direct effects of environmental concern and the persuasion outcomes in Hypothesis 6, bivariate correlations were calculated between the measure of ‘environmental concern’ and each dependent variable. Table 10 shows that environmental concern correlated significantly with the outcomes for the organization, such as credibility and attitude, and behavioral intentions (p <.001). For the message outcomes, environmental concern correlated significantly with credibility (p <.05), however, not with attitudes towards the message. These results partly support Hypothesis 6 and show that participants’
environmental concern positively influenced their responses towards the organization and their behavioral intentions but did not impact on their attitudes towards the message. Table 10
Correlations between environmental concern and the persuasion outcomes
Message Organization Behavioral
intentions Credibility Attitude Credibility Attitude
Environmental
concern .156* .102 .186*** .237*** .235***
Conclusion and discussion
Following the scientific literature, this study developed a conceptual model which was then tested in an online experiment. The model aimed at gaining insights into the
effectiveness of personalization in the context of environmental advertising by examining three different personalization levels (generic, group and individual) as well as their influence on underlying mechanisms, privacy concern and perceived message relevance. Further, the role of participants’ environmental concern was investigated to identify a possible moderator of the effectiveness of personalization.
First, the model hypothesized that individual-level personalization leads to more positive outcomes for the message, the organization, and the environment than group-level or generic personalization. The findings partly confirm this hypothesis and demonstrate that individual-level personalization significantly increased the credibility of the organization and behavioral intentions. Next, it was proposed that the levels of personalization influence privacy concern and perceived message relevance differently, which subsequently produce contradictory effects on the persuasion outcomes. Contrary to the expectation, neither privacy concern, nor perceived message relevance were significantly influenced by the level of personalization and therefore, they did not mediate the personalization effects. Privacy concern and message relevance yet directly influenced the outcomes in the expected directions, so that participants with high privacy concern evaluated the outcomes for the message, the organization, and the environment more negative, while participants who perceived the message as relevant showed significantly more positive evaluations.
Further, the model hypothesized that participants’ environmental concern moderates the relationship between personalization and its mechanisms, privacy concern and perceived relevance. However, the results find that participants’ environmental concern did not increase participants perceptions of message relevance nor did it reduce privacy concern. Moreover,
the expectation was that environmental concern directly influences the persuasive outcomes. In conformity with the hypothesis, environmental concern increased the credibility and attitudes of the organization, behavioral intentions, and the credibility of the message. This indicates that consumers who are concerned about the environment are particularly receptive to the environmental message regardless of personalization.
The results are partly in line with previous research on the effectiveness of generic, group-level, and individual-level personalization. On the one hand, they support scholars that found superiority of individual-level personalization (Maslowska et al., 2016; Bang et al., 2019), although the effects were small. On the other hand, they could not confirm that privacy concern and perceived message relevance are mechanisms of these effects, which is
contradictory to previous findings (Bleier & Eisenbeiss, 2015; Zhu & Chang, 2016). A reason for this inconsistency can be ascribed to the conceptualization of
personalization in this study. The personalized levels were manipulated based on participants’ issue relevance or their national identity. Maslowska et al. (2016) for instance, realized the individual-level by making a direct reference to the participant’s name. While ones’ name has been found to activate ones’ self-scheme (Dijkstra, 2008), the reference to relevant
environmental issues in this study was less obvious and affected by additional limitations. Many participants rated more than one out of five environmental issues as “very relevant” (the average number of issues with the score ‘very relevant’ was 4.25). This indicates, that they could have perceived the selected issue as a relevant issue, but only as one of the many relevant issues they saw. Therefore, the manipulation was not distinctive enough to create a strong link to their indicated most relevant issue.
Similarly, the manipulation of group-level personalization differed from previous scholars and, in retrospect, was not very suited to create a strong link to the participants’ social identity. Bang et al. (2019) conceptualized group-level personalization by addressing
the participants as students, which led to higher attention and favorable attitudes. This study focused on social identity in terms of nationality, which was manipulated by adjusting the message according to ones’ native language. The remaining questionnaire, however, was presented in English and may have initiated an international orientation, so that the sudden change to Dutch or German was either not strongly perceived or led to negative feelings of confusion.
Given the above, the manipulations were not recognized as ‘too’ personal to raise privacy concern or to increase the relevance of the message, which were identified as mediators in previous scholars (Bleier & Eisenbeiss, 2015; Zhu & Chang, 2016). Future research could enhance personalization by using phrases that directly stress ones’ nationality (e.g., “we as Germans …”) or by including additional information (e.g., the participants’ name) that is more strongly related to the self (Dijkstra, 2008).
A factor that could have played a confounding role as well could be the choice of a familiar organization as source of the environmental message. Brand familiarity and trust have been found to affect participants’ privacy concern and possibly to moderate the effects of personalization (Wottrick, Verlegh, & Smit, 2017). Thus, the participants could have had an opinion about WWF and have processed the message on the basis of their pre-existing experiences. To overcome this possible influence, a follow-up study should consider a less familiar organization.
The examination of environmental concern followed the call of previous studies to investigate the effects of personalization in different subgroups of the population (Stiglbauer & Kovacs, 2019). While environmental concern was not confirmed as a moderator, the findings hint to alternative factors. Smit, Van Noort, and Voorveld (2014) found that
however, participants were highly educated in comparison to the overall Dutch and German population (OECD, 2019), which could explain the weak effects.
Also, the participants’ cultural orientation could have had an influence. The study focused only on Western European participants from Germany and the Netherlands. Drawing from the individualism typology suggested by Hofstede (1983), Yu and Cude (2009)
compared personalization effects in Western and Eastern countries and found that American consumers had generally more negative perceptions of personalization than Korean
consumers. This indicates, that the effects found in the current study may not apply to consumers with a different educational or cultural background. Thus, future research should consider a sample of participants with diverse education, cultural characteristics, or
nationalities.
Besides the described limitations of the operationalization, the choice of method came with significant disadvantages, such that the participants were exposed to the stimulus in an artificial and one-time manner. This could have affected participants’ perceptions and attitude strength (Germelmann & Groeppel-Klein, 2009). To provide insights about long-term effects that may arise from the positive relationship based on individual-personalization, the
experiment should be repeated in a more natural setting which retargets the participants over a period of time.
The current study can also be extended from a conceptual point of view, as it examines the individual-level and group-level personalization separately. However, it seems worthwhile to investigate a condition that combines both levels as Maslowska et al. (2016) previously indicated stronger persuasion effects for combined appeals. This conceptualization would also create a realistic environment where marketers already vary the use of personal information (Bleier & Eisenbeiss, 2015).
In addition to the theoretical implications, this study provides useful insights for a practical application of personalization. Altogether, the findings emphasize the difficulty to target individuals’ characteristics or social identities and that environmental advertisements should be personalized carefully and based on intensive consumer research. Moreover, they imply that the presence of personalization on different levels is meaningful: When applied on an individual-level, personalization has the potential to increase credibility and behavioral intentions. Yet, group-level personalization can elicit more negative reactions on the environmental message and organization compared to individual-level and generic
personalization. To conclude, the findings suggest that personalization based on personally relevant issues seems to be a good fit for environmental organizations that want to appear more credible or encourage consumers’ intentions to support their campaign. Further research is needed to identify the factors that clearly address the consumers’ social identity as well as personal predispositions that may enhance the personalization effectiveness.
References
Aguirre, E., Mahr, D., Grewal, D., de Ruyter, K., & Wetzels, M. (2015). Unraveling the personalization paradox: The effect of information collection and trust-building strategies on online advertisement effectiveness. Journal of Retailing, 91(1), 34-49. Baek, T., & Morimoto, M. (2012). Stay Away From Me. Journal of Advertising, 41(1), 59–
76.
Bang, H., Choi, D., Wojdynski, B., & Lee, Y. (2019). How the level of personalization affects the effectiveness of personalized ad messages: the moderating role of
narcissism. International Journal of Advertising, 38(8), 1116–1138.
Bang, H., & Wojdynski, B. (2016). Tracking users’ visual attention and responses to personalized advertising based on task cognitive demand. Computers in Human Behavior, 55, 867–876.
Bartsch, A., & Kloß, A. (2019). Personalized charity advertising. Can personalized prosocial messages promote empathy, attitude change, and helping intentions toward
stigmatized social groups? International Journal of Advertising, 38(3), 345–363. Boerman, S., Kruikemeier, S., & Zuiderveen Borgesius, F. (2017). Online Behavioral
Advertising: A Literature Review and Research Agenda. Journal of Advertising, 46(3), 363–376.
Bleier, A., & Eisenbeiss, M. (2015). The Importance of Trust for Personalized Online Advertising. Journal of Retailing, 91(3), 390–409.
Chang, C. (2011). Feeling Ambivalent About Going Green. Journal of Advertising, 40 (4), 19-32.
Dijkstra, A. (2005). Working mechanisms of computer-tailored health education: evidence from smoking cessation. Health Education Research, 20(5), 527–539.
Dijkstra, A. (2008). The Psychology of Tailoring‐Ingredients in Computer‐Tailored Persuasion. Social and Personality Psychology Compass, 2(2), 765–784. Dijkstra, A., & Ballast, K. (2012). Personalization and perceived personal relevance in
computer-tailored persuasion in smoking cessation. British Journal of Health Psychology, 17(1), 60–73.
Dinev, T., & Hart, P. (2004). Internet privacy concerns and their antecedents - measurement validity and a regression model. Behaviour & Information Technology, 23(6), 413– 422.
Dobber, T. (2020). Measuring effects of issue-based microtargeted messages on vote likelihood: Reporting a field experiment in a multi-party setting. [Unpublished doctoral dissertation]. Amsterdam School of Communication Research (ASoCR), Amsterdam, Netherlands.
Germelmann, C. C., & Gröppel-Klein, A. (2009). Forciert Forced-exposure Fehler bei der Datenerhebung?. Die Betriebswirtschaft, 69(2), 229.
Hayes, A. (2018). Partial, conditional, and moderated mediation: Quantification, inference, and interpretation. Communication Monographs: Special Issue: Advances in Methods and Statistics, 85(1), 4–40.
Hirsh, J., Kang, S., & Bodenhausen, G. (2012). Personalized Persuasion: Tailoring Persuasive Appeals to Recipients’ Personality Traits. Psychological Science, 23(6), 578–581. Hofstede, G. (1983). The Cultural Relativity of Organizational Practices and Theories.
Journal of International Business Studies, 14(2), 75–89.
Kalyanaraman, S., & Sundar, S. (2006). The Psychological Appeal of Personalized Content in Web Portals: Does Customization Affect Attitudes and Behavior? Journal of
Communication, 56(1), 110–132.
Kim, H., & Huh, J. (2017). Perceived Relevance and Privacy Concern Regarding Online Behavioral Advertising (OBA) and Their Role in Consumer Responses. Journal of Current Issues & Research in Advertising, 38(1), 92–105.
Kollmuss, A., & Agyeman, J. (2002). Mind the Gap: Why do people act environmentally and what are the barriers to pro-environmental behavior? Environmental Education Research, 8(3), 239–260.
Laczniak, R., & Muehling, D. (1993). The Relationship between Experimental Manipulations and Tests of Theory in an Advertising Message Involvement Context. Journal of Advertising, 22(3), 59–74.
Lauren, N., Smith, L., Louis, W., & Dean, A. (2019). Promoting Spillover: How Past Behaviors Increase Environmental Intentions by Cueing
Lee, A., & Aaker, J. (2004). Bringing the Frame Into Focus: The Influence of Regulatory Fit on Processing Fluency and Persuasion. Journal of Personality and Social
Psychology,86(2), 205–218.
Lee, N., Choi, Y., Youn, C., Lee, Y., Ha-Brookshire, J., & Wilson, L. (2012). Does Green Fashion Retailing Make Consumers More Eco-friendly? The Influence of Green Fashion Products and Campaigns on Green Consciousness and Behavior. Clothing and Textiles Research Journal, 30(1), 67–82.
Maslowska, E. (2013). “Just for you!”: a study into the effectiveness and the mechanism of customized communication. [s.n.].
Maslowska, E., Smit, E., & van Den Putte, B. (2016). It Is All in the Name: A Study of Consumers’ Responses to Personalized Communication. Journal of Interactive Advertising, 16(1), 74–85.
McCroskey, J., & Teven, J. (1999). Goodwill: A reexamination of the construct and its measurement. Communication Monographs, 66(1), 90–103.
McDonagh, P., & Prothero, A. (2014). Sustainability marketing research: past, present and future. Journal of Marketing Management: JMM 30th Anniversary Special Issue: Pushing the Boundaries, Sketching the Future, 30(11-12), 1186–1219.
McKeen, J., Guimaraes, T., & Wetherbe, J. (1994). The Relationship between User
Participation and User Satisfaction: An Investigation of Four Contingency Factors. MIS Quarterly, 18(4), 427–451.
OECD. (2019). Education at a Glance, 2019: OECD Indicators. Paris: OECD Publishing. Pavlou, P., & Stewart, D. (2000). Measuring the Effects and Effectiveness of Interactive
Advertising: A Research Agenda. Journal of Interactive Advertising, 1(1), 61–77. Perez, D., & Steinhart, Y. (2014). Not so personal: The benefits of social identity ad appeals
Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York, NY: Springer-Verlag.
Rimer, B., & Kreuter, M. (2006). Advancing Tailored Health Communication: A Persuasion and Message Effects Perspective. Journal of Communication, 56((Supp.)), S184– S201.
Schuhwerk, M. E., & Lefkoff-Hagius, R. (1995). Green or Non-Green? Does Type of Appeal Matter When Advertising a Green Product? Journal of Advertising, 24 (Summer), 45-54.
Smith, H., Milberg, S., & Burke, S. (1996). Information Privacy: Measuring Individuals’ Concerns about Organizational Practices. MIS Quarterly, 20(2), 167–196. Stiglbauer, B., & Kovacs, C. (2019). Need for Uniqueness Determines Reactions to
Web-Based Personalized Advertising. Psychological Reports, 122(1), 246–267. Van Doorn, J., & Hoekstra, J. (2013). Customization of online advertising: The role of
intrusiveness. Marketing Letters, 24(4), 339–351.
Van Noort, G., Smit, E. G., & Voorveld, H. A. (2013). The online behavioural advertising icon: two user studies. In Advances in Advertising Research (Vol. IV) (pp. 365-378). Springer Gabler, Wiesbaden.
Venable, B., Rose, G., Bush, V., & Gilbert, F. (2005). The role of brand personality in
charitable giving: An assessment and validation. Journal of the Academy of Marketing Science, 33(3), 295–312.
Westin, A. (1967). Privacy and freedom. Atheneum, New York
White, T., Zahay, D., Thorbjørnsen, H., & Shavitt, S. (2008). Getting too personal: Reactance to highly personalized email solicitations. Marketing Letters, 19(1), 39–50.
Wonneberger, A. (2018). Environmentalism – a question of guilt? Testing a model of guilt arousal and effects for environmental campaigns. Journal of Nonprofit & Public Sector Marketing, 30(2), 168-186.
Wottrich, V., Verlegh, P., & Smit, E. (2017). The role of customization, brand trust, and privacy concerns in advergaming. International Journal of Advertising: International Conference on Research in Advertising (ICORIA) 2015, 36(1), 60–81.
WWF. (2020, June 20). Our work. Retrieved from: https://www.worldwildlife.org/initiatives. Yu, J., & Cude, B. (2009). Possible Disparities in Consumers’ Perceptions Toward
Personalized Advertising Caused by Cultural Differences: U.S. and Korea. Journal of International Consumer Marketing, 21(4), 251–269.
Zhu, Y. Q., & Chang, J. H. (2016). The key role of relevance in personalized advertisement: Examining its impact on perceptions of privacy invasion, self-awareness, and
Appendix A. Stimulus material
Individual-level personalized advertisements (specific issues: endangered species, freshwater, forests, climate, sustainable food).
Appendix B. Questionnaire experiment Introduction
Dear participant,
I would like to invite you to participate in a study, that is part of my Master thesis, to be conducted under the auspices of the Graduate School of Communication, a part of the University of Amsterdam.
In this research, I would like to ask your opinion about environmental issues. The study will take around 5 to 7 minutes and can be conducted on a laptop, tablet or mobile phone.
As this research is being carried out under the responsibility of the ASCoR, I can guarantee that:
1. Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions unless you first give your express permission for this.
2. You can refuse to participate in the research or cut short your participation without having to give a reason for doing so.
3. You also have up to 24 hours after participating to withdraw your permission to allow your answers or data to be used in the research. Participating in the study will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any explicitly offensive material.
4. No later than five months after the conclusion of the research, I will be able to provide you with a research report that explains the general results of the research.
For more information about the research and the invitation to participate, you are welcome to contact me at any time:
Tabea Filipski
tabea.filipski96@gmail.com +4915757815404
Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this study, you can contact the designated member of the Ethics Committee representing ASCoR, at the
following address: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl. Any complaints or comments will be treated in the strictest confidence.
I hope that I have provided you with sufficient information. I would like to take this opportunity to thank you in advance for your assistance with this research, which I greatly appreciate.
Kind regards, Tabea Filipski
Informed consent
I hereby declare that I have been informed in a clear manner about the nature and method of the research, as described in the introduction of this study.
I agree, fully and voluntarily, to participate in this research study. With this, I retain the right to withdraw my consent, without having to give a reason for doing so. I am aware that I may halt my participation in the experiment at any time.
If my research results are used in scientific publications or are made public in another way, this will be done in such a way that my anonymity is completely safeguarded. My personal data will not be passed on to third parties without my express permission. If I wish to receive more information about the research, either now or in the future, I can contact Tabea Filipski. Should I have any complaints about this research, I can contact the designated member of the Ethics Committee representing the ASCoR, at the following address: ASCoR secretariat, Ethics Committee, University of
Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐ 525 3680; ascor‐secr‐ fmg@uva.nl.
o I understand the text presented above, and I agree to participate in the research.
Demographic questions
What is your age in years? o (open answer) What is your gender?
o Female o Male
o I’d prefer not to say What is your nationality?
o German o Dutch
o Other (led to debriefing and end of survey)
What is your highest level of education you are currently enrolled in or have completed?
o Elementary school o High school
o Post-secondary education / MBO o Associate degree
o Bachelor’s degree (HBO/WO) o Master’s degree (HBO/WO) o Doctoral degree
Environmental concern
To what extent do you agree or disagree with the following statements? − "I am concerned about the environment."
o Strongly disagree o Disagree
o Somewhat disagree o Neither agree nor disagree o Somewhat agree
o Agree
o Strongly agree
− "I am concerned the condition of the environment affects the quality of my life."
o (Et cetera.)
− "I am willing to make sacrifices to protect the environment." o (Et cetera.)
− "My actions have an impact on the environment." o (Et cetera.)
Issue importance
How concerned are you about the following environmental issues? − Freshwater
o Strongly disagree o Disagree
o Neither agree nor disagree o Somewhat agree o Agree o Strongly agree − Renewable energy o (Et cetera.) − Sustainable food o (Et cetera.) − Forests o (Et cetera.) − Endangered animals o (Et cetera.)
Most relevant issue
Which of the following environmental issues are you most concerned about? Please choose one of the topics.
− Freshwater − Renewable energy − Sustainable food − Forests
− Endangered animals
What do you do yourself in your daily life for the environment? Please give one or two examples. If you don't do anything special, you can skip this question and continue with the questionnaire.
Introduction 2 and stimulus exposure
On the next page, you will see an environmental advertisement for the non-profit organization World Wide Fund for Nature (WWF).
Please look at it carefully and take as much time as you need.
Introduction 3 questionnaire
In the following, I would like to ask you several questions about the organization WWF and the environmental advertisement you just saw.
Attitudes towards the message
How did you feel about the advertisement? " I think the advertisement was…"
− Bad O O O O O O O Good − Unfavorable O O O O O O O Favorable − Negative O O O O O O O Positive
Credibility of the message
To what extent do you agree or disagree with the following statements? "I think the advertisement was..."
− Believable
o Strongly disagree o Disagree
