The effect of message framing on young adult consumers ’ sustainable fashion consumption: The role of anticipated emotions and perceived ethicality

The debate about which communication strategies are most effective for inducing consumers to behave sus-tainably remains open, despite growing attention on more sustainable forms of fashion consumption. To further this understanding, we investigate the effectiveness of positive and negative message framing in promoting sustainable fashion consumption, where the beneficial versus detrimental environmental consequences of choosing second-hand clothing were highlighted. Across two experiments, positively framed messages were more effective than negatively framed ones in prompting consumers to engage in sustainable fashion consumption. Elevation was the anticipated emotion responsible for this effect, while consumers ’ subjective beliefs about the ethicality of advertising messages promoting sustainable consumption served as a moderator. We discuss the theoretical and managerial implications of this research together with its limitations and directions for future research.


Introduction
Over the last years, sustainability has become a vital business goal for many stakeholders (Kim et al., 2020) and has been recognized as a major global concern (Wang et al., 2019).In the fashion industry, achieving sustainability success has become a crucial issue for companies, as this industry is considered to be among those that have the greatest negative impact on the environment (Pal and Gander, 2018;Shrivastava et al., 2021).While research on fashion sustainability is new and emerging, current evidence suggests that consumers are more inclined to implement ethical forms of consumption (Carrington et al., 2021;Kong et al., 2021;Pangarkar et al., 2021).Young consumers are particularly interested in these topics (McNeill and Venter, 2019;Park and Lin, 2020), and market research has identified an ongoing trend in which sustainability and more conscious shopping have become important priorities in the purchasing decisions of the younger generation (McKinsey, 2019;Roberts, 2019).The new affluent generations are influenced by ethical principles (Stern, 2011) and display higher expectations for fashion brands to be sustainable in their value chains (Deloitte, 2019).Companies are trying to develop more circular business models, and many fashion companies and online fashion platforms are opening sections dedicated to second-hand clothing in their online channels (e.g., Levi's, Patagonia, and Zalando).Furthermore, an increasing number of websites and apps are dedicated to clothing rental (e.g., By Rotation and Sisterly) and second-hand clothing purchases (e.g., Vinted, Depop, and Vestiaire Collective).
Based on these premises, this research examines a specific type of sustainable fashion consumption practice, the purchase of second-hand clothing via specialized apps, and focuses on young adult consumers, as most initiatives in support of second-hand fashion occur online (Lim et al., 2021) and primarily reach young audiences (Styvén and Mariani, 2020).Four consumer responses are examined: (a) intention to buy second-hand clothes via specialized apps; (b) intention to spread positive word of mouth (WOM) about this behavior; (c) actual consumer information-seeking; and (d) consumer intentions to volunteer time for related causes, a prosocial "secondary" outcome closely related to the topic.
Our research aims to examine the impact of online persuasive messages on these four responses.To accomplish this, we adopt the framework of framing theory (Goffman, 1974), which is useful for understanding the effects of differently framed messages on consumer behavior (Bullock and Shulman, 2021).The goal is to investigate how to encourage responsible consumption behaviors through persuasive and effectively framed messages.Prosocial research does not always achieve clear results, with some works highlighting that negatively framed messages are highly effective in fostering responsible behavior (Amatulli et al., 2019;Brennan and Binney, 2010), while others claim that positively framed messages produce more positive reactions (Randle et al., 2016;Yang et al., 2018).We propose that persuasive messages that are positively framed and draw attention to the opportunity to buy secondhand clothing will lead to more positive responses compared to persuasive messages that are negatively framed.In doing so, we participate in the recent debate regarding the role of anticipated emotions in the underlying psychological process that explains the effect of message framing on sustainable consumer responses (Amatulli et al., 2019;Baek and Yoon, 2017;Duhachek et al., 2012;Jin and Atkinson, 2021;Nabi et al., 2018;Randle et al., 2016;Zubair et al., 2020).Using two experimental studies, we demonstrate that message framing significantly affects consumer responses through the mediation of two anticipated emotions, elevation and guilt, and the moderation of the subjective beliefs of consumer-perceived ethicality (CPE; Brunk. 2010) regarding advertising promoting sustainable consumption.
This study makes several contributions.First, our findings contribute to the research on message framing and how to influence consumers' decisions to adopt sustainable behaviors (Amatulli et al., 2019;Brennan and Binney, 2010;Defazio et al., 2021;Randle et al., 2016;Yang et al., 2018).Here, we demonstrate that using a positively framed message stimulates specific emotions that, in turn, lead to more positive, sustainable fashion consumption responses.Thus, we contribute to the existing research regarding the role of positive emotions in fostering sustainable behaviors (Yan et al., 2023;Jin and Atkinson, 2021;Nabi et al., 2018;Randle et al., 2016).In doing so, we contribute to the emerging trend of positive sustainability initiatives (Peter and Honea, 2012;Wang et al., 2017), which show promise in increasing sustainable behaviors (White et al., 2019;Winterich et al., 2019).Our research demonstrates that sustainable responses occur due to the inspiration consumers experience through elevation.This is in line with research connecting inspiration to the elicitation of positive emotions (Yan et al., 2023;Thrash et al., 2014).In documenting message framing's ability to elicit inspiration, we contribute to the limited research on inspiration in marketing and consumer behavior (Böttger et al., 2017;Liang et al., 2016) and simultaneously provide further research on the role of positive emotions in sustainable behavior (White et al., 2019).
Secondly, we consider the role of an important dimension that has been largely unexplored in explaining consumer responses to advertising messages: consumers' subjective beliefs about the ethicality of advertising promoting specific behaviors, known as CPE (Burk, 2010).Thus, we contribute to the debate about the types of messages that are considered ethical for promoting responsible behaviors (Antonetti and Maklan, 2014;Hastings et al., 2004).We show that subjective consumer beliefs about the ethicality of advertising promoting sustainable behaviors shape the emotions that consumers anticipate in their adoption of these behaviors.Thus, our research not only acknowledges the need to communicate ethically to consumers but also demonstrates the underlying mechanism by which consumers' subjective ethical beliefs regulate anticipated emotions and the consequent adoption of sustainable behavior.
Thirdly, four different sustainable consumer responses are considered: three "primary" impacts of messages directly related to the promoted sustainable practice (i.e., intention to buy second-hand clothes via specialized apps, positive WOM, and actual consumer informationseeking) and one "secondary" impact connected to the topic (i.e., intention to volunteer for related causes).Both self-reported intentions and relevant actual behavior are considered (i.e., actual consumer information-seeking is measured by a dichotomous variable detecting whether respondents provided an email address for contact to obtain more information).In doing so, a broad spectrum of consumer responses is considered, providing a robust test of the proposed model and also disclosing the broad social return of the message.
In the following sections, the theoretical background, proposed hypotheses, and experimental studies are described.Finally, implications and directions for further research are discussed.

The context: The fashion industry
The global apparel industry is worth $3 trillion, accounting for 2 % of the world's gross domestic product and employing 33 million workers (Wang et al., 2019).Despite its economic value, the fashion industry showed also to be responsible for 10 % of carbon and GHG emissions (Boykoff et al., 2021) and its negative environmental impacts have recently gained attention (Changing Markets Foundation, 2021;Han et al., 2017;Mukendi et al., 2020;Pal and Gander, 2018;Styvén and Mariani, 2020).Studies have highlighted the need for fundamental changes in the industry (Mukendi et al., 2020;Park and Lin, 2020), such as moving from a linear to a circular system (Styvén and Mariani, 2020).Collaborative fashion consumption practices that recirculate goods (e.g., gifting, sharing, swapping, second-hand use) are examples of sustainable consumption practices in a circular system (Camacho-Otero et al., 2019).Of these, we focus on the purchase of second-hand clothing via specialized apps, with the aim of identifying the impact of different message framing on consumers' sustainable fashion consumption behavior.

Message framing and sustainable consumption behavior
The theory of message framing refers to the way people react or feel about a certain message depending on the way it is presented (Kahneman, 2011).Several studies have shown that individuals respond differently to the same information depending on whether it is framed in positive (gains) or negative (losses) terms (Tversky and Kahneman, 1981).Positively framed advertising (i.e., encouraging sustainable consumption) emphasizes environmental benefits in a prosocial marketing context, such as the positive consequences of the promoted action.Negatively framed advertising focuses on the negative aspects (Maheswaran and Meyers-Levy, 1990).
Message framing impacts consumer responses.Although framing theory states that consumers generally tend to overweight the probability of a loss over a gain (Tversky and Kahneman, 1981), prosocial research does not always achieve clear results.Inconsistent results can occur due to the types of framing used, as valence frames are not equal and different types of valence framing (e.g., risky choice, attribute, goal framing) can lead to conflicting results (Levin et al., 1998).However, additional psychological reflections must be considered, such as how people view themselves and how prosocial messages leverage this perception, to understand the framing effect.One line of research argues that negatively framed messages are highly effective in fostering responsible behavior as they lead to negative emotions, which encourages people to engage in coping behaviors aimed at regaining a positive view of themselves (Amatulli et al., 2019;Brennan and Binney, 2010).Other scholars claim that positively framed messages produce lower perceived manipulative intent and, thus, more positive reactions (Randle et al., 2016).Negative framing may produce boomerang effects (Hyland and Birrell, 1979), whereby it could result in people reacting in the opposite way to intended (Cox and Cox, 2001;Reinhart et al., 2007).This might occur as message framing can be inconsistent with preexisting perceptions related to the specific behavior (Cox and Cox, 2001) or because an individual might feel pressured (e.g., through elicited guilt) and limited in their freedom of choice (Reinhart et al., 2007).In these situations, consumer backlash and resistance to reform might occur in response to a perceived attempt at moral persuasion or a communication intervention too intrusive (Lee et al., 2020).
In the interpretative model adopted here, positively framed messages are assumed to be more acceptable to consumers and, therefore, more effective at inspiring consumers to act responsibly (Yang et al., 2018).Supported by previous studies on the effectiveness of positive framing (Defazio et al., 2021), we hypothesize the following: H1: Positive message frames are more likely to lead to consumers engaging in positive sustainable fashion consumption responses than negative message frames.

The moderating role of CPE
Scholars have highlighted ethical issues concerning the moral rightness of messages promoting sustainable consumption based on consumer responsibility (Antonetti and Maklan, 2014), creating debate regarding the ethicality of promoting responsible behaviors through messages that might imply negative self-perception (Hastings et al., 2004).Scholars increasingly question whether relying on negative message frames is morally justifiable, suggesting that marketers should first try to achieve their goals through positive message frames that are much less likely to generate unexpected negative consequences (Hastings et al., 2004).Consequently, companies might have a smaller range of communication options to effectively stimulate good consumer practices.We argue that in the domain of communication for sustainability, it is relevant to consider how consumers build their moral system and react to communication stimuli based on their personal beliefs and moral judgments (Reynolds, 2008;Reynolds et al., 2014).Consistent with social cognitive theory, behavioral outcomes are a function of individuals, stimuli, and the interaction between both (Bandura, 1986).Consumers vary in terms of their attention devoted to themes of morality and how they process stimuli and information (Reynolds et al. 2014).
Morality is a distinct category in an individual cognitive framework used to recognize and encode stimuli (Wurthmann, 2013).Morality depends on different beliefs based on social norms and personal experience (Antonetti and Maklan, 2014) and shapes consumers' reactions to stimuli.How subjective beliefs regarding the ethicality of corporate behaviors form has received attention in consumer behavior literature (Brunk, 2012;Street et al., 2001).According to Reynolds (2008) and the literature regarding ethical decision-making (Street et al., 2001), there is explanatory value in considering how consumers make decisions based on individual perceptions of ethicality.Consumers' subjective beliefs and ethical perceptions affect attitude formation and, therefore, behaviors (Van Quaquebeke et al., 2019).CPE typically focuses on whether a practice is deemed as morally "right" or "wrong" and is primarily based on an individual's stable inner values.Consumers' unwavering responses are based on their moral values, which are shaped by social backgrounds and life experiences (Ferrell and Gresham, 1985;Fraedrich and Ferrell, 1992).Based on these premises, we consider the construct of CPE (Brunk, 2010(Brunk, , 2012)), here regarded as a consumer's perception of morality when examining advertising messages promoting sustainable behaviors.This subjective construct might vary according to individual factors, including personality, cognitive prerequisites, demographic characteristics, moral identity (Aquino and Reed, 2002), and personal moral sensitivity (Sparks and Hunt, 1998).Brunk (2010) identified several different domains that can be involved in CPE.In the consumer domain, these include the advertising messages that companies direct to their targets.In this research, CPE relates to the communication strategies used by companies to promote sustainable behaviors, which can range from stimulating higher-order values to leveraging consumer responsibility and negative self-perception.Consumer evaluations of the ethicality of communication strategies may differ and potentially affect their responses (Hyman and Tansey, 1990).CPE acts as a lens for interpreting communication stimuli (Brunk, 2012).CPE is a chronically accessible cognitive interpretative lens used in the elaboration of stimuli to inform an individual's ethical decision process.
People with high CPE assign companies the right to use the widest spectrum of communication options to stimulate higher-level goals, such as environmental protection.People with low CPE do not believe that advertising messages should leverage consumer responsibility (Antonetti and Maklan, 2014) or negative self-perception (Hastings et al., 2004) and are not prone to consider such stimuli.Therefore, through this ethical belief interpretative lens, communication strategies can impact consumers' behaviors differently.
Previous research demonstrated that ethical attributes influence consumer behaviors, such as WOM (Markovic et al., 2018), purchase intention (Van Quaquebeke et al., 2019), boycott (Farah and Newman, 2010), and retaliation (Shea, 2010).Furthermore, subjective CPE affects consumer purchase intentions (Brunk, 2010;Shah et al., 2020) and positive WOM (Markovic et al., 2018).We contend that communication messages stimulate consumers' responses differently, depending on their subjective ethical beliefs.A positive frame is accepted either by people with high or low CPE, as it does not involve any morally or ethically problematic or ambiguous issues.On the contrary, when consumers are exposed to a negatively framed message, CPE operates as a significant interpretative lens.Those experiencing high CPE are expected to be affected by the negative message frame, as it is perceived to be in line with their ethical norms.Those experiencing low CPE are expected to evade the same message, as it is not in line with their ethical norms.
H2: The effect of a negative message frame on sustainable fashion consumption responses depends on CPE; it is stronger when the CPE is high.Instead, the effect of a positive message frame on sustainable fashion consumption responses does not depend on CPE.

The mediating role of anticipated emotions
Research considering the role of emotions in explaining the effect of message framing on sustainable consumption is quite recent (see Table 1 for an overview).Some studies primarily focused on negative emotions (Amatulli et al., 2019;Baek and Yoon, 2017;Duhachek et al., 2012), documenting that negative emotional appeals can be persuasive in leading recipients to adopt responsible behavior.Other studies (Jin and Atkinson, 2021;Nabi et al., 2018;Randle et al., 2016) have demonstrated the role positive emotions play in fostering prosocial consumption behaviors.This is in line with previous works suggesting that, when faced with an ethical decision, consumers consider the subsequent effects of their actions and choose those they anticipate to have the most pleasurable emotions (Mellers and Mcgraw, 2001).Anticipated moral emotions affect intentions in consumers' ethical (Steenhaut and Van Kenhove, 2006), pro-environmental (Onwezen et al., 2013), and goaldirected (Perugini and Bagozzi, 2001) behaviors.
We contend that the effect of message framing on sustainable consumer responses is explained by the role of anticipated emotions.A losses (negative) frame leverages an individual's sense of responsibility and fosters emotions connected with the negative consequences of not engaging in pro-environmental practices.A gains (positive) frame underlines the positive effects of adopting pro-environmental practices and leverages self-enhancing goals and related emotions (Trudel et al., 2020).Positive and negative emotions are different (Haidt, 2003).Negative emotions act as "red-alert buttons," focusing attention on a problem and setting a corrective procedure in motion; positive emotions typically arise in safer situations in which focused action is not required, and their purpose is to broaden a person's "momentary thought-action repertoire" (Haidt 2003, p. 862).We acknowledge the need to focus on both negative and positive emotions associated with specific actions that consumers can take and subsequent behavioral coping mechanisms.Therefore, we hypothesize a role of two specific emotions that previous research has found to be affected by message frames (Amatulli et al., 2019;Randle et al., 2016).Specifically, the negative emotion of guilt and the positive emotion of elevation.Guilt and elevation are both moral S. Grappi et al.

Table 1
Overview of previous studies focused on message framing, sustainable consumption, and the role of emotions.

Amatulli et al., 2019
Examine the effectiveness of negative vs. positive message framing in promoting green products, considering the role of anticipated negative emotions.

Shame (mediator).
Negatively framed messages are more effective in prompting consumers to engage in pro-environmental behaviors.Anticipated shame is the emotion responsible for this effect.Baek and Yoon, 2017 Examine how negative emotions influence responses to environmental ad messages framed as gains or losses.
Guilt (or shame) paired with a gainframed (or loss-framed) message is more persuasive in fostering an eco-friendly attitude and behavioral intention.Duhachek et al., 2012 Examine how negative emotions differentially influence the effectiveness of health messages framed as gains or losses.
Guilt and shame (independent variables).
Guilt appeals are more effective when paired with gain frames, whereas shame appeals are more effective when paired with loss frames.Jin and Atkinson, 2021 Examine how individuals' positive emotions influence the persuasive effectiveness of climate change news framing techniques (thematic vs. episodic).
Positive emotional engagement plays a central role in processing episodically framed messages, attenuating the effects of risk perception and news believability.

Nabi et al., 2018
Examine the role of emotions in the gain/ loss framing of environmental policy initiatives.

Hope (mediator). Fear (mediator).
Loss-framed messages led to fear, while gain-framed messages led to hope.Both emotions are mediators of framed messages and desired behaviors, with hope playing a key role.

Randle et al., 2016
Examine whether message framing (positive vs. negative) leads to emotions affecting positive responses to advertisements for a specific case of highcognitive-elaboration donation.
Positive framing elicits significantly more positive emotions, whereas negative framing elicits more negative emotions.
Positive emotions cause stronger reactions to the advertisements.Zubair et al., 2020 Examine the neural mechanism of message framing on self-conscious emotions regarding consumer purchase intention in the context of green marketing.

Guilt (independent variable).
Positively framed messages attract the attention of consumers at first glance, while negatively framed messages are more effective when customers go deeper to understand the message slogan.emotions, which are "emotions that are linked to the interests or welfare either of society as a whole or at least of persons other than the judge or agent" (Haidt, 2003, p. 854).Both have been shown to support and motivate prosocial behavior (Haidt, 2003).
Guilt is a central moral emotion aroused by the violation of moral rules and imperatives, particularly violations causing harm or suffering to others.The cognitive antecedent of guilt is a negative judgment about an action (e.g., "I did something wrong") (Amatulli et al., 2019).Guilt is felt when individuals believe they have personally caused a negative outcome, as opposed to when they perceive a negative event is caused by others, in which case they are likely to feel anger (Antonetti and Maklan, 2014;Lindenmeier et al., 2012).Consumers rationalize guilt and try to regulate their behavior through problem-focused coping mechanisms.This is in contrast to what is associated with different negative emotions, such as shame, which leads to emotion-focused coping, in which consumers try to regulate the emotional experience (Antonetti and Maklan, 2014).Generally, guilt is considered a good moral emotion because it motivates one to help the victim or make up for transgressions.Research shows that guilt motivates direct helping behavior, supports prosocial action, and promotes ethical consumer behavior (Antonetti and Maklan, 2014;Haidt, 2003).
Elevation is a positive emotion elicited by moral beauty.Acts of charity, kindness, loyalty, and self-sacrifice are powerful triggers for elevation that can provoke one's desire to become a better person (Haidt, 2003).Based on recent research on inspiration in marketing (Böttger et al., 2017;Winterich et al., 2019), the process of being inspired by an external stimulus, such as a message, can activate the emotion of elevation (Rudd et al., 2018;Thrash et al., 2014).According to Haidt (2003), elevation initiates action tendencies.Thus, when consumers are inspired by marketing stimuli and consequently feel elevated, they are inspired to act (Böttger et al., 2017;Shiota et al., 2014).Elevation motivates individuals to perform prosocial and affiliated actions (Landis et al., 2009;Romani and Grappi, 2014), acting as a transmission element between the first phase of inspiration and the connected actions.Prior research demonstrated that, compared with other positive emotions (e.g., happiness and pride), moral elevation is the most effective in improving pro-environmental behavior (Li et al., 2021;Zelenski and Desrochers, 2021) as it has an empowerment effect that can avert the danger of moral licensing (Schnall and Roper, 2012).Recently, van Van Kleef and Lelieveld (2022) reviewed the literature on how the experience of emotions influences people's prosocial behavior, confirming that the "appreciation and self-transcendence" group of emotions (among which elevation is classed) better explains the intrapersonal prosocial behaviors (i.e., people's own prosocial behaviors), compared to the "dominance and status assertion" group of emotions (i.e., anger, disgust, contempt, envy, and pride).
We contend that elevation and guilt, anticipated emotions felt by consumers thinking about taking (or not taking) actions related to a message, drive subsequent behavior.In line with previous research demonstrating that message framing can elicit emotions (Hesz and Neophytou, 2010;Tong et al., 2021) and that emotions can mediate the effect of framing on sustainable consumer behavior (Amatulli et al., 2019;Randle et al., 2016), we propose that the anticipated emotions of elevation and guilt mediate the influence of message framing on sustainable fashion consumption.
The use of emotions to stimulate consumer responses is connected to CPE, as it represents individual beliefs about the ethics of the way behaviors are promoted (Burk, 2012;Hastings et al., 2004).Consumers showing high CPE are prone to accept either positive or negative framed messages, as their belief system accepts both in favor of higher-level goals.Those showing low CPE are more selective in considering external stimuli and morally disputing negative frames.We contend that the moderation mechanism exerted by CPE on consumers' sustainable responses develops through the anticipated emotions of guilt and elevation and occurs when respondents are exposed to negative, not positive, frames.
H3: Message framing influences sustainable fashion consumption responses through the mediation of the anticipated emotion of elevation, whose strength depends on CPE; H3a: CPE does not moderate the effect of the positive message frame on elevation, whereas it moderates the impact of the negative message frame.The effect of a negative message frame on elevation is stronger when the CPE is high; H3b: The greater the anticipated emotion of elevation, the stronger the sustainable fashion consumption response.H4: Message framing influences sustainable fashion consumption responses through the mediation of the anticipated emotion of guilt, whose strength depends on CPE: H4a: CPE does not moderate the effect of the positive message frame on guilt, whereas it moderates the impact of the negative message frame.The effect of a negative message frame on guilt is stronger when the CPE is high; H4b: The greater the anticipated emotion of guilt, the stronger the sustainable fashion consumption response.

Overview of empirical research
We ran two complementary empirical studies to test H1 to H4. Fig. 1 and Table 2 provide an overview of the studies and of the hypotheses tested.To enhance the generalizability of our findings, we analyzed four different sustainable consumer responses.In Study 1, we used (a) intentions to buy second-hand clothes via specialized apps and (b) intentions to spread positive WOM about this behavior.In Study 2, we added (c) a measure of actual consumer information-seeking about the topic and (d) a relevant spillover effect, namely consumer intention to volunteer for related causes.Considering this "secondary" outcome opens new opportunities for promoting sustainable behaviors, demonstrating that the message can also induce "secondary" consumer responses, such as the willingness to volunteer for related causes, evidencing its broader social return.Ultimately, by considering four different dependent variables, two self-reported intentions directly related to the message (intention to buy and positive WOM), one selfreported intention indirectly related to the message (volunteering), and one actual behavior (actual consumer information-seeking), the study's external validity and results' robustness are enhanced.
Each study presents the same scenario (an advertising message for a specialized app for buying second-hand clothing) framed as a positive, neutral, or negative message (see Appendix A).We developed messages with realistic layouts and content based on actual posts and ads on similar topics to maximize the reliability and external validity of the study.We developed stimuli by matching realistic images with text messages inspired by previous studies (Amatulli et al., 2019).This is in line with recent research contending that images are useful to overcome the "problem of abstractness" (White et al., 2019, p. 35) related to sustainable consumption and that visual information helps to communicate how environmental issues can affect others, thus favoring the elicitation of concrete emotions and responses.All messages refer to a fictitious organization to avoid potential confounding effects associated with existing companies.The three scenarios have been pretested, as detailed in Appendix B.
Data were collected via online questionnaire surveys.We selected young adult respondents for both studies, whom we identify as subjects aged between 18 and 40 years, consistent with the aim of the study and with previous research (Han et al., 2017;Park and Lin, 2020).

Study 1
Study 1 tested H1 and H2.We compared the effects of the negative, neutral, and positive frames to reliably test the effects of the message framing and perceived message ethicality on consumer responses.We ran two distinct moderation analyses, one for each dependent variable S. Grappi et al. (intention to buy second-hand clothes via specialized apps and positive WOM).

Participants and procedures
Each participant read only one of the three fictitious advertising messages (see Appendix A) and then answered the questionnaire.We indicated that the company name was masked to comply with university ethics regulations.The sample was recruited via Prolific to complete an online survey hosted on Qualtrics.The survey was administered to a sample of 135 UK-resident young-adult consumers.We retained only questionnaires from those participants who answered the attention check correctly.Ten respondents were eliminated due to failing this check.The final sample of 125 respondents was sufficient for a power analysis calculated by G*power (power = 99 %, medium effect size, 5 % alpha margin error).The sample's characteristics are detailed in Table 2.

Measures
All variables were measured on 7-point Likert scales.CPE was measured through three items adapted from Brunk (2012) and Septianto et al. (2021) (i.e., "I think advertising messages promoting sustainable consumption are ethical"; α = 0.86; no difference emerged across conditions; F(2, 122) = 2.99; p >.05).After the manipulation, respondents rated their intention to buy second-hand clothes via specialized apps using three items adapted from Möhlmann (2015; i.e., "The next time I need a clothing item, I would prefer second-hand clothes via specialized apps"; α = 0.93).They rated their positive WOM intentions using a three item scale adapted from Romani et al. (2013;i.e., "I intend to say positive things about buying second-hand clothing via specialized apps to other people"; α = 0.95).We used additional measures as controls, including environmental concern (six items adapted from Polonsky et al., 2014; i.e. "I am concerned about the condition of the environment"; α = 0.93); disgust associated with wearing second-hand clothes (three items adapted from Grappi et al., 2013; i.e., "I would feel disgusted in wearing second-hand clothes"; α = 0.94); perceived costliness (five items adapted from Yu et al., 2021; i.e., "Buying second-hand clothing via the app would require my effort"; α = 0.92); frequency of clothing purchase and frequency of second-hand clothing purchase (both measured with one item).Finally, we collected several manipulation checks and demographic characteristics regarding the respondents.
The items of the focal variables were subjected to maximum likelihood exploratory factor analysis (EFA) with oblique rotation (promax).The EFA showed that the items loaded correctly onto the corresponding factors, with loadings greater than 0.50 on the focal factors and below 0.25 on other factors.We then performed a confirmatory factor analysis (CFA) on these measures to assess convergent and discriminant validity.The fit of the model was excellent (χ 2 (24) = 19.30;p =.74; CFI = 1.00;NNFI = 1.00;RMSEA = 0.00; SRMR = 0.02; Hu and Bentler, 1999), and the factors had adequately high discriminant validity, which was assessed using the χ 2 difference test (all χ 2 difference tests [Δχ 2 (df = 1)] were higher than 3.98, p <.05).The items were averaged to form the corresponding dimensions.Manipulation checks (detailed in Appendix C) confirmed that participants correctly perceived the stimuli as differently framed and realistic.No differences emerged regarding the perceived relevance or vividness of the messages.

Results
To test H1, the data were analyzed using a one-way multivariate analysis of variance (MANOVA), where consumer responses were expressed as a function of message frame valence.There was a statistically significant difference in consumer responses based on framing (F (4, 242) = 3.28, p <.05; Wilk's Λ = 0.90, partial η 2 = 0.05).Results confirm that the conditions demonstrated significant mean differences in intention to buy (F (2, 122) = 4.31, p <.05) and positive WOM (F (2, 122) = 4.63, p <.05) (see Table 3).Pairwise comparisons were conducted to evaluate the differences between groups.Results indicate that respondents were more inclined to buy second-hand clothes via specialized apps and to spread positive WOM about this behavior when exposed to a positive message frame.Thus, H1 is supported.
Given the difference between the experimental groups in terms of consumer responses, it is appropriate to test H2.We performed a moderation analysis for each outcome (PROCESS Model 1; Hayes, 2022), where the message framing effect on the dependent variable was moderated by individual CPE.The manipulated variable was coded to enable multicategorical moderation analyses.The negative frame was used as the comparison group for the analyses, and specific comparisons were run between negative and neutral frames (identified as X1 in Table 4) and between negative and positive frames (identified as X2 in Table 4).
Message framing and CPE interact significantly to influence consumer responses, represented here in the form of intention to buy (X1 [negative vs.   4).The conditional effects of the predictor at values of the moderator, along with the corresponding bootstrap confidence intervals (CIs), showed exactly when the hypothesized effects occur.When consumers were exposed to a negative frame, the effect of the message on their intention to buy and to spread positive WOM is stronger when the CPE is high.The negative message frame stimulated these consumer responses only when they showed high levels of CPE.On the contrary, when consumers were exposed to a positive frame, the effect of the message on intention to buy and WOM did not depend on consumers' CPE (Fig. 2).Hence, H2 is supported.

Study 2
Study 2 aimed to test H3 and H4.We tested the underlying process linking consumer exposure to a specific (negative, neutral, or positive) message frame, consumer CPE, consumer anticipated emotions (elevation and guilt), and four relevant consumer responses (intention to buy second-hand clothes via specialized apps, positive WOM, actual consumer information-seeking, and volunteering intention).The manipulated variable (i.e., message framing) was coded as in Study 1 to enable multicategorical comparisons.

Participants and procedures
We used the same three stimuli as in Study 1 (see Appendix A).The sample was recruited via Prolific to complete an online survey hosted on Qualtrics.The survey was administered to a sample of 159 UK-resident young-adult consumers.We retained questionnaires where participants correctly answered the attention checks.Fourteen respondents were eliminated due to failing this check.The final sample comprised 145 respondents.The sample size was sufficient based on a power analysis calculated by G*power (power = 99 %, medium effect size, 5 % alpha margin error).The sample's characteristics are detailed in Table 2.

Measures
All variables were measured on 7-point Likert scales.CPE (α = 0.94; no difference emerged across conditions: F(2,141) = 2.18; p >.05), intention to buy second-hand clothes via specialized apps (α = 0.93), and positive WOM (α = 0.95), were measured as in Study 1.Additionally, respondents were asked to answer a four-item scale adapted from Aquino et al. (2011) to assess their sense of anticipated elevation ("I feel touched," "I feel inspired," "I feel moved," and "I feel in awe"; α = 0.84) and a three-item scale adapted from Amatulli et al. (2019) to assess their sense of anticipated guilt ("I feel guilty," "I feel culpable," and "I feel remorseful"; α = 0.83).To measure anticipated feelings of elevation and guilt accurately, we prompted respondents to merely think about taking (or not taking) an action related to the message they read (i.e., if they decided to download the app and buy or not buy second-hand clothing, they would contribute to the protection of or the collapse of the environment), although that action had not yet been taken.
Participants were also asked if they would like to receive more information about the specialized app privately, and if so, they were asked to provide their email addresses.This measure assessed their actual information seeking, as respondents voluntarily gave their email addresses to receive more information, thus adopting a real behavior relevant to the topic analyzed.With this variable, we answered the call for considering actual behaviors in sustainable consumption analyses (Carrington et al., 2010).Inspired by recent work (Park and Lin, 2020), we measured the actual behavior through a dichotomous variable to detect if respondents were willing to provide their contact addresses to obtain more information.
We also examined a prosocial "secondary" outcome closely related to the topic of the advertising message.Namely, consumer intentions to volunteer for related causes (e.g., joining environmental movements, volunteering for organizations working for environmental improvement).Volunteering is considered a prosocial act that implies a form of direct exchange between individuals and non-profit organizations (Romani and Grappi, 2014), in this case, organizations working for relevant environmental causes.As volunteers often meet face-to-face with recipients, and their activities usually require time, energy, and initiative (Lee et al., 1999), this type of "secondary" behavior is of a different nature compared to the other dimensions examined here.Considering volunteering together with intentions and actual behavior can provide insight into how message framing might affect both consumer responses directly related to the topic at hand and potential spillover effects.In this study, willingness to volunteer was measured using a three-item scale adapted from Romani and Grappi (2014;e.g., "In the future, I am likely to join environmental movements"; α = 0.87).The items were subjected to maximum likelihood EFA with oblique rotation (promax).Results showed that the items loaded correctly on the corresponding factors, with loadings greater than 0.50 on focal factors and less than 0.25 on other factors.CFA was performed to assess the psychometric characteristics of the measures.The fit of the model was adequate (χ 2 (150) = 210.28;p =.00; CFI = 0.98; NNFI = 0.97; RMSEA = 0.05; SRMR = 0.04; Hu and Bentler, 1999), and the factors had adequately high discriminant validity based on the χ 2 difference test (all χ 2 difference tests [Δχ 2 (df = 1)] were higher than 22.28, p <.05).The items were averaged to form the corresponding dimensions.Manipulation checks (detailed in Appendix C) confirmed that participants correctly perceived the stimuli as differently framed and realistic.No  differences emerged in the relevance or vividness of the messages.

Results
We performed separate moderated mediation analyses (PROCESS Model 7; Hayes 2022) for each dependent variable to examine the proposed model (Fig. 1).Tables 5 and 6 present the results of these analyses.
Table 5 describes the mediator variable model.The moderation effect of CPE on the impact of positive vs. negative message framing on elevation (b = − 0.55; t = − 2.73, p <.01; CI = − 0.88; − 0.22) and guilt (b = 0.40, t = 1.96, p =.05; CI = 0.06; 0.75) was significant.When the negative and neutral frames were compared, the moderation effect was marginally significant regarding its impact on guilt (b = 0.40, t = 1.82, p <.10; CI = 0.04; 0.77).However, there was no significant effect on elevation.The CIs of the conditional effects of message framing on elevation were significant when CPE was high (X1: effect = − 0.82; CI = − 1.57, − 0.07; X2: effect = − 1.00; CI = − 1.70, − 0.32) and not when CPE was low (X1: effect = 0.11; CI = − 0.54, 0.76; X2: effect = 0.64; CI = − 0.02, 1.30).The opposite trend occurred when considering the conditional effects of message framing on guilt.CIs were significant when CPE was low (X1: effect = − 0.68; CI = − 1.35, − 0.01; X2: effect = − 0.69; CI = − 1.37, − 0.01) and not when CPE was high (X1: effect = 0.52; CI = − 0.25, 1.30; X2: effect = 0.53; CI = − 0.19, 1.24).To ease the interpretation of the results, we plotted the effects of message framing on elevation and guilt for low, medium, and high levels of CPE (Fig. 3).The plot of the interaction effects on elevation (Fig. 3, Panel A) demonstrates that this emotion is experienced at a high level on average (above 3.5 in all conditions).Results suggest that when respondents were exposed to a positive message frame, the effect of the message on elevation did not depend on CPE.In contrast, when respondents were exposed to a negative message frame, the effect of the message on elevation was stronger when the CPE was high.Thus, H3a is supported.Considering the plot of the interaction effects on guilt (Fig. 3, Panel B), this emotion was experienced in a limited way on average (equal to or below 3.5 in all conditions).When respondents were exposed to a positive message frame, the effect of the message on guilt was stronger when the CPE was high.The opposite occurred when they were exposed to a negative message frame.Thus, H4a is not supported.
Table 6 shows that elevation influenced all consumer responses.Therefore, H3b is supported.However, the anticipated emotion of guilt only affected volunteering intentions.Thus, H4b is only partially supported.The indexes of the moderated mediation (Hayes, 2015) formally tested the effects.As shown in Table 6, Panel A, the mediation of elevation depending on the message framing (X2-negative vs. positive) on intention to buy was moderated by CPE: the index of moderated mediation was − 0.09 and the CI was between − 0.19 and − 0.01.The same pattern occurred for WOM (Table 6, Panel B; index = − 0.18; CI = − 0.32, − 0.05), actual consumer information-seeking (Table 6, Panel C; index = − 0.54; CI = − 1.06, − 0.19), and volunteering intention (Table 6, Panel D; index = − 0.16; CI = − 0.29, − 0.03).The moderated mediation effects of guilt were not significant across the dependent variables, with the exception of volunteering intention (Table 6, Panel D; index = 0.07; CI = 0.00, 0.17).We conclude that the moderated mediation path hypothesized in H3 is fully supported.Message framing influences consumer responses through the mediation process of anticipated elevation, whose strength depends on CPE.The hypothesized moderated mediation path of anticipated guilt (H4) is verified to some extent for volunteering but not for the other dependent variables. 1

General discussion
We examined the impact of message framing on young adults' sustainable fashion consumption responses to explore the most effective X = manipulation, W = moderator, M = mediator, C = control.Continuous variables are mean centered for the analysis.Bolded parameter estimates correspond to the significant focal interactions hypothesized in the model illustrated in Fig. 1.
1 Although we agree with recent research contending the importance of using realistic stimuli to reliably measure and analyze consumer responses to advertising messages (i.e., White et al., 2019), we also acknowledge the importance of ruling out the risk of the possible presence of confounding factors introduced by the use of both visuals and text in the experiment.We addressed this issue by running an additional study that used a text-only manipulation (the stimuli used the text illustrated in Appendix A, without visuals).Respondents, 123 UK-resident young-adult consumers, correctly recognized messages with neutral, negative, and positive frames as different, even without visuals; F(2, 120) = 113.23,p <.01.The method and analyses are consistent with those presented in Study 2, as well as the results obtained: (a) elevation is the primary emotion stimulated by the advertising message; (b) a positive message frame favors the raising of elevation regardless of CPE; (c) a negative message frame stimulates elevation depending on CPE; (d) elevation fosters all consumer responses.Guilt showed no role in this case.Thus, this additional study corroborated our findings and excluded the risk of potentially confounding effects connected with the presence of visuals.(Details of the additional study are available upon request to the authors.).
S. Grappi et al. framing technique.Findings show that consumers are generally more inclined to adopt sustainable fashion consumption responses when exposed to a positive message frame.Subjective beliefs about the ethicality of communication strategies explain how this occurs.Consumers have an emotional reaction first, which is consistent with the role played by consumers' emotional needs for positive experiences in fashion consumption (Bishnoi and Singh, 2021;Trudel et al., 2020).Elevation plays a central role in the underlying process hypothesized here, as it is the primary emotion stimulated by the advertising message.Further, a positive message frame favored the raising of elevation regardless of consumers' subjective beliefs about the ethicality of advertising promoting sustainable consumption, whereas a negative message frame stimulated elevation depending on CPE.Elevation fosters all consumer responses.These results, although should be interpreted while considering the research context (i.e., the fashion context in which the purchasing process is characterized by the search for positive experiences), corroborate previous research demonstrating that individuals in a positive emotional state are motivated to maintain this state and are less emotionally connected to frames that might elicit negative emotions and, thus, less persuaded by them (Jin and Atkinson, 2021).The role of guilt was more limited, emerging only for the prosocial "secondary" outcome of volunteering intentions.

Theoretical implications
While prior research often supports the supremacy of negative message framing in fostering sustainable behaviors (Amatulli et al., 2019;Brennan and Binney, 2010), we challenge the efficacy of negative frames in favoring sustainable consumption in all situations.In the fashion context, we show that positive message frames are effective in supporting sustainable fashion consumption responses and that this effect is superior to that of negative message frames.In response to calls for more evidence about how different message framing affects sustainable consumption and ethical behavior (Randle et al., 2016;Reinhart et al., 2007), we establish that the positive message frames can better foster sustainable fashion consumption and that elevation plays a pivotal role in the underlying process.As such, we contribute to an emerging trend of designing positive sustainability initiatives (Peter and Honea, 2012;Wang et al., 2017;Winterich et al., 2019) and address calls for research to examine how positive emotions enhance sustainable behavior (White et al., 2019).
Furthermore, we contribute to the emerging literature on inspiration in marketing (Rudd et al., 2018).Recent research shows that inspiration can be elicited from marketing communications (Böttger et al., 2017;Winterich et al., 2019).We extend this by proposing novel elicitors of inspiration; positively framed messages (and negatively framed messages reaching consumers with high CPE) soliciting sustainable consumption elicit elevation, inspiring consumers to engage in the desired behaviors.In line with recent research where inspiration transforms the reception of a stimulus to the intrinsic pursuit of a consumption-related goal (Böttger et al., 2017), we demonstrate that the process of being inspired involves the self-transcendent emotion of elevation.Results confirm the nature of elevation as igniting action tendencies (Haidt, 2003).Elevation is the main driver of consumer responses to messages supporting sustainable behavior.The influence of elevation on sustainable consumption responses is determined by these responses being perceived as connected with higher-level outcomes (Fry, 2010;Piacentini and Banister, 2009) and are able to inspire consumers to improve themselves (e.g., improving self-esteem and being an example worth following).As a result, the expected emotional benefits associated with these outcomes strongly affect an individual's subsequent actions.Thus, our research shows that marketing communications have the potential to increase consumer discovery of new possibilities (i.e., sustainable consumption), which can activate components of inspiration and corresponding feelings of elevation.In doing so, we offer insights into the complex task of eliciting authentic, inspirational experiences (Thrash   S. Grappi et al. et al. 2014).
Our findings also showed that the negative consequence (i.e., feeling guilty) connected with the failure to adopt sustainable behavior plays a minor role, with only volunteering intentions affected by this emotion.Failure to adopt sustainable consumption responses likely decreases current self-esteem to a limited extent.Thus, the negative anticipated consequences of this failure drive behaviors to a limited degree.A possible explanation can be found in the neutralization techniques that individuals use to deflect blame for deviant or undesirable behavior (De Bock and Van Kenhove, 2010).Neutralization techniques help explain how consumers act in ways that contradict their prosocial beliefs and still preserve a positive self-image (Antonetti and Maklan, 2014), for example, by feeling less guilty when adopting unsustainable behaviors.Consumers might implement denial mechanisms (such as denial of responsibility or injury) that justify deviant behaviors (i.e., not acting as suggested) without feeling guilty.These mechanisms could explain the marginal role played by anticipated guilt in this study.Further research could examine the possible effects of neutralization techniques that consumers might implement on the proposed model.
Our findings demonstrate the key role of consumer subjective perception of the ethicality of advertising promoting sustainable consumption (i.e., CPE).This dimension has been largely unexplored until now; however, its effects are relevant in explaining consumer responses to advertising messages (Brunk and Blümelhuber, 2011).We show the extent to which individual beliefs regarding the ethicality of advertising differently shape consumers' emotional reactions to a framed message; emotions in turn, affect consumer responses.Thus, our research not only acknowledges the need to communicate ethically to consumers as recent research suggests (Antonetti and Maklan, 2014) but also demonstrates that consumers have different and subjective inclinations when evaluating the ethicality of advertising.When respondents are exposed to a positive message frame, the positive emotional response is always high, regardless of consumers' perceived ethicality.When respondents are exposed to a negative message frame, the moderating role of perceived ethicality is crucial in defining the effect of framing on emotional response.Supporting the significant moderating role of subjective beliefs, the current results contribute to explaining, to some extent, the incongruity seen in the results of previous studies, which alternatively claim the superiority of positive or negative message frames.Our findings demonstrate the need to consider individual belief systems, suggesting a more in-depth view regarding which message frame is most effective in stimulating sustainable consumption.Further research is needed to strengthen these insights.
It is worth emphasizing that we examined different types of sustainable fashion consumption responses: intentions, actual behavior, and the prosocial "secondary" consumer response of volunteering time to related causes.Thus, a broad spectrum of consumer responses is examined, providing a strong and reliable test of the proposed model.
Finally, this research examines sustainable fashion consumption of young adults in a specific buying environment, online apps, which are attracting more consumers.This allows the identification of insights whose potential lies both in the present and the future.We expect our study to stimulate further research considering other sustainable fashion consumption behaviors and broadening the population examined.

Practical implications
Through the analysis of consumer responses to different invitations to participate in ethical fashion consumption, this research suggests practical guidelines on how to increase the effectiveness of communication strategies for the promotion of sustainable fashion consumption.It suggests how marketers and policymakers can develop successful tools to encourage sustainable consumption behavior.Results indicate that the positive anticipated emotion of elevation engenders a stronger reaction to the featured advertisement than the negative anticipated emotion of guilt.
Operationally, this suggests that sustainable fashion consumption should be promoted through positive emotions and presents an opportunity to re-conceptualize sustainable consumption in a positive and inspiring way.The findings show that a positive message frame elicits elevation regardless of CPE, while a negative message frame stimulates elevation in a CPE-dependent manner.Thus, practitioners should favor positively framed messages or, in the case of negatively framed messages, ensure they reach a target audience that perceives negatively framed messages to be ethically justified to achieve the desired goals, thus avoiding possible boomerang effects.To successfully elicit sustainable behaviors, marketers and practitioners should consider using messages that positively depict sustainable consumption and the emotion of elevation connected with sustainable behavior implementation rather than negative messages related to the non-adoption of such behaviors (e.g., pollution and climate change problems).This has been seen in recent campaigns from big brands, such as Levis and Starbucks, who promoted sustainable consumption behaviors by relying on positive feelings (e.g., "Do good, Feel good" for the Levi's campaign) and on positive consequences for the individual and for the environment (e.g., "Save your money.Save our environment" for the Starbucks campaign).
It is useful to consider that even though the negative anticipated emotion of guilt plays no role in influencing direct sustainable behavior, it affects a relevant "secondary" consumer response: volunteering.From a practical point of view, policymakers and non-profit organizations could promote active citizenship spillover behavior, such as volunteering.However, they should reflect on the ethics of manipulating perceptions of individual agency through the elicitation of guilt, considering that volunteering intention is affected by positive emotions too.

Limitations and future research directions
Although this study provides insights into how to encourage the adoption of sustainable fashion consumption, several limitations and hence opportunities for future research should be mentioned.First, we use scenario-based experiments to test the proposed model, featuring boundary conditions of the effects of message framing on sustainable consumer fashion consumption.However, scenario-based experiments that maximize internal validity provide less ecological validity than field data.Due to our focus on testing the psychological mechanism underlying the effects of message framing, this design appears satisfactory.We sought to address its weaknesses by studying cases inspired by actual communication campaigns and making the stimuli as realistic as possible.Nonetheless, further research should establish experimental studies that involve real campaigns using positive and negative message framing.
Second, we masked the company's name to avoid any confounding brand-and company-related effects.Using actual cases of communication campaigns would require controlling for these effects, such as brand familiarity, brand love/hate, brand attitudes, and relationship quality.These elements could impact consumers' emotional reactions and responses.To test the robustness of the effects observed, future research could include actual companies and real brands and potentially brandand company-related moderating factors.
Third, additional dimensions could play a role in the process we proposed, such as an individual's guilt-proneness (Agrawal and Duhachek, 2010) and elevation-proneness (Thrash et al., 2014).These could affect the extent to which consumers feel these emotions when exposed to specific stimuli (e.g., messages).Future research could consider these two traits to enhance the robustness of the findings presented.
Study 2. We used the same descriptions objectively framed as positive, negative, or neutral in Study 1 as scenarios.Respondents correctly perceived the message frames (F(2, 141) = 5.83, p <.01).The pairwise comparisons showed statistically significant differences between scenarios (M

Fig. 1 .
Fig. 1.Conceptual model and overview of hypothesis testing.Four sustainable fashion consumption responses are examined: Intention to buy second-hand clothes (examined in Study 1 and Study 2); Positive WOM intentions (examined in Study 1 and Study 2); Consumers' actual information seeking (examined in Study 2); Consumer intentions to volunteer for related causes (examined in Study 2).Each consumer response is analyzed separately.
Panel A) Intention to buy second-hand clothes via specialized apps Panel B) Positive WOM about buying second-hand clothing via specialized apps

Fig. 2 .
Fig. 2. Study 1: Interactions between the message framing and CPE for intention to buy second-hand clothes via specialized apps (Panel A) and positive WOM (Panel B).
dependent variable, C = control.Continuous variables are mean centered for the analysis.Bolded parameter estimates correspond to the significant key effects hypothesized in the model illustrated in Fig.1.S.Grappi et al.

Fig. 3 .
Fig. 3. Study 2: Interactions between the message framing and CPE for elevation (panel A) and guilt (panel B). * in Panel B (guilt), the two interpolation lines for the positive frame message and the neutral frame message are almost overlapping, as very similar.

Table 2
Overview of empirical studies.
Manipulation: Message framing.Dependent variables: Intention to buy second-hand clothes via specialized apps and positive WOM.Moderator: CPE.Controls: Environmental concerns, disgust associated with wearing second-hand clothing, perceived costliness, frequency of purchase of clothing, frequency of purchase of second-hand clothing, age, and gender.Research design: Between-subjects experiment.Stimuli: Fictitious advertising message objectively framed as negative, positive, or neutral (Appendix A).Debrief: At the end of the study, respondents were debriefed regarding the research purpose, confirming that the questionnaire was anonymous, they were randomly assigned to different experimental conditions, and was for scientific research to measure characteristics or intentions.Respondents: 125 UK-resident young-adult consumers.Sample characteristics: 55 men (44 %); the participants' average age was 28 (SD = 10.23;min=18;max = 40).People with a high school education (or lower) accounted for 13.2 % of the sample, those with a bachelor's degree accounted for 49.6 %, and those with higher degrees represented 37.2 %.Of the respondents, 34.6 % were from Northern England and Scotland, 31.2 % from Mid England, Wales, and Northern Ireland, and 34.2 % from Southern England and Greater London.The experimental groups did not vary in gender (χ 2 (4) = 5.72, p =.22), age (F(2, 122) = 2.01, p =.14), educational level (χ 2 (10) = 8.70, p =.56), or region of residence (χ 2 (12) = 14.75, p =.26).Thus, any differences between them was not the result of demographic characteristic differences.Manipulation: Message framing.Dependent variables: Intention to buy second-hand clothes via specialized apps, positive WOM, actual consumer information-seeking, and volunteering.Moderator: CPE.Mediators: Elevation and guilt.Controls: Environmental concerns, disgust associated with wearing second-hand clothing, perceived costliness, frequency of purchase of clothing, frequency of purchase of second-hand clothing, age, and gender.Research design: Between-subjects experiment.Stimuli: Fictitious advertising message objectively framed as negative, positive, or neutral (Appendix A).Debrief: Participants were debriefed about the research purpose at the end of the study, confirming they were completing an anonymous questionnaire and the intent was purely scientific.Researchers clarified that any contact addresses that the respondents provided would be deleted from the database immediately, and that no one would have access to this data.Respondents: 145 UK-resident young-adult consumers.Sample characteristics: 78 men (53.8 %) participated; the average age was 28 (SD = 8.23; min = 18; max = 40).People with a high school education (or lower) accounted for 12.5 % of the sample, those with a bachelor's degree accounted for 51.6 %, and those with higher degrees accounted for 35.9 %.Of the respondents, 31 % were from Northern England and Scotland, 35.2 % from Mid England, Wales, and Northern Ireland, and 33.8 % from Southern England and Greater London.The experimental groups did not differ in age (F(2,141) = 0.47, p =.63), gender (χ 2 (4) = 1.06, p =.90), educational level (χ 2 (10) = 13.51,p =.20), or region of residence (χ 2 (12) = 15.23,p =.23).Thus, any differences between them did not result from demographic characteristic differences.

Table 3
Study 1: multivariate analysis of variance.
X = manipulation, W = moderator, Y = dependent variable, C = control.Continuous variables are mean centered for the analysis.S.Grappi et al.