Journal of Retailing and Consumer Services: Theresa Maria Rausch, Daniel Baier, Stefanie Wening

Journal of Retailing and Consumer Services 63 (2021) 102681
Contents lists available at ScienceDirect
Journal of Retailing and Consumer Services

journal homepage: www.elsevier.com/locate/jretconser
Does sustainability really matter to consumers? Assessing the importance of

online shop and apparel product attributes
Theresa Maria Rausch a, *, Daniel Baier a, Stefanie Wening b
a
University of Bayreuth, Universitätsstraße 30, 95447 Bayreuth, Germany
b
BAUR Versand, 96222 Burgkunstadt, Germany
A R T I C L E I N F O A B S T R A C T
Keywords: Although there is a shift in consumers’ consumption behavior towards more sustainable patterns across a variety
Sustainable apparel of different contexts, sustainable apparel has still not become a mainstream trend despite the textile industry’s
Online shopping excessive usage of valuable resources. Albeit extant research found different potential barriers elucidating why
Willingness to pay
consumers hesitate to purchase such apparel, it remains unclear whether sustainability really matters to con
Best-worst scaling
E-commerce
sumers in a clothing context and further, which aspects are of relevance during consumers’ purchase decision.
We thus conducted two studies with four best-worst scaling experiments in which 4,350 online shoppers assessed
the importance of both conventional and sustainable apparel attributes, as well as sustainable apparel attributes
only, and the willingness to pay for sustainable product attributes. We further inquired the importance of con
ventional as well as sustainable online shop attributes. Our findings indicate that conventional apparel attributes
such as fit and comfort, price-performance ratio, and quality are of higher relevance to consumers than sus
tainable attributes. The most important sustainable apparel attributes are the garment’s durability, fair wages
and working conditions, as well as an environmentally friendly production process. Consumers also indicated to
prefer the latter three attributes to a 20% discount. Moreover, consumers demand less as well as sustainable
packaging, free returns, and discount campaigns. Our findings reveal a gender gap regarding green consumerism
with female respondents assessing most sustainable attributes as more important than male respondents do.
1. Introduction is still marginal: Particularly consumers older than 35 years were found
to have a neutral rather than a supportive attitude towards sustainable
Across a wide variety of industries, sustainability developed from a fashion (KPMG, 2019). Apparently, in the clothing context, a vast ma
fringe to a mainstream issue for manufacturers, managers, marketers, jority of consumers struggle to translate its overall green attitude into
and further stakeholders throughout the past years. The shift in con green actions (Kollmuss and Agyeman, 2002; Young et al., 2010). Albeit
sumers’ mindsets and increasing awareness regarding their environ research frequently took a macro-perspective and found single aspects
mental impact triggered research to gather a better understanding of such as perceived higher prices (Hiller Connell, 2010; Joergens, 2006) as
green purchase behavior, particularly in an online context as consumer well as perceived aesthetic risk (Hiller Connell, 2010; Rausch and
behavior is increasingly shifting towards a virtual one (Kautish et al., Kopplin, 2021) to deter consumers from buying sustainable clothing,
2019; Maichum et al., 2016; Paul et al., 2016; Taufique and Vaithia literature still lacks a micro-perspective on the topic: a comprehensive
nathan, 2018; Yadav and Pathak, 2016, 2017). understanding regarding the importance of concrete sustainable apparel
Especially the textile industry can be considered a black sheep, and attributes is needed and, further, their importance in comparison with
its environmental harm is frequently underestimated by consumers: it conventional apparel attributes to successfully target the most crucial
caused approximately 2.1 billion tons of greenhouse gas emissions in aspects. It remains unclear whether sustainability really matters to
2018 and thus, accounted for around 4% of the annual emissions glob consumers in a clothing context and, more specifically, which aspects
ally, which is equivalent to the combined annual emissions of France, are of importance to consumers during their purchase decision. Partic
Germany, and United Kingdom (McKinsey & Company and Global ularly during disruptive situations, e.g., throughout the global
Fashion Agenda, 2020). Nevertheless, sustainable apparel consumption COVID-19 pandemic, sustainability seems to play only a minor role to
* Corresponding author.
E-mail addresses: theresa.rausch@uni-bayreuth.de (T.M. Rausch), daniel.baier@uni-bayreuth.de (D. Baier), stefanie.wening@baur.de (S. Wening).
https://doi.org/10.1016/j.jretconser.2021.102681
Received 12 January 2021; Received in revised form 23 June 2021; Accepted 13 July 2021
Available online 24 July 2021
0969-6989/© 2021 Elsevier Ltd. All rights reserved.
T.M. Rausch et al. Journal of Retailing and Consumer Services 63 (2021) 102681
consumers (Barbier und Burgess, 2020; Hakovirta und Denuwara, Morgan and Birtwistle, 2009; Scherer et al., 2018), but good working
2020). Moreover, as sustainable apparel is frequently perceived as more conditions and fair wages (Fulton and Lee, 2013; Stöckigt et al., 2018)
expensive than conventional apparel (Hiller Connell, 2010), research on may also be an important sustainability facet. Thereby, increasing in
consumers’ willingness to pay for sustainable product attributes is formation complexity surrounding sustainable products and, as a result,
needed as initial attempts mainly cover surcharges for materials (Ellis higher search costs for consumers were frequently identified to be
et al., 2012; Ha-Brookshire and Norum, 2011). As the textile industry among the main barriers to sustainable clothing consumption (Ellen,
yields the highest online sales across all e-businesses (Statista, 2020), 1994; Harris et al., 2016). To shift consumption behaviors towards more
assessing the importance of conventional as well as sustainable online sustainable ones, prior literature thus recommended providing better
shop attributes may be of relevance to trigger sustainable online con information and education about materials used or manufacturing
sumer behavior. To sum up, we identify the following research gaps: conditions, e.g., with eco or social labels, to consumers (D’Souza et al.,
RG1. A micro-perspective on sustainable clothing consumption 2007; Hiller Connell, 2010).
behavior in terms of a comprehensive understanding regarding the Considering the later stages of the garment’s lifecycle, and specif
importance of sustainable apparel attributes, particularly in comparison ically, the product’s discard, it may be of importance to consumers
with conventional apparel attributes. during their purchase decision whether the material is recyclable or
RG2. A sound understanding of consumers’ willingness to pay for biodegradable to close the garment’s lifecycle (Fulton and Lee, 2013).
sustainable apparel attributes. Further, companies offering return programs to recycle the garment
RG3. A thorough investigation of the importance of both sustainable were found to appear attractive to consumers (Baier et al., 2020).
as well as conventional online shop attributes. Overall, female consumers were frequently found to be more likely to
To fill these research gaps, we intend to answer the following engage in sustainable behaviors1 than male consumers (Eisler and Eisler,
research questions: 1994; Luchs and Mooradian, 2012) as women tend to be more altruistic
RQ1. Which is the most important sustainable clothing attribute to as well as prosocial (Lee and Holden, 1999). Men associate greenness
consumers? with femininity and thus, avoid sustainable behaviors to maintain their
RQ2. How important are specific sustainable clothing attributes to gender-identity (Brough et al., 2016).
consumers compared to conventional clothing attributes? Despite these initial exploratory findings on potentially essential
RQ3. How important are specific sustainable online shop attributes sustainable apparel attributes, there is little knowledge on the impor
to consumers compared to conventional online shop attributes? tance of these different attributes with regard to consumers’ clothing
RQ4. What are consumers willing to pay for specific sustainable purchase behavior. Extant research primarily investigated the impor
clothing attributes? tance of conventional intrinsic as well as extrinsic attributes for con
We answer these research questions by conducting two studies with sumers’ purchase decision of apparel products: There is vast consensus
four different best-worst scaling (BWS) experiments: We gathered data within literature that aesthetic criteria with regard to the garment’s
of 4,350 online shoppers assessing the importance of both conventional physical appearance such as color, pattern, and design have the greatest
and sustainable apparel attributes as well as sustainable apparel attri impact on consumers’ purchase decision (Abraham-Murali and Littrell,
butes only and further, the willingness to pay for sustainable product 1995; Baier et al., 2020; Fiore and Damhorst, 1992; Viciunaite and
attributes. Moreover, we asked the respondents to assess the importance Alfnes, 2020; Zhang et al., 2002). Visual characteristics are assumed to
of conventional as well as sustainable online shop attributes. To gain fulfill implicit expectations such as fashionability, aesthetic appeal, and
deeper insights, we further compare the results of male and female self-expression (Eckman et al., 1990). This seems particularly critical
respondents. with regard to sustainable clothing consumption, as consumers
The remainder of this paper is structured as follows: Section 2 re frequently perceive sustainable apparel as unfashionable (Hiller Con
views related literature and Section 3 outlines the theoretical basics of nell, 2010; Joergens, 2006), deterring consumers from purchasing
the BWS methodology. We then present our methodology, followed by despite initial purchase intention (Rausch and Kopplin, 2021). Sus
the conceptualization and results of our four BWS experiments. Next, we tainable clothing is mostly associated with specific stereotypes inhibit
discuss our findings in the light of theoretical contribution, practical ing mainstream consumption (Connolly and Prothero, 2003).
implications, and limitations. Lastly, we draw a conclusion. Aside from aesthetic criteria, the garment’s quality and physical
performance is another essential purchase criterion, i.e., attributes
2. Theoretical background which are instrumental outcomes of the product’s physical aspects
(Abraham-Murali and Littrell, 1995; Viciunaite and Alfnes, 2020; Zhang
2.1. Apparel attributes affecting consumers’ purchase decision et al., 2002). Physical performance comprises – among others – aspects
concerning the overall fit, comfort, the extent to which the garment can
Sustainable clothing implies pro-environmental aspects throughout be maintained in a wearable condition during care, and workmanship (i.
the whole lifecycle of a garment, from the pre-purchase, purchase, to the e., the level of excellence regarding construction as well as materials)
post-purchase phase (Jacoby et al., 1977; Lundblad and Davies, 2016; (Eckman et al., 1990; Hopfer and Istook, 2016; Zhang et al., 2002).
Morgan and Birtwistle, 2009). Albeit sustainable clothing is mostly Despite the striking importance of aesthetic and performance
associated with logistics or material usage and the respective environ criteria, research found extrinsic criteria to exhibit a comparable impact
mental impact, there is no industry standard uniformly regulating the on consumers’ purchase decision: Price and perceived value for money
notion (Joergens, 2006; Lundblad and Davies, 2016). Within literature, are often among the most decisive purchase criteria for most consumers
the concept of sustainable clothing extends well beyond environmental (Abraham-Murali and Littrell, 1995; Jegethesan et al., 2012; Viciunaite
aspects (e.g., usage of environmentally friendly materials) by further and Alfnes, 2020; Zhang et al., 2002). Moreover, the garment’s brand
comprising social aspects (e.g., working conditions of employees) (Ful and, accordingly, brand familiarity are crucial to some consumers
ton and Lee, 2013; Goworek et al., 2012; Lundblad and Davies, 2016). (Abraham-Murali and Littrell, 1995; Eckman et al., 1990; Jegethesan
Throughout the early stages of the apparel lifecycle, employee et al., 2012). Although it is of less importance than the garment’s price,
wages, working conditions, amount of pesticide usage, material usage, it still plays an important role during the purchase decision process
and the country-of-origin are among the key aspects determining the
product’s sustainability (Goworek et al., 2012; Hustvedt and Dickson,
2009; Lundblad and Davies, 2016): Intuitively, sustainable clothing is 1
Sustainable (consumption) behavior refers to environmentally friendly
frequently referred to recycled/upcycled or bio-based materials and behavior across all facets of consumers’ daily life such as nutrition, means of
locally manufactured products by most consumers (Allwood et al., 2008; transportation, waste disposal, etc. aside from clothing.
2
(Abraham-Murali and Littrell, 1995; Hopfer and Istook, 2016; Zhang as they are to changes in the product’s prices. Further, prior positive
et al., 2002). return experiences were found to enhance the customer’s future buying
Research comparing the importance of both sustainable and con behavior and thus, his or her lifetime value for the organization
ventional product attributes is sparse and mostly focused on materials as (Petersen and Kumar, 2009; Wood, 2001). Despite the striking impor
a sustainable product attribute: Although research indicated that envi tance of delivery speed and costs, the environmental impact of the
ronmentally friendly materials as a sustainable product attribute may shipping procedure can potentially influence consumers’ purchase de
influence consumers’ purchase decision, conventional product attri cisions (Stöckigt et al., 2018).
butes were found to be more important (Viciunaite and Alfnes, 2020). Particularly in an e-commerce context, the availability and channels
Albeit the study of Viciunaite and Alfnes (2020) provides first insights to contact customer service seem to play an important role (Cao et al.,
into the significance of sustainable apparel attributes compared to 2018). It is well known that perceived service quality affects customer
conventional apparel attributes, they only consider few sustainable loyalty (Parasuraman et al., 1985; Zeithaml et al., 1996). This relation
apparel attributes (particularly materials) and, thus, lack a holistic may be even more critical in an online context as consumers do not have
understanding. the chance to interact face-to-face with agents like in brick-and-mortar
stores and hence, may assess service quality as more intangible.
2.2. Willingness to pay for sustainable clothing attributes In a sustainability context, further online store attributes may be of
relevance: E.g., consumers’ demand for less packaging waste and/or
Consumers often exhibit an attitude-behavior gap regarding sus sustainable packaging alternatives are steadily increasing and has been a
tainable clothing, i.e., albeit they pretend environmentally friendly at fruitful path to investigate within literature throughout the past years
titudes, they frequently struggle to translate this into green actions (Nguyen et al., 2020; Prakash and Pathak, 2017; Schwepker and Corn
(Kollmuss and Agyeman, 2002; Young et al., 2010). One of the main well, 1991). Thereby, purchase intention for eco-friendly packaging was
barriers towards buying sustainable clothes is the perceived increased particularly found to be determined by altruistic motives, i.e., con
economic risk associated with such clothes: Preceding research found sumers’ environmental concerns (Prakash et al., 2019; Prakash and
consumers to consider the price of sustainable apparel as higher Pathak, 2017).
compared to conventional clothes, and thus, they do not purchase such Generally, consumers value a large range of sustainable products as
clothes despite their initial positive attitude (Hiller Connell, 2010; one of the main barriers towards sustainable clothing consumption is the
Hustvedt and Dickson, 2009; Joergens, 2006). Since sustainable clothes lack of availability: Sustainability in a clothing context frequently im
are hardly produced for the mass market, they often carry higher prices, plies fewer fashion cycles as well as collections, few different styles (e.g.,
and thus, are perceived as less affordable compared to conventional in terms of color), and a smaller product range (Hiller Connell, 2010;
clothes (Hiller Connell, 2010). However, extant research determining Lundblad and Davies, 2016; Pookulangara and Shephard, 2013) as it is
consumers’ willingness to pay for different sustainable clothing attri restricted to natural materials.
butes and thus, trying to overcome the attitude-behavior gap is still Table 1 summarizes the findings of the most recent literature on the
sparse. importance of sustainable apparel and online shop attributes. Overall,
Mostly, research investigated consumers’ willingness to pay for albeit the importance of single aspects has been indicated by extant
materials: Ellis et al. (2012) found consumers to pay 25% more for research, literature still lacks a holistic comparison and evaluation of the
t-shirts made of organic cotton, aligning with Ha-Bookshire and importance of both conventional as well as sustainable store attributes.
Norums’ (2011) findings that consumers are willing to pay a higher
price for domestic-grown organic cotton shirts. Similarly, Hustvedt and 3. Best-worst scaling
Bernard (2008) found consumers to be willing to pay a premium for
socks labeled as organic and produced locally. Similar findings were BWS is a rather new measurement approach for the subjective value
gathered in the food context, where research found a higher willingness (or importance, utility) of items (objects or attribute-levels for objects)
to pay for local and organic food (de-Magistris and Gracia, 2016; Gil that is based on the random utility framework by Thurstone (1927) and
et al., 2000). McFadden (1974). In this framework, it is assumed that individuals
Aside from materials and country-of-origin, consumers were found to evaluate stimuli (objects, attribute-levels, or
pay a premium in case the product had a labor-related labeling indi attribute-level-combinations) on a subjective utility scale and that these
cating social responsibility and fair trade (Hustvedt and Bernard, 2010). evaluations form their basis for choosing the ‘best’ stimulus (with
An increased willingness to pay for (e.g., fair trade or eco) labels was maximum utility) among presented ones. These assumptions allow
further found to be applicable in a food context (Delmas and Lessem, deriving unknown subjective values of items from observed choice fre
2017; Paetz and Guhl, 2017; van Loo et al., 2015). quencies among stimuli (Louviere et al., 2013). The individual evalua
tion might be superimposed by an additive random error, but when
2.3. Online shop attributes affecting consumers’ purchase decision distributional assumptions are justified and individuals choose repeat
edly, means and variances of these subjective values can be estimated
Compared to brick-and-mortar stores, e-commerce businesses are (Louviere et al., 2015). Based on this framework, Jordan J. Louviere
confronted with several issues inherent to the online context. As online developed BWS in 1987 and applied the new approach together with
transactions are frequently perceived as risky (particularly in terms of Adam Finn for the first time (Finn and Louviere, 1992; Louviere and
privacy as well as data security or product performance) and were Woodworth, 1991).
associated with time as well as convenience loss (Forsythe and Shi, BWS is an alternative to rating, ranking, and to other choice-based
2003; Miyazaki and Fernandez, 2001), it is crucial for e-businesses to procedures that are common for measuring subjective values (Mühl
assess the importance of different attributes regarding the online store bacher et al., 2013): For BWS, the basic idea is to repeatedly confront
itself to mitigate consumers’ doubts. respondents with ‘choice sets’ of stimuli from which they have to select
Considering online store attributes which may be of importance to the ‘best’ and the ‘worst’ one (Cohen, 2003). In each of these choice
consumers during the purchase decision, research found particularly tasks, the respondents are assumed to compare and evaluate all stimuli
factors associated with shipping such as shipping fees, shipping speed, and select those two that they perceive to provide them the greatest and
and return policy to be essential decision criteria (Bower and Maxham, the least individual benefit. During this choice task, various cognitive
2012; Cao et al., 2018; Lewis, 2006; Ma, 2017; Oghazi et al., 2018; processes have to take place: Respondents identify and evaluate all
Stöckigt et al., 2018). E.g., Smith and Brynolfsson (2001) found con possible stimulus pairs. The utility differences between all pairs are
sumers to be approximately twice as sensitive to changes in shipping fees calculated and the pair is selected with the maximum difference (Flynn
3
Table 1 Table 1 (continued )

Recent literature’s findings on importance of sustainable apparel and online Author(s) Research objective(s) Method Key finding(s)
shop attributes.
sustainable apparel previous experience,
Author(s) Research objective(s) Method Key finding(s) and store attributes tactile features, and
Baier et al. Which sustainable (Segmented) A possibility to return the environmental
(2020) apparel and online Kano method disposed garments, a impact of materials.
shop attributes discount for Regarding store
enhance customer sustainable garments, attributes, consumers
satisfaction? a sustainability level put most emphasis on
indicator, and bio- the selection, nice
based as well as staff, and a positive
recycled materials are previous experience.
highly attractive
apparel and online
shop attributes to et al., 2007). This stochastic selection process can be expressed formally
consumers by the following so-called MaxDiff formulation:
Cao et al. Identification of Structural Customer service and (
(2018) determinants of Equation return service are the ( ) ( ) (( )
customer satisfaction Modeling most important Prob((j1 ; j2 )|CSik ; i) = Prob uij1 + εij1 − uij2 − εij2 ≥ ′ max ′ uij + εij
j,j ∈CSik ,j∕
=j
when purchasing determinants of )
online customer satisfaction
( ))
− uij′ + εij′
when purchasing
online and further,
future purchase Prob((j1 ; j2 )|CSik ; i) corresponds to the probability with which
intention respondent i (i = 1, …,I) assesses stimulus j1 ∈ CSik as the best and
Oghazi et al. Identifying the role Structural A lenient return policy stimulus j2 ∈ CSik as the worst when the choice set CSik is presented (k =
(2018) of a retailer’s return Equation enhabnces consumer
policy on consumers’ Modeling trust and thus,
1, …,K). The expression (uij1 +εij1 ) − (uij2 − εij2 ) represents the difference
purchase decision consumers’ online between the evaluations of stimulus j1 and stimulusj2 on the underlying
purchase decision scale. uij is the mean subjective value respondent i allocates to stimulus j
Lundblad Identification of In-depth Consumers perceive and εij is the additive random error. The MaxDiff expression
and values and interviews; sustainable clothing
Davies motivations Means-End as more durable, of
max ′ ((uij +εij ) − (uij′ +εij′ )) ultimately contains the greatest difference
j,j ∈CSik ,j∕
=j
′
(2016) underpinning Theory higher quality, and

of all possible paired differences in CSik (cf. Finn and Louviere, 1992).
sustainable apparel Approach they expect a higher
consumption value for their money. The pairwise comparisons within the choice sets force the respondent to
Timeless cuts are make trade-off decisions between different alternatives (Lee et al.,
demanded 2007). As a result, the respondent can neither accept all possible an
Ma (2017) Determining the Scenario-based High delivery time
swers, evaluate them as well, reject them, nor classify them as bad.
effect of delivery role playing has negative effect on
time and shipping experiment consumers’ purchase
Nevertheless, the search for a pair with the highest difference in sub
charges on intention, but free jective values is closely related to the search for the ‘best’ stimulus with
consumers’ purchase shipping can the highest and the ‘worst’ stimulus with the lowest subjective value in
intention compensate for this. the set CSik . Ultimately, it is possible to estimate the means and variances
Rausch and Macro-perspective Partial Least Consumers’ purchase
of the subjective values from repeated choice tasks (Cohen, 2003).
Kopplin on sustainable Squares intention is rather
(2021) clothing Structural determined by According to Flynn and Marley (2014) a distinction can be made
consumption: Which Equation intrinsic (i.e., between three cases of measurement, i.e., (1) object scaling (Case 1), (2)
factors influence Modeling attitudinal) than profile scaling (Case 2), and (3) multiprofile scaling (Case 3). Object
consumers’ purchase extrinsic (i.e., social
scaling describes the basic archetype of BWS as discussed above. There is
intention for pressure from peers)
sustainable clothing? motives. Design or
a (small) set of objects (say 6 to 30) and the presented choice sets are
aesthetics are subsets of this set (Mühlbacher et al., 2013). The respondents have to
extremely important select the best and the worst stimulus/object in each subset (Finn and
to consumers and can Louviere, 1992). Further, if an i.i.d. exponential distribution for the
deter them from
additive random disturbance is assumed, we get a multinomial logit
buying despite an
initial purchase (MNL) formulation for the probability Prob((j; .)|CSik ; i) that stimulus j is
intention selected as the best (or most important) one by individual i in choice set
Scherer Determining the Choice-based Origin of materials, CSik :
et al. importance of Conjoint price, and weight are ( )
(2018) sustainable shoe analysis the three most exp uij
attributes for important attributes Prob((j; .)|CSik ; i) = ∑
exp(uij′ )
purchase decision when purchasing a ′
j ∈CSik
bio-based shoe
Stöckigt Determining the Choice-based Shipping costs, Finn and Louviere (1992) argue that the probability Prob((.; j)|CSik ; i)
et al. importance of Conjoint working conditions,
that stimulus j is selected as the worst (or least important) one in choice
(2018) sustainable apparel analysis and the
attributes for environmental impact
set CSik can be formulated similarly:
purchase decision are the three most ( )
important attributes
exp − uij
Prob((; j)|CSik ; i) = ∑ ( )
when purchasing exp − uij′
sustainable apparel
′
j ∈CSik
Viciunaite Identification of Best-Worst Regarding sustainable
and Alfnes consumer Scaling apparel attributes, For object scaling with I individuals, J objects, K choice tasks and
(2020) preferences for consumers put most choice sets CSik ⊂{1, ..., J} ∀ i ∈ {1, ..., I}, j ∈ {1, ..., J}, the observed se
emphasis on a positive lections of best and worst stimuli/objects across all choice sets and re
spondents can be used to derive a likelihood function across all
4
observations and the model parameters (the unknown mean subjective choice task and they have to select a best and a worst
values of the objects uij ∀ i ∈ {1,...,I},j ∈ {1,...,J}) are estimated using a attribute-level-combination (Mühlbacher et al., 2013). In both cases, the
Maximum Likelihood approach. The choice sets may vary across re subjective value of a stimulus/profile is assumed to be the sum of its
spondents, but (best possible) fulfillment of criteria like balance (objects attribute-level partworths. Again, using the above MNL formulation,
appear with the same frequency across all choice sets) and orthogonality Maximum Likelihood can be used for estimating means and variances of
(object pairs appear with the same frequency across all choice sets) subjective values, but with the attribute-level partworths as model pa
support the efficient and uncorrelated estimation of the model param rameters (Sawtooth Software, 2009a, 2009b, 2013, 2009b).
eters. It needs to be mentioned that the underlying MNL assumption of i. BWS has various advantages compared to conventional survey
i.d. errors across the selections of best and worst stimuli has been methods with rating scales. The crucial difference between BWS and
questioned by some authors since it may lead to biased estimates of traditional approaches is that BWS is an indirect approach where no
subjective values (e.g., Dyachenko et al., 2013). However, Horne and direct evaluation of objects or attribute-levels is necessary, i.e., the risk
Rayner (2013) showed in several analyses that this bias should not of possible distortion of the results can be overcome (Auger et al., 2007).
justify more advanced estimation procedures that additionally model The following biases can thus be counteracted:
ordering issues (as proposed, e.g., by Dyachenko et al. (2013)).
Instead, particularly Hierarchical Bayes (HB) with underlying MNL - Social desirability bias: Respondents tend to give answers that they
assumption is wide-spread for estimating the subjective values think are correct and socially accepted (Fisher, 1993).
(Sawtooth Software, 2009a, 2013). The advantage of this approach is - Acquiescence bias: tendency of respondents to generally agree to
the sharing of observations across respondents by assuming a higher questions regardless of their content (Schuman and Presser, 1981).
(aggregate) level model across all respondents besides the lower (indi - Extreme response bias: the respondents’ tendency to extreme re
vidual) level model. This assumption allows reducing the number of sponses (Culpepper and Zimmerman, 2006).
necessary observations (choice sets) per respondent dramatically. Soft
ware packages like MaxDiff from Sawtooth Software (2009a) provide With BWS, these distortions can be avoided, and there is no distor
easy access to BWS and HB estimation. When the same data is analyzed tion of mean values, which allows for valid comparisons (Cohen and
with Maximum Likelihood and HB, both approaches lead to very similar Orme, 2004). In addition, significant differences in the response styles or
aggregate results (mean subjective values across all respondents), as the use of rating scales between different countries have been empiri
Cheung et al. (2018) have demonstrated. According to this comparison, cally proven (Chen et al., 2016). Therefore, this method is frequently
even the Average BW Score (Louviere and Flynn, 2010; Mühlbacher used in market research in a country comparison (Lee et al., 2007).
et al., 2016) is useful to derive valid proxies for the mean subjective Further advantages are ease of use for the respondents (Marley and
values: Louviere, 2005) as well as pleasant implementations that allow a largely
uncomplicated analysis with the help of standard software (Cohen,
nBest − nWorst
Average BW Scorej =
j j 2003). Finally, BWS leads to more precise results than standard rating
nj scales (Cohen and Orme, 2004). This is the conclusion gathered by
This Average BW Score for object j is calculated by relating the Chrzan and Golovashkina (2006), who examined customer satisfaction
number of times j was selected as best stimulus (nBest ) minus the number with restaurant services using various methods of comparative scaling.
j
BWS turned out to be the model with the highest predictive validity. It is
of times object j was selected as worst stimulus (nWorst j ) to the total
also best suited to differentiate between the importance of objects
number of times object j was presented (nj ). However, the advanced HB
(Chrzan and Golovashkina, 2006). Further advantages that are partic
technique additionally allows deriving subjective values at the indi ularly relevant for this work are the high level of forecast reliability, the
vidual level – even with few observations per respondent – and thus, can achievement of reliable results, and the ability to identify a precise
be used to discuss segment-specific results by averaging them cross a subjective value structure between the items (Mühlbacher et al., 2013).
priori defined groups of respondents or by applying clustering ap This is particularly important with regard to the determination of the
proaches. Before averaging or clustering, the estimated mean subjective heterogeneity of the individual evaluations with regard to the objects
values uij ∀ i ∈ {1, ..., I}, j ∈ {1, ..., J} have to be zero-centered so that examined.
∑J
j=1 uij = 0 ∀ i ∈ {1, ..., I} holds. This transformation does not affect the However, there are also weaknesses: Aside from an increased
above individual choice probabilities. In some cases, the zero-centered expenditure of time, there is a certain pressure to make a decision for the
mean subjective values are then further transformed into probabilities respondents. In addition, the respondents cannot express likes or dislikes
( ) of all items. However, this can be counteracted by adding a rejection
exp uij
pij = ( ) option (Mühlbacher et al., 2013). Further disadvantages are the
exp uij + a − 1 increased need for explanation of the method, the more complex data
collection and analysis, and the availability of specific knowledge
where a is the size of the choice sets (number of stimuli presented in a
(Simon, 2010).
choice task). pij reflects the probability that object j is selected when
presented together with a − 1 other stimuli j’ with uij′ = 0 (the average
4. Methodology
subjective value if zero-centered). In contrast to the subjective values uij ,
these ratio-scaled probabilities pij can be interpreted more easily (j has We conducted our two studies in cooperation with BAUR, a large
an x-times higher probability to be selected than j’). Further, these German online retailer for fashion and furniture: BAUR generated 361.8
probabilities pij can additionally be normalized to Probability Scores so million Euros online sales in 2018 (Statista, 2019), and thus, was among
that they sum up to 100. the top 10 fashion e-businesses in Germany. BAUR is primarily targeting
Profile Scaling – Case 2 – differs from object scaling in that stimuli women, which are between 40 and 55 years old.
are attribute-level-combinations (profiles). The respondents are pre To analyze the importance of the attributes, we developed two online
sented one profile in a choice task and they have to select the best and questionnaires with Sawtooth Software’s Lighthouse Studio 9.8.1.
worst attribute-level in the profile (Flynn and Marley, 2014). Based on Before conducting the main study, the questionnaires were pre-tested
this assessment, it is possible to determine the overall benefit by adding with experienced participants and researchers to assess appropriate
up the respective partworths for each attribute-level (Marley and Lou ness, clarity, completeness, wording, and structure. Only minor
viere, 2005). Finally, with multiprofile scaling, the respondents are amendments were made. The final questionnaires were structured
presented more than one profile (attribute-level-combination) in a identically and consisted of three major sections. To gather deeper
5
insights into the participants’ consumption behavior and perception of discounts for sustainable products (mean = 4.50, SD = 1.12) to draw
sustainable consumption, we developed five introductory questions. We more attention towards sustainable products and to communicate sus
inquired the respondents’ general perception of sustainability (multiple tainability transparently. Respondents assess an organization’s sustain
selections with a maximum of three choices), their sustainable behav ability level mainly with regard to its manufacturing conditions (n =
ioral patterns in daily life (evaluation on a 6-point Likert type scale from 1,532, 72.7%) and use of pesticides as well as chemicals in its products
‘1 = Totally disagree’ to ‘6 = Totally agree’), barriers towards making (n = 1,385, 65.8%). Table A.2 outlines the descriptive statistics of Study
sustainable purchases (multiple selections), assessment of options to 2.
transparently communicate sustainability (evaluation on a 6-point Lik
ert type scale from ‘1 = Very unimportant’ to ‘6 = Very important’), and 5. Study 1a: Assessing sustainable and conventional apparel
assessment of attributes indicating an organization’s sustainability attributes
(multiple selections with a maximum of three choices). The main part
comprised the respective BWS experiments (object scaling): Within 5.1. Conceptualization
Study 1, we inquired sustainable and conventional apparel attributes
(Study 1a) as well as the willingness to pay for sustainable apparel at As a first step, for both studies, we cooperated with several experts of
tributes (Study 1b). Within Study 2, sustainable apparel attributes BAUR to appropriately determine practically relevant attributes, which
(Study 2a) and sustainable and conventional online shop attributes should then be assessed by the respondents. This procedure comple
(Study 2b) were assessed. Lastly, we collected the respondents’ ments our literature findings from a practical perspective. To evaluate
demographics. the identified attributes, we draw on a BWS experiment like preceding
The two surveys were spread via the BAUR newsletter from July 10, literature on sustainable clothing (Viciunaite and Alfnes, 2020). In
2020 to July 19, 2020. The surveys were mailed to an equal number of contrast to Viciunaite and Alfnes (2020), we investigate not only more
different BAUR customers (approximately 260,000 for each question sustainable clothing attributes but further use it to measure consumers’
naire), i.e., each customer could only participate in one survey. Among willingness to pay. From a methodological point of view, BWS repre
all participants, we gave away 15 vouchers á 15 Euros for the BAUR sents an alternative to rating, ranking, and to other choice-based data
online shop. collection procedures due to its simplicity of the question format and its
The newsletter containing the first survey was opened by 51,578 parsimony when collecting data (Flynn and Marley, 2014). Both are
customers. In total, 5,149 customers opened the first survey, and 2,244 valid arguments when collecting data in a real-world setting without
responses were considered for further analysis. In line with the target paid online-panel respondents but – nevertheless – the intention to
customer segment of BAUR, our sample consisted of 1,770 (76.7%) fe collect responses from a large and representative sample of customers. It
males. Most respondents were between 50 and 54 years old (n = 423, forces respondents to make trade-off decisions between different alter
18.3%) and 60 years or older (n = 549, 23.8%). The majority of the natives and, thus, prevents respondents from developing and expressing
respondents’ households had a total net income between 1,501 and an inflation of expectations (compared to e.g. Likert-type scales).
2,000 Euros (n = 287, 12.8%). Mostly, respondents associated sustain Further, BWS experiments allow drawing conclusions about the attri
ability with a decreased environmental impact (n = 1,055, 47.0%) and butes’ relative importance and provides information not only about the
durable as well as reparable products (n = 1,054, 47.0%). Regarding best alternative but also about the worst alternative. Moreover, since
their sustainable behavioral patterns in daily life, respondents mainly BWS relies on the collection of real-life forced choices between (mone
indicated avoiding plastic packaging when buying groceries (mean = tary and non-monetary) options (see Louviere et al., 2015), it can be
4.95, standard deviation SD = 1.27) and purchasing environmentally assumed to provide valid willingness to pay measurements for the sus
friendly cleaning products (mean = 4.61, SD = 1.45). The most tainable product attributes.
frequently mentioned barriers towards sustainable consumption Within the first BWS experiment of the first study’s main part, we
behavior were the lack of information (n = 1,248, 55.6%) and high asked the respondents to evaluate both sustainable and conventional
prices (n = 1,139, 50.8%). To draw more attention towards sustainable apparel attributes in the main part of the survey (see Table 2). To cap
products and to communicate sustainability transparently, respondents ture ecological, social, and economic sustainability we incorporated the
demand information about the product’s environmental impact items bio-based materials, fair wages and working conditions, and a
attached directly to the product (mean = 4.70, SD = 1.05) and discounts return program respectively. In total, there were ten items in the study
for sustainable products (mean = 4.65, SD = 1.10). Lastly, respondents and we displayed four items per choice set. We asked as many choice sets
indicated to assess an organization’s sustainability level with regard to per respondent such that each item appeared at least three times per
its manufacturing conditions (n = 1,524, 67.9%) and use of pesticides as respondent. Respondents were asked to choose the most important and
well as chemicals in its products (n = 1,519, 67.7%). The descriptive the least important attribute within each choice set while considering
statistics of Study 1 are depicted in Table A.1 in the Appendix. purchasing clothes.
The newsletter containing Study 2 was opened by 51,668 customers.
In total, 5,065 customers opened the first survey, and 2,106 responses 5.2. Results
were considered for further analysis. The sample comprised 1,678
(79.7%) females. Mostly, respondents were between 50 and 54 years old For analysis, we applied HB estimation and ran 30,000 iterations
(n = 379, 18.0%) and 60 years or older (n = 503, 23.9%). Most re (including 20,000 burn-in iterations). Complementary, we calculated
spondents’ households had a total net income between 3,001 and 4000 average BW scores. Internal consistency is given, considering that the
Euros (n = 241, 11.4%). Respondents mainly associated sustainability model’s root likelihood (RLH) is 0.561, which is higher than the null
with a decreased environmental impact (n = 1,030, 48.9%) and durable model’s RLH of 0.25.
as well as reparable products (n = 1,003, 47.6%). The majority of the Across the whole sample, fit and comfort was chosen the most
respondents indicated avoiding plastic packaging when buying groceries important attribute, just before price-performance ratio. The quality and
(mean = 4.84, SD = 1.33) and purchasing environmentally friendly number of reviews are rather unimportant, as well as the garment’s
cleaning products (mean = 4.56, SD = 1.45) as common sustainable brand. The most important sustainable clothing attribute was social-
behavioral patterns in their daily life. Respondents stated the lack of facetted, i.e., fair wages and working conditions. Nevertheless, it was
information (n = 1,107, 52.6%) and high prices (n = 1,015, 48.2%) to be approximately only half as important as the garment’s fit and comfort.
the main barriers towards sustainable consumption behavior. Further, Bio-based materials and the return program even had a negative average
respondents demand information about the product’s environmental BW score. The results are reported in Table 3.
impact attached directly to the product (mean = 4.62, SD = 1.06) and Overall, consumers apparently rather value intrinsic conventional
6
Table 2 Table 4
Sustainable and conventional apparel attributes used in Study 1a. Gender-specific best-worst scaling results for Study 1a.
Attribute Description References Attribute Zero-centered utility Probability score
Brand The garment’s brand or (Eckman et al., 1990; Female (n Male (n = Female (n Male (n =
reference to manufacturer Jegethesan et al., 2012) = 1,770) 470) = 1,770) 470)
Design The garment’s appearance in (Abraham-Murali and
Fit and comfort*** 3.501 2.570 22.934 21.326
terms of design and style Littrell, 1995; Fiore and
(1.341) (1.288) (4.209) (5.375)
Damhorst, 1992)
Price-performance 2.212 2.186 18.447 19.655
Fit and comfort How well the garment fits and (Eckman et al., 1990;
ratio (1.481) (1.353) (6.762) (6.528)
conforms to the shape of the Zhang et al., 2002)
Quality 1.360 1.230 14.426 14.728
body
(1.570) (1.439) (7.070) (7.143)
Number of The number of reviews written Park et al. (2007)
Design** 0.788 0.521 11.950 11.104
customer by other customers
(1.581) (1.356) (7.118) (6.554)
reviews
Fair wages and 0.511 0.094 10.785 9.583
Price- Subjective evaluation about (Abraham-Murali and
working conditions s (1.873) (1.604) (7.728) (7.457)
performance whether the price is Littrell, 1995; Zhang et al.,
***
ratio appropriate for the perceived 2002)
product performance (value for Bio-based materials s − 0.539 − 1.108 6.841 5.090
money) *** (1.870) (1.634) (6.617) (5.495)
Quality Level of excellence of the (Abraham-Murali and Quality of customer − 1.581 − 1.134 3.852 5.015
construction or material in the Littrell, 1995; Zhang et al., reviews*** (1.687) (1.544) (5.011) (5.477)
garment 2002) Brand*** − 1.652 − 0.828 4.881 6.553
Quality of The quality of reviews written Park et al. (2007) (2.355) (1.904) (6.699) (6.985)
customer by other customers Return program s ** − 1.797 − 1.501 3.977 4.265
reviews (2.097) (1.710) (5.634) (5.541)
Bio-based Use of bio-based materials or (Baier et al., 2020; Fulton Number of customer − 2.803 − 2.030 1.907 2.680
materials s fibres and Lee, 2013) reviews*** (1.825) (1.535) (3.452) (3.659)
Fair wages and Manufactured under humane (Fulton and Lee, 2013;
Note: Standard deviations in parentheses. t-Test conducted with zero-centered
working working conditions; employees Goworek et al., 2012;
utilities. s = sustainable apparel attribute, ***= p < 0.001, **= p < 0.01,
conditions s obtain appropriate wages Stöckigt et al., 2018)
*= p < 0.05.
Return program s Garment can be returned to the Baier et al. (2020)
retailer for disposal to prolong
the life cycle; it will be the opinion of other third parties in terms of number (t = − 8.429, p <
recycled, re-used, or 0.001) and quality of customer reviews (t = − 5.193, p < 0.001)
environmentally friendly
compared to female respondents. Further, they value a return program
disposed
more than female respondents (t = − 2.824, p = 0.005). Apparently,
Note: s
= sustainable apparel attribute. male respondents generally care less about sustainable aspects, partic
ularly when it comes to material usage and manufacturing conditions. In
contrast, female respondents put more emphasis on fit and comfort
Table 3
when purchasing clothes (t = 13.495, p < 0.001), the garment’s design
Best-worst scaling results for Study 1a (n = 2,244).
and appearance (t = 3.352, p = 0.001), bio-based materials (t = 6.008,
Attribute Average BW Zero-centered Probability
p < 0.001), and the social facet of sustainability, i.e., fair wages and
score utility score
working conditions (t = 4.419, p = 0.003).
Fit and comfort 0.621 3.314 (1.365) 22.631 (4.556)
Price-performance ratio 0.408 2.215 (1.458) 18.720 (6.746)
Quality 0.233 1.330 (1.544) 14.485 (7.109)
6. Study 2a: Assessing sustainable apparel attributes
Design 0.114 0.729 (1.536) 11.765 (7.034)
Fair wages and working 0.087 0.426 (1.825) 10.515 (7.716) 6.1. Conceptualization
conditions s
Bio-based materials s − 0.132 − 0.670 (1.850) 6.475 (6.475) Within the first BWS experiment of the second study’s main part, we
Quality of customer − 0.257 − 1.502 (1.665) 4.074 (5.139) asked the respondents to evaluate only sustainable apparel attributes in
reviews
the main part of the survey (see Table 5). We included nine items in the
Brand − 0.262 − 1.460 (2.280) 5.231 (6.821)
Return program s − 0.324 − 1.740 (2.025) 4.038 (5.644) study and displayed four items per choice set. Every item was supposed
Number of customer − 0.487 − 2.642 (1.795) 2.066 (3.535) to appear at least two times per respondent. This can be considered
reviews sufficient, as we expected our sample to be respectively large. For every
choice set, respondents were asked to choose the most important and the
Note: Standard deviations in parentheses, s
= sustainable apparel attribute.
least important attribute when purchasing clothes.
apparel attributes than sustainable ones. Intrinsic attributes (i.e., fit and
comfort, quality, and design) were found to be extremely important as 6.2. Results
well as the value consumers get for their money. Surprisingly, the con
ventional extrinsic attribute (brand) is of inferior relevance and only half Analogously to Study 1a, we applied hierarchical Bayes (HB) esti
as important as fair wages and manufacturing conditions (with respect mation and ran 30,000 iterations (including 20,000 burn-in iterations).
to their probability scores). Further, we calculated the average BW scores. The model’s root likeli
Additionally, we performed a post-hoc Tukey HSD test (on the zero- hood (RLH) is 0.560, which is higher than the null model’s RLH of 0.25,
centered utilities) to test the differences between the attributes. All at and thus, internal consistency is provided.
tributes’ utilities were found to be significantly different from each For our respondents, the garment’s durability, fair wages and
other, except for brand versus quality of customer reviews. manufacturing conditions, a product causing few emissions, and an
To gather further insights, we investigated the gender-specific at environmentally friendly production process are the most important
tributes’ importances (see Table 4). Apparently, male respondents are sustainable apparel attributes. These attributes’ probability scores are
caring more about the garment’s brand (t = − 6.997, p < 0.001), and comparable, see Table 6. Durability can be considered not only a sus
tainable attribute reducing consumption levels of consumers and thus,
7
Table 5 Table 7
Sustainable apparel attributes used in Study 2a. Gender-specific best-worst scaling results for Study 2a.
Attribute Description References Attribute Zero-centered utility Probability score
Bio-based materials Use of bio-based materials or (Baier et al., 2020; Female (n Male (n = Female (n Male (n
fibres Fulton and Lee, 2013) = 1,678) 417) = 1,678) = 417)
Country-of- Country in which product was Fulton and Lee (2013)
Durability** 1.892 2.277 18.257 21.562
manufacture manufactured
(2.336) (1.864) (10.211) (9.366)
Durability Product longevity in terms of (Baier et al., 2020;
Fair wages and working 1.835 1.529 18.808 18.535
robustness, reparability, Goworek et al., 2012)
conditions*** (1.405) (1.163) (7.558) (6.960)
timelessness
Environmentally friendly 1.584 1.231 17.652 16.636
Sustainability label Product’s sustainability is (D’Souza et al., 2007;
production process*** (0.978) (0.853) (5.291) (5.145)
guaranteed with a label Hiller Connell, 2010)
Low-emission product*** 1.724 1.117 17.977 15.819
Environmentally Production process saved Fulton and Lee (2013)
(1.438) (1.228) (7.121) (6.720)
friendly production valuable resources and caused
Recyclable materials 0.085 0.155 9.102 9.590
process few emissions
(1.161) (0.788) (5.789) (4.549)
Fair wages and Manufactured under humane (Fulton and Lee, 2013;
Bio-based materials** − 0.782 − 1.025 5.876 4.151
working conditions working conditions; employees Goworek et al., 2012;
(1.477) (0.795) (5.730) (3.038)
obtain appropriate wages Stöckigt et al., 2018)
Sustainability label** − 1.542 − 1.273 4.241 5.237
Low-emission product Product caused few emissions Fulton and Lee (2013)
(1.646) (1.583) (5.529) (6.326)
throughout the whole supply
Country-of- − 2.285 − 1.931 3.927 4.237
chain
manufacture** (2.219) (1.838) (6.541) (6.595)
Recyclable materials Use of recyclable materials or Fulton and Lee (2013)
Return program** − 2.512 − 2.080 4.161 4.232
fibres
(2.524) (2.020) (7.300) (6.479)
Return program Garment can be returned to the Baier et al. (2020)
retailer for disposal to prolong Note: Standard deviations in parentheses. t-Test conducted with zero-centered
the life cycle; it will be recycled, utilities. ***= p < 0.001, **= p < 0.01, *= p < 0.05.
re-used, or environmentally
friendly disposed
7.924 p < 0.001), and bio-based materials (t = 3.243, p = 0.001).
Similar to the findings of Study 1a, male consumers tend to put more
emphasis on the opinion of third parties as they consider sustainability
Table 6
Best-worst scaling results for Study 2a (n = 2,106).
labels as more important than female consumers (t = − 3.008, p =
0.003), and they rather value return programs (t = − 3.240, p =
Attribute Average BW Zero-centered Probability
0.001). Further, the country-of-manufacture is of higher importance to
score utility score score
them compared to female respondents (t = − 3.013, p = 0.003).
Durability 0.370 1.986 (2.275) 18.871
(10.174)
Fair wages and working 0.359 1.787 (1.378) 18.800 7. Study 1b: Assessing willingness to pay for sustainable apparel
conditions (7.529) attributes
Environmentally friendly 0.298 1.503 (0.952) 17.381
production process (5.240)
7.1. Conceptualization
Low-emission product 0.267 1.598 (1.425) 17.552
(7.163)
Recyclable materials 0.010 0.101 (1.086) 9.183 (5.581) Within the second BWS experiment of the first study’s main part, we
Bio-based materials − 0.131 − 0.831 (1.364) 5.550 (5.318) determined how much consumers are willing to pay more for sustain
Sustainability label − 0.291 − 1.492 (1.662) 4.474 (5.779)
able apparel attributes. In extension to the non-monetary attributes
Country-of-manufacture − 0.406 − 2.219 (2.158) 4.004 (6.598)
Return program − 0.426 − 2.434 (2.440) 4.184 (7.169)
investigated in Study 2a, we added three monetary items – 10%, 15%,
and 20% discount – to answer this question. In the highly competitive
Note: Standard deviations in parentheses. context of fashion online shopping, customers are used to monetary- and
non-monetary product-related offerings. Discounts in this range are
decreasing overall apparel production, but further an economic attribute appropriate and can be used to comparatively measure the subjective
allowing consumers to wear their clothes longer with fewer re value of the sustainable apparel attributes as alternative objects. Table 8
placements and hence, fewer investments. Respondents emphasize the summarizes the list of items used in our BWS study.
social and environmental facets of sustainability, particularly those The usage of these discounts is in line with the approach that Kaas
environmental aspects, which are obviously associated with CO2 emis (1977) introduced to random utility theory and was later further elab
sions. Material usage or the country-of-manufacture, which are only orated by Gaul (1989) and Gierl and Schwanenberg (1997): The idea is
indirectly linked to CO2 emissions (e.g., due to longer transportation to confront respondents with choice tasks that comprise monetary ob
routes), are considered only half as important. Moreover, programs for jects (e.g., certain amounts of money or discounts) as well as
returning used clothes and labels indicating sustainable products are non-monetary objects (e.g., real or hypothetical products) and to use the
rather unimportant. collected choice frequencies to derive mean subjective values of the
Post-hoc Tukey HSD test revealed that all attributes’ zero-centered objects. Then, by comparing the subjective values of monetary and
utilities are significantly different from each other, except for environ non-monetary objects, it is possible to calculate the monetary value of
mentally friendly production process versus low-emission product. the non-monetary objects (via linear transformations). In their experi
With regard to the respondents’ gender, both groups consider the ments with consumer goods, Kaas (1977), Gaul (1989), as well as Gierl
garment’s durability as the most important attribute, and for male re and Schwanenberg (1997) demonstrated that this approach leads to
spondents it is even more decisive for their purchase decision (t = − valid monetary values of the non-monetary objects.
3.130, p = 0.002) (see Table 7). In contrast, the social facet of sus The chosen approach circumvents the traditional critique of choice-
tainability, i.e., fair wages and working conditions, is more important based conjoint analysis where the willingness to pay for attribute-levels
for female respondents than for male respondents (t = 4.119, p < is calculated by comparing partworths for attribute-levels with price
0.001). Furthermore, female consumers put more emphasis on envi coefficients or partworths for price-levels (Allenby et al., 2013, 2014),
ronmental aspects such as an environmentally friendly production since there assumptions of a base product (as a specified
process (t = 6.672, p < 0.001), a product causing few emissions (t = attribute-level-combination) and a competitive environment are made,
8
Table 8 Table 9
Sustainable apparel and price-related attributes used in Study 1b. Best-worst scaling results for Study 1b (n = 2,244) and calculation of monetary
Attribute Description References
values (discount equivalents) for non-monetary attributes.
Attribute Zero-centered Probability Discount
10% discount A price reduction of 10% is (Gaul, 1989; Gierl and
utility score equivalents
offered Schwanenberg, 1997;
Kaas, 1977) 10% discount − 2.246 5.603 (8.647) –
15% discount A price reduction of 15% is (Gaul, 1989; Gierl and (3.563)
offered Schwanenberg, 1997; 15% discount − 1.091 8.184 (9.844) –
Kaas, 1977) (3.953)
20% discount A price reduction of 20% is (Gaul, 1989; Gierl and 20% discount 0.033 (4.830) 11.089 –
offered Schwanenberg, 1997; (10.887)
Kaas, 1977) Durability 2.493 (2.376) 16.038 30.80%
Sustainability label Product’s sustainability is (D’Souza et al., 2007; (7.709)
guaranteed with a label Hiller Connell, 2010) Low-emission product 1.457 (2.049) 12.755 26.25%
Bio-based materials Use of bio-based materials or (Baier et al., 2020; Fulton (6.962)
fibres and Lee, 2013) Environmentally friendly 1.250 (1.779) 11.889 25.34%
Fair wages and Manufactured under (Fulton and Lee, 2013; production process (6.085)
working conditions humane working conditions; Goworek et al., 2012; Fair wages and working 1.132 (2.058) 11.280 24.83%
employees obtain Stöckigt et al., 2018) conditions (6.722)
appropriate wages Recyclable materials − 0.209 6.664 (5.401) 18.94%
Environmentally Production process saved Fulton and Lee (2013) (1.812)
friendly production valuable resources and Sustainability label − 0.513 6.540 (6.510) 17.61%
process caused few emissions (2.464)
Low-emission product Product caused few Fulton and Lee (2013) Bio-based materials − 0.849 5.208 (5.166) 16.13%
emissions throughout the (2.103)
whole supply chain Country-of-manufacture − 1.457 4.748 (6.109) 13.46%
Recyclable materials Use of recyclable materials Fulton and Lee (2013) (2.670)
or fibres
Durability Product longevity in terms of (Baier et al., 2020; Note: Standard deviations in parentheses.
robustness, reparability, Goworek et al., 2012)
timelessness
respondent. However, it has to be mentioned that this discount was not
Country-of- Country in which product Fulton and Lee (2013)
manufacture was manufactured
offered directly to the respondents in the study.
Post-hoc Tukey HSD test showed that all attributes’ zero-centered
utilities were different from each other, except for environmentally
which limits the general validity of the derived willingness to pay and friendly production process versus (1) low-emission product and (2) fair
results in a tendency to overestimate the effects of improved wages and working conditions; recyclable materials versus 20% dis
attribute-levels. In our BWS approach, we collect choices with respect to count; as well as bio-based materials versus 15% discount.
all monetary and non-monetary objects in Table 8 and therefore, do not We gathered similar results as in Study 2a (without monetary attri
assume additive models, competitors, or attribute-level-combinations as butes) when comparing the subsamples of female and male respondents
base products. (see Table 10): Female respondents value every sustainable apparel
Based on the list of attributes (objects) in Table 8, a BWS experiment attribute on average significantly higher than the male respondents
with eight choice tasks for each respondent, four objects in each choice (except for durability), whereas male respondents put more emphasis on
task, and 300 versions of questionnaires was designed. Similar to the the discounts and the garment’s durability. The comparisons were made
paired comparisons experiment in Kaas (1977), Gaul (1989), as well as based on zero-centered utilities, using the Maximum Likelihood across
Gierl and Schwanenberg (1997) three prohibitions were specified: A the (sub)sample and assuming that identical subjective values lead to
maximum of one discount attribute was allowed to appear in one choice similar results. More specifically, the discount equivalents yielded for
tasks. Consequently, it was only possible to develop a slightly unbal female respondents are higher for every attribute than for male
anced and orthogonal choice task design. The monetary (discount) ob respondents.
jects appeared less often (800 times in the 8⋅4⋅300 = 9.600 choice sets)
than the non-monetary objects (900 times) and pairs of monetary objects 8. Study 2b: Assessing sustainable and conventional store
did not appear at all in the choice sets, whereas pairs of monetary and attributes
non-monetary objects appeared slightly more often than pairs of
non-monetary objects. 8.1. Conceptualization
7.2. Results Within the second BWS experiment of the second study’s main part,
we asked the respondents to assess sustainable and conventional online
The choice frequencies were analyzed using HB for estimation. The shop attributes (see Table 11). We included 14 items (i.e., five sustain
mean RLH value of 0.615 across the 2,244 respondents (SD: 0.135) can able and nine conventional) in the study and displayed five items per
be considered good compared to the random RLH value of 0.25. choice set. Every item was supposed to appear at least three times per
The results are summarized in Table 9. First, it is obvious that the respondent. For every choice set, respondents were asked to choose the
ranking of the zero-centered utilities and the probability scores are most important and the least important attribute of an online shop.
similar to the results in Table 6, implying quite valid results of Study 2
(without considering the monetary attributes): The attributes durability, 8.2. Results
low-emission product, environmentally friendly production process, as
well as fair wages and working conditions are by far the highest-ranked We applied hierarchical Bayes (HB) estimation with 30,000 itera
attributes, whereas the attributes country-of-manufacture, bio-based tions (including 20,000 burn-in iterations). Additionally, we calculated
materials, and sustainability label are the lowest-ranked attributes. the average BW scores. Internal consistency is given, considering the
Concerning the monetary attributes, it seems that the 10% discount is model’s root likelihood (RLH) is 0.483, which is higher than the null
ranked rather low on average. The most important sustainable apparel model’s RLH of 0.2. The results of Study 2b are reported in Table 12.
attribute durability is equivalent to a 30.80% discount for the average From an overall perspective, consumers demand less packaging of
9
Table 10
Gender-specific best-worst scaling results for Study 1b and calculation of monetary values (discount equivalents) for non-monetary attributes.
Attribute Zero-centered utility Probability score Discount equivalents
Female (n = 1,770) Male (n = 470) Female (n = 1,770) Male (n = 470) Female (n = 1,770) Male (n = 470)
10% discount** − 2.364 (3.549) − 1.798 (3.593) 5.308 (8.521) 6.744 (9.051) – –
15% discount** − 1.209 (3.938) − 0.643 (3.985) 7.819 (9.593) 9.578 (10.318) – –
20% discount* − 0.081 (4.613) 0.465 (4.681) 10.801 (10.789) 12.166 (11.220) – –
Durability** 2.423 (2.394) 2.747 (2.294) 15.771 (7.780) 16.997 (7.334) 30.97% 30.09%
Low-emission product*** 1.533 (2.030) 1.166 (2.099) 13.038 (6.912) 11.658 (7.035) 27.07% 23.10%
Environmentally friendly production process*** 1.323 (1.758) 0.972 (1.832) 12.172 (6.013) 10.818 (6.249) 26.15% 22.24%
Fair wages and working conditions*** 1.209 (2.050) 0.846 (2.060) 11.502 (6.680) 10.461 (6.807) 25.65% 21.69%
Recyclable materials − 0.192 (1.819) − 0.278 (1.784) 6.711 (5.395) 6.482 (5.427) 19.51% 16.72%
Sustainability label − 0.490 (2.500) − 0.594 (2.332) 6.606 (6.524) 6.312 (6.477) 18.21% 15.32%
Bio-based materials*** − 0.748 (2.089) − 1.236 (2.107) 5.437 (5.268) 4.343 (4.656) 17.08% 12.49%
Country-of-manufacture* − 1.406 (2.675) − 1.648 (2.654) 4.835 (6.123) 4.440 (6.071) 14.19% 10.67%
Note: Standard deviations in parentheses. t-Test conducted with zero-centered utilities. ***= p < 0.001, **= p < 0.01, *= p < 0.05.
Table 11 Table 12
Sustainable and conventional online shop attributes used in Study 2b. Best-worst scaling results for Study 2b (n = 2,106).
Attribute Description References Attribute Average BW Zero-centered Probability
score utility score
Assurance seal Online shop’s security is (Kovar et al., 2000; Odom
verified by an assurance seal et al., 2002) Less packaging s 0.287 1.835 (1.914) 11.793 (6.495)
Availability of Online shop provides customer (Chevalier and Mayzlin, Free returns 0.204 1.309 (2.082) 10.602 (7.080)
customer reviews for its products 2006; Park et al., 2007) Discount campaigns 0.132 0.751 (2.469) 9.324 (7.687)
reviews Sustainable packaging s 0.091 0.789 (1.739) 8.141 (5.802)
Availability of Customer service is easily Cao et al. (2018)
Data security 0.071 0.517 (2.198) 7.999 (6.704)
customer available and online shop has
Free shipping 0.052 0.499 (2.326) 8.486 (7.246)
service sufficient channels for contact
Broad sustainable product 0.051 0.401 (1.972) 7.304 (6.156)
Broad product Online shop offers a broad
range s
range variety of different products
Data security Customers’ data are treated (Forsythe and Shi, 2003; Availability of customer 0.006 − 0.155 (1.860) 6.456 (5.934)
confidentially and are not Ingaldi and Ulewicz, 2019) service
misued or passed to third Assurance seal 0.005 0.019 (2.405) 7.160 (7.036)
parties Broad product range 0.0002 − 0.145 (2.366) 7.009 (6.943)
Discount Online shop offers frequent (Abraham-Murali and Climate-neutral shipping s − 0.056 0.0865 (1.853) 6.117 (5.377)
campaigns discount campaigns Littrell, 1995; Eckman et al., Sustainability label s − 0.114 − 0.636 (2.524) 5.392 (6.139)
1990) Availability of customer − 0.322 − 2.162 (1.967) 2.413 (3.983)
Fast shipping Orders are processed and (Ma, 2017; Stöckigt et al., reviews
shipped quickly 2018) Fast shipping − 0.459 − 3.110 (2.118) 1.804 (3.743)
Free returns Returns are free of charge (Bower and Maxham, 2012;
Cao et al., 2018) Note: Standard deviations in parentheses, = sustainable apparel attribute.
s
Free shipping Shipping is free of charge (Bower and Maxham, 2012;

Lewis, 2006; Ma, 2017;
The post-hoc Tukey HSD test proved that – except for broad sus
Stöckigt et al., 2018)
Broad sustainable Online shop offers a broad (Lundblad and Davies, tainable product range versus (1) data security, (2) free shipping, and (3)
product range s variety of different sustainable 2016; Pookulangara and assurance seal; data security versus (1) free returns and (2) discount
products Shephard, 2013) campaigns; availability of customer service versus (1) assurance seal, (2)
Climate-neutral Online shop offers climate- Stöckigt et al. (2018) broad product range, and (3) climate-neutral shipping; climate-neutral
shipping s neutral shipping
shipping versus (1) assurance seal and (2) broad product range; dis
Less packaging s Online shop reduces packaging (Nguyen et al., 2020;
count campaigns versus (1) free shipping and (2) sustainable packaging
of its orders Schwepker and Cornwell,
1991) – all attributes’ zero-centered utilities are different from each other.
Sustainable Packaging is environmentally (Fulton and Lee, 2013; Comparing both female as well as male respondents (see Table 13),
packaging s friendly Nguyen et al., 2020; we found female respondents to put more emphasis on less (t = 7.647,
Prakash and Pathak, 2017)
p < 0.001) and sustainable (t = 8.159, p < 0.001) packaging, returns
Sustainability Online shop is certified by a
label s sustainability label free of charge (t = 5.091, p < 0.001), data security (t = 3.264, p =
0.001), a broad sustainable product range to choose from (t = 2.723,
Note: s
= sustainable apparel attribute. p = 0.007), and climate-neutral shipping (t = 4.967, p < 0.001). Male
respondents, in turn, are rather interested in the availability of customer
their orders, returns free of charge, and frequent discount campaigns. service (t = − 6.745, p < 0.001), a broad product range to choose from
They also wish for more sustainable packaging, free shipping, and data (t = − 3.588, p < 0.001), the availability of customer reviews (t = −
security. Apparently, consumers mainly care about the ‘how’ (in terms 7.630, p < 0.001), and a fast shipping procedure (t = − 11.910, p <
of packaging) and ‘how much’ (in terms of fees) of their shipping. 0.001).
Nevertheless, consumers care less about climate-neutral shipping (as Overall, all five sustainable online shop attributes were more
this frequently implies additional costs) and overall shipping speed. This important for female respondents than for male respondents (except for
indicates that our respondents are price-sensitive when it comes to the sustainability label). Further, female respondents seem to be more
shipping and return fees. They consider reducing packaging waste and risk-averse regarding their personal data. Again, male consumers were
using sustainable packaging as fruitful possibilities to reduce the envi found to emphasize more the opinion of third parties, i.e., other cus
ronmental impact of online shopping. Further, consumers consider trust tomers, and rather value customer touchpoints, i.e., the availability of
and sustainability labels of online shops as rather unimportant and do customer service, to have the option of contacting in case of complaints
not mind other customers’ reviews. and problems during the shopping process. They seem to be more price-
10
Table 13 and the social facet of sustainability, i.e., fair wages and working con
Gender-specific best-worst scaling results for Study 2b. ditions. Materials were of inferior importance, contradicting the findings
Attribute Zero-centered utility Probability score of Viciunaite and Alfnes (2020). With respect to the respondents’
gender, we found female consumers to consider almost every sustainable
Female (n Male (n = Female (n Male (n =
= 1,678) 417) = 1,678) 417) attribute as more important than male consumers, particularly when it
comes to the social facet of sustainability, i.e., fair wages and working
Less packaging s *** 1.835 1.038 12.193 10.238
(1914) (1.851) (6.346) (6.867)
conditions, as well as bio-based materials, and environmental aspects
Free returns*** 1.309 0.730 10.785 9.715 such as an environmentally friendly production process and a
(2.082) (2.062) (7.023) (7.142) low-emission product. This aligns with prior research indicating that
Discount campaigns 0.751 0.714 9.248 9.506 female consumers rather exhibit environmentally friendly behaviors and
(2.469) (2.281) (7.686) (7.725)
are more likely to engage in sustainable behaviors (Brough et al., 2016;
Sustainable 0.789 0.029 8.561 6.372
packaging s *** (1.739) (1.535) (5.857) (5.132) Lee and Holden, 1999). Female consumers further value fit and comfort
Data security** 0.517 0.131 8.137 7.346 as well as design more than male consumers. In contrast, male re
(2.198) (2.007) (6.644) (6.674) spondents put more emphasis on the garment’s durability and brand, the
Free shipping 0.500 0.598 8.285 9.198 opinion of third parties (i.e., quality and number of customer reviews),
(2.326) (2.185) (7.191) (7.328) signaling (i.e., sustainability label), and return programs.
Broad sustainable 0.401 0.117 7.418 6.810
product range s ** (1.972) (1.604) (6.239) (5.671)
Second, we extend previous research on willingness to pay for sus
Availability of − 0.155 0.511 6.031 8.274
tainable apparel attributes: we found durability as the most important
customer service*** (1.860) (1.538) (5.789) (6.090) sustainable attribute to be equivalent to a 30.80% discount. Regarding
Assurance seal 0.019 − 0.112 7.198 7.032 materials, we found bio-based as well as recyclable materials to be
(2.405) (2.205) (7.065) (6.927) equivalent to a 16.13% and 18.94% discount, respectively. This is
Broad product − 0.145 0.309 6.729 8.191
slightly lower compared to the results of Ellis et al. (2012), who found
range*** (2.366) (2.076) (6.866) (7.170)
Climate-neutral 0.086 − 0.409 6.443 5.184 consumers to pay 25% more for organic cotton. Similar to Hustvedt and
shipping s *** (1.853) (1.675) (5.489) (4.939) Bernard (2010), who indicated that consumers are willing to pay a
Sustainability label s − 0.636 − 0.597 5.428 5.239 premium for labels indicating social responsibility and fair trade, we
(2.524) (1.977) (6.238) (5.663) found fair wages and working conditions to yield among the highest
Availability of − 2.161 − 1.351 2.142 3.405 discount equivalent. Female respondents were again found to value
customer (1.967) (1.818) (3.661) (4.693)
sustainable apparel attributes more, as almost all attributes’ discount
reviews***
Fast shipping*** − 3.110 − 1.708 1.401 3.489 equivalents were higher for female than for male respondents.
(2.118) (2.270) (3.203) (5.166) Third, we provide first insights regarding the importance of both
conventional and sustainable online shop attributes. Extant literature
Note: Standard deviations in parentheses. t-Test conducted with zero-centered
utilities. s = sustainable apparel attribute, ***= p < 0.001, **= p < 0.01, *= focusing on packaging waste and/or sustainable packaging in the e-
p < 0.05. commerce context flourished throughout the past years (Nguyen et al.,
2020; Prakash and Pathak, 2017; Schwepker and Cornwell, 1991),
sensitive regarding shipping fees than female consumers. In turn, male indicating its importance to consumers. Indeed, we found less as well as
respondents care less about return fees. sustainable packaging to be among the four most important online shop
attributes. Besides, we found returns and shipping free of charge to be
9. Discussion emphasized by the respondents, which is in line with preceding findings
(Bower and Maxham, 2012; Smith and Brynjolfsson, 2001; Stöckigt
9.1. Theoretical contribution et al., 2018). Nevertheless, we found shipping speed and climate-neutral
shipping to be less important, contradicting prior findings (Stöckigt
While the importance of single conventional apparel attributes et al., 2018). Female respondents were more interested in almost all
(Abraham-Murali and Littrell, 1995; Eckman et al., 1990) as well as sustainable online shop attributes (i.e., less and sustainable packaging,
online shop attributes (Stöckigt et al., 2018) for consumers has been broad sustainable product range, climate-neutral shipping) compared to
discussed in several studies, research lacks comparing a comprehensive male respondents, highlighting the greenness of female consumers
set of different conventional as well as sustainable apparel and online (Brough et al., 2016; Lee and Holden, 1999).
shop attributes: Extant research on comparing apparel attributes focused
on materials as sustainable attributes only (Viciunaite and Alfnes, 2020). 9.2. Practical implications
Further, literature on willingness to pay for sustainable apparel attri
butes is limited to materials (Ellis et al., 2012; Ha-Brookshire and Our findings imply several practical implications for e-commerce
Norum, 2011; Hustvedt and Bernard, 2008), country-of-manufacture managers, (sustainable) apparel retailers as well as manufacturers, and
(Hustvedt and Bernard, 2008), and labels (Hustvedt and Bernard, further stakeholders.
2010). Thus, our contribution to literature is multifold: Overall, conventional apparel attributes are of striking importance to
First, we are the first study to compare the importance of an exten consumers, particularly fit and comfort, price-performance ratio, qual
sive set of both conventional as well as sustainable apparel attributes ity, and design. Nevertheless, in the context of sustainability, apparel
and, additionally, sustainable attributes only. Overall, we align with retailers should keep the social facet of sustainability (i.e., fair wages
prior research, which compared the importance of sustainable materials and working conditions) in mind. To avoid image damages, manufac
and environmentally friendly production process with conventional at turers are recommended to guarantee fair working standards for their
tributes within a BWS experiment, finding that conventional apparel employees. Further, the garment’s durability is decisive for the con
attributes are of higher importance to consumers (Viciunaite and Alfnes, sumers’ purchase: Consumers value robustness and reparability of gar
2020): We found fit and comfort, price-performance ratio, quality, and ments and further, seek timeless styles. Additionally, it seems of high
design to be the most important attributes when purchasing garments, importance to minimize the environmental impact of the production
which was previously indicated by research on conventional apparel process and the product in general. Retailers and manufacturers are
attributes (Abraham-Murali and Littrell, 1995; Eckman et al., 1990). recommended to disclose this information for a transparent communi
Concerning sustainable apparel attributes, consumers value durability cation with their customers. Material usage, which primarily affects
consumers’ health, is of inferior importance. Apparently, consumers
11
value altruistic (i.e., environmental and social) facets of sustainable 10. Conclusion
clothing more than egoistic facets (i.e., aspects affecting health). Thus,
retailers and e-commerce managers should highlight or even visualize Within this study, we intended to answer the research questions
the consumer’s environmental and social impact when purchasing the regarding (1) the importance of specific sustainable clothing attributes
respective sustainable garment rather than highlighting the consumer’s to consumers (and further, (2) their importance compared to conven
personal benefit. To overcome the gender gap in green consumerism, e- tional clothing attributes), (3) the importance of specific online shop
commerce and retailer managers should consider addressing male con attributes to consumers compared to conventional online shop attri
sumers with male testimonials and ad campaigns targeting men. butes, and (4) consumers’ willingness to pay for specific sustainable
E-commerce managers should further avoid unnecessary packaging clothing attributes.
and use biodegradable or recycled alternatives. Moreover, consumers In terms of the second research question, we found that conventional
were found to be sensitive in terms of return and shipping fees: While apparel attributes (e.g., fit and comfort, price-performance ratio, qual
free returns are more important to women, men are rather interested in ity, and design) are still more important to consumers than sustainable
free shipping. Apparently, men’s intention to return their order when apparel attributes like fair wages and working conditions, bio-based
purchasing apparel products is lower compared to women. The re materials, and a return program. Regarding the first research question,
spondents’ focus on monetary aspects is further reflected in the high the most important sustainable apparel attributes are a garment’s
importance of discount campaigns. durability, fair wages and working conditions, as well as an environ
Policymakers can trigger sustainable clothing consumption among mentally friendly production process, while a return program, the
consumers by raising awareness about the environmental impact of country-of-manufacture, and a sustainability label are of inferior
conventional clothing compared to sustainable clothing with public importance. Moreover, regarding the third research question, consumers
campaigns. Social-facetted advantages of sustainable clothing can be demand less as well as sustainable packaging when ordering online and
depicted with provocative images and documentations. Further, sus free returns. Customer reviews and shipping speed are rather unim
tainable clothing could be subsidized or taxation of sustainable could be portant to consumers. Consumers are willing to pay the highest sur
decreased to stimulate a higher demand among consumers. charge for durability (30.80%), a low-emission product (26.25%), and
an environmentally friendly production process (25.34%) in terms of the
9.3. Limitations and future research fourth research question.
With this micro-perspective on sustainable clothing, we provided
Notwithstanding our theoretical and practical implications, our detailed insights into consumers’ preferences and desires. However,
research is subject to several limitations stimulating further studies. further research is needed as consumers’ demands are constantly
First, albeit our samples are appropriately reflecting the target popula changing, particularly within the fast-moving clothing context.
tion of the online retailer, younger as well as male consumers are rather
underrepresented. Future research could replicate our studies with a Funding
younger sample and a higher number of male respondents.
Furthermore, we did not directly determine consumers’ willingness This research was funded by the Bayerische Staatsministerium für
to pay, but discount equivalents. We did not conduct the assessment Wirtschaft, Landesentwicklung and Energie.
based on a specific apparel product, and discount equivalents may
depend on the product’s base price. Declaration of competing interest
Sustainable product attributes may have been underrepresented in
Study 1a, and hence, future research could conduct a BWS experiment None.
with a balanced number of conventional and sustainable apparel
attributes. Acknowledgements
Aside from gender-specific results, further research could investigate
potential differences arising from different age groups. This research was funded by the Bayerische Staatsministerium für
Moreover, during disruptive situations like the global COVID-19 Wirtschaft, Landesentwicklung and Energie. The authors would like to
pandemic, aspects concerning sustainability may be of inferior impor thank the 34 students (Yvonne Brodtmann, Anna-Maria Christel, Annika
tance to consumers and, from an organizational perspective, aspects Dienst, Lena Grieβhammer, Maximilian Happich, Simon Hardt, Anna-
such as resiliency to ensure an organization’s survival may be more Marie Harnisch, Dara Hentze, Roman Herasymenko, Magdalena Hölzl,
important than sustainability (Dohale et al., 2021). Our findings might Pauline Kaβner, Timo Koch, Hoai-Giang Le-Thi, Liu Siyu, Michael Mal
thus be less generalizable within a disruptive context like the current ter, Alina Meier, Jakob Merschroth, Malin Meyer, Ralph Neubauer,
pandemic. Investigating consumers’ sustainable clothing consumption Xuan-Anh Nguyen, Henrik Peine, Maja Pham, Laura Rödig, Tobias
behavior during disruptive situations can be a fruitful path for future Schichtel, Elena Schiller, Lukas Schmidt, Stefan Schneider, Lisa Marie
research. Steinhaus, Kaj-Johanna Stichnoth, Julia Stumpp, Antonie Teuchert,
Anne-Sophie Thiele, and Amelie Zwickel) for participating in the un
derlying research project.
Appendix. Descriptive statistics

Table A.1
Descriptive statistics of Study 1 (n = 2,244).
Demographics/Characteristics Frequency Proportion (in %)
Gender Female 1,770 78.9

Male 470 20.9
Diverse 4 0.2
Age ≤29 years 109 4.9
30–39 years 299 13.3
(continued on next page)
12
Table A.1 (continued )

40–49 years 501 22.3

50–59 years 786 35.1
≥60 years 549 24.5
Employment Full-time employee 1,006 44.8
Part-time employee 525 23.4
Student 43 1.9
Unemployed 52 2.3
Retired 492 21.9
Others 126 5.6
Household’s total net income ≤2,000 Euros 714 31.9
2,001–3,000 Euros 425 18.9
3,001–4,000 Euros 239 10.7
4,001–5,000 Euros 110 4.9
≥5,001 Euros 64 2.9
No information provided 692 30.8
General perception of sustainability Fair wages and working conditions 584 26.0
No child labor 500 22.3
Decreased environmental impact 1,055 47.0
Buying local 328 14.6
Avoid packaging waste 895 39.9
Recycling 627 27.9
Upcycling 157 7.0
Durable and reparable products 1,054 47.0
Reduce consumption level 289 12.9
Animal welfare 408 18.2
Reduce resource usage 740 33.0
Barriers towards sustainable consumption behavior Lack of trust 515 23.0
Lack of transparency 1,081 48.2
High prices 1,139 50.8
Lack of availability 637 28.4
No clear marking as such 965 43.0
Lack of information 1,248 55.6
I always purchase sustainable products 72 3.2
Other 67 3.0
Assessment of an organization’s sustainability Labels 366 16.3
Number of products with sustainability labels 327 14.6
Return program 970 43.2
Transparency 635 28.3
Carbon footprint 270 12.0
Use of regenerative energy 92 4.1
Pesticide and chemicals usage 1,519 67.7
Sustainable suppliers 556 24.8
Fair manufacturing conditions 1,524 67.9
Image of country-of-manufacturing 123 5.5
Level of social responsibility 330 14.7
Other 20 0.9
Table A.2
Descriptive statistics of Study 1 (n = 2,106).
Gender Female 1,678 79.7

Male 417 19.8
Diverse 11 0.5
Age ≤29 years 119 5.7
30–39 years 241 11.5
40–49 years 488 23.2
50–59 years 755 35.9
≥60 years 503 23.9
Employment Full-time employee 954 45.3
Part-time employee 499 23.7
Student 34 1.6
Unemployed 52 2.5
Retired 443 21.0
Others 124 5.9
Household’s total net income ≤2,000 Euros 615 29.2
2,001–3,000 Euros 419 19.9
3,001–4,000 Euros 241 11.4
4,001–5,000 Euros 106 5.0
≥5,001 Euros 62 2.9
No information provided 663 31.5
General perception of sustainability Fair wages and working conditions 623 29.6
(continued on next page)
13
Table A.2 (continued )

No child labor 475 22.6

Decreased environmental impact 1,030 48.9
Buying local 309 14.7
Avoid packaging waste 772 36.7
Recycling 576 27.4
Upcycling 122 5.8
Durable and reparable products 1,003 47.6
Reduce consumption level 231 11.0
Animal welfare 375 17.8
Reduce resource usage 709 33.7
Barriers towards sustainable consumption behavior Lack of trust 406 19.3
Lack of transparency 919 43.6
High prices 1,015 48.2
Lack of availability 604 28.7
No clear marking as such 863 41.0
Lack of information 1,107 52.6
I always purchase sustainable products 66 3.1
Other 49 2.3
Assessment of an organization’s sustainability Labels 262 12.4
Number of products with sustainability labels 194 9.2
Return program 1,118 53.1
Transparency 524 24.9
Carbon footprint 249 11.8
Use of regenerative energy 83 3.9
Pesticide and chemicals usage 1,385 65.8
Sustainable suppliers 493 23.4
Fair manufacturing conditions 1,532 72.7
Image of country-of-manufacturing 131 6.2
Level of social responsibility 330 15.7
Other 17 0.8
References Culpepper, R.A., Zimmerman, R.A., 2006. Culture-based extreme response bias in
surveys employing variable response items: an investigation of response tendency
among Hispanic-Americans. J. Int. Bus. Res. 5, 75–84.
Abraham-Murali, L., Littrell, M.A., 1995. Consumers’ conceptualization of apparel
Delmas, M.A., Lessem, N., 2017. Eco-premium or eco-penalty? Eco-labels and quality in
attributes. Cloth. Text. Res. J. 13, 65–74.
the organic wine market. Bus. Soc. 56, 318–356.
Allenby, G., Brazell, J., Howell, J., Rossi, P., 2013. Using conjoint analysis to determine
de-Magistris, T., Gracia, A., 2016. Consumers’ willingness-to-pay for sustainable food
the market value of product features. In: Sawtooth Software Conference Proceedings.
products: the case of organically and locally grown almonds in Spain. J. Clean. Prod.
Sawtooth Software Inc., pp. 341–355
118, 97–104.
Allenby, G.M., Brazell, J.D., Howell, J.R., Rossi, P.E., 2014. Economic valuation of
Dohale, V., Ambilkar, P., Gunasekaran, A., Verma, P., 2021. Supply chain risk mitigation
product features. Quant. Market. Econ. 12, 421–456.
strategies during COVID-19: exploratory cases of “make-to-order” handloom saree
Allwood, J.M., Laursen, S.E., Russell, S.N., Rodríguez, C.M. de, Bocken, N., 2008. An
apparel industries. International Journal of Physical Distribution & Logistics
approach to scenario analysis of the sustainability of an industrial sector applied to
Management ahead-of-print.
clothing and textiles in the UK. J. Clean. Prod. 16, 1234–1246.
D’Souza, C., Taghian, M., Lamb, P., Peretiatko, R., 2007. Green decisions: demographics
Auger, P., Devinney, T.M., Louviere, J.J., 2007. Using best–worst scaling methodology to
and consumer understanding of environmental labels. Int. J. Consum. Stud. 31,
investigate consumer ethical beliefs across countries. J. Bus. Ethics 70, 299–326.
371–376.
Baier, D., Rausch, T.M., Wagner, T.F., 2020. The drivers of sustainable apparel and
Dyachenko, T., Naylor, R.W., Allenby, G., 2013. The ballad of best and worst. In:
sportswear consumption: a segmented kano perspective. Sustainability 12, 2788.
Sawtooth Software Conference Proceedings. Sawtooth Software Inc., pp. 357–366
Barbier, E.B., Burgess, J.C., 2020. Sustainability and Development after COVID-19, vol.
Eckman, M., Damhorst, M.L., Kadolph, S.J., 1990. Toward a model of the in-store
135. World development, p. 105082.
purchase decision process: consumer use of criteria for evaluating women’s apparel.
Bower, A.B., Maxham, J.G., 2012. Return shipping policies of online retailers: normative
Cloth. Text. Res. J. 8, 13–22.
assumptions and the long-term consequences of fee and free returns. J. Market. 76,
Eisler, A.D., Eisler, H., 1994. Subjective time scaling: influence of age, gender, and type A
110–124.
and type B behavior. Chronobiologia 21, 185–200.
Brough, A.R., Wilkie, J.E.B., Ma, J., Isaac, M.S., Gal, D., 2016. Is eco-friendly unmanly?
Ellen, P.S., 1994. Do we know what we need to know? Objective and subjective
The green-feminine stereotype and its effect on sustainable consumption. J. Consum.
knowledge effects on pro-ecological behaviors. J. Bus. Res. 30, 43–52.
Res. 43, 567–582.
Ellis, J.L., McCracken, V.A., Skuza, N., 2012. Insights into willingness to pay for organic
Cao, Y., Ajjan, H., Hong, P., 2018. Post-purchase shipping and customer service
cotton apparel. Jnl of Fashion Mrkting and Mgt 16, 290–305.
experiences in online shopping and their impact on customer satisfaction: an
Finn, A., Louviere, J.J., 1992. Determining the appropriate response to evidence of
empirical study with comparison. Asia Pac Jnl of Mrkting & Log 30, 400–416.
public concern: the case of food safety. J. Publ. Pol. Market. 11, 12–25.
Chen, C., Lee, S., Stevenson, H.W., 2016. Response style and cross-cultural comparisons
Fiore, A.M., Damhorst, M.L., 1992. Intrinsic cues as predictors of perceived quality of
of rating scales among east asian and north American students. Psychol. Sci. 6 (3),
apparel. J. Consumer Satisfaction, Dissatisfaction Complain. Behav. 5, 168–178.
170–175. Psychological Science 6, 170–175.
Fisher, R.J., 1993. Social desirability bias and the validity of indirect questioning.
Cheung, K.L., Mayer, S., Simon, J., Vries, H. de, Evers, S.M., Kremer, I.E., Hiligsmann, M.,
J. Consum. Res. 20, 303–315.
2018. Comparison of statistical analysis methods for object case best–worst scaling.
Flynn, T.N., Louviere, J.J., Peters, T.J., Coast, J., 2007. Best–worst scaling: what it can do
J. Med. Econ.
for health care research and how to do it. Journal of Health Economics, 26(1.
Chevalier, J.A., Mayzlin, D., 2006. The effect of word of mouth on sales: online book
J. Health Econ. 26, 171–189, 171–189.
reviews. J. Market. Res. 43, 345–354.
Flynn, T.N., Marley, A., 2014. Best-worst scaling: theory and methods. In: Hess, S.,
Chrzan, K., Golovashkina, N., 2006. An empirical test of six stated importance measures.
Daly, A. (Eds.), Handbook of Choice Modelling. Edward Elgar Publishing, Glos,
Int. J. Mark. Res. 48, 717–740.
pp. 178–202.
Cohen, S., 2003. Maximum difference scaling: improved measures of importance and
Forsythe, S.M., Shi, B., 2003. Consumer patronage and risk perceptions in Internet
preference for segmentation. In: Sawtooth Software Conference Proceedings.
shopping. J. Bus. Res. 56, 867–875.
Sawtooth Software Inc., pp. 61–74
Fulton, K., Lee, S.-E., 2013. Assessing sustainable initiatives of apparel retailers on the
Cohen, S., Orme, B., 2004. What’s your preference? Asking survey respondents about
internet. Jnl of Fashion Mrkting and Mgt 17, 353–366.
their preferences creates new scaling decisions. Marketing Research 16, 32–37.
Gaul, W., 1989. Probabilistic choice behavior models and their combination with
Connolly, J., Prothero, A., 2003. Sustainable consumption: consumption, consumers and
additional tools needed for applications to marketing. In: Klauer, K.C., Soete, G. de,
the commodity discourse. Consum. Mark. Cult. 6, 275–291.
Feger, H. (Eds.), New Developments in Psychological Choice Modeling. North-
Holland, Amsterdam, pp. 317–337.
14
Gierl, H., Schwanenberg, S., 1997. Die Schätzung der Preisresponsefunktion aus Miyazaki, A.D., Fernandez, A.N.A., 2001. Consumer perceptions of privacy and security
Paarvergleichsdaten. J. Market. 36, 3–10. risks for online shopping. J. Consum. Aff. 35, 27–44.
Gil, J.M., Gracia, A., Sánchez, M., 2000. Market segmentation and willingness to pay for Morgan, L.R., Birtwistle, G., 2009. An investigation of young fashion consumers’ disposal
organic products in Spain. Int. Food Agribus. Manag. Rev. 3, 207–226. habits. Int. J. Consum. Stud. 33, 190–198.
Goworek, H., Fisher, T., Cooper, T., Woodward, S., Hiller, A., 2012. The sustainable Mühlbacher, A.C., Kaczynski, A., Zweifel, P., 2013. Experimentelle Präferenzmessung im
clothing market: an evaluation of potential strategies for UK retailers. Intl J of Retail Gesundheitswesen mit Hilfe von Best-Worst Scaling (BWS). PharmacoEcon. Ger. Res.
& Distrib Mgt 40, 935–955. Artic. 11, 101–117.
Ha-Brookshire, J.E., Norum, P.S., 2011. Willingness to pay for socially responsible Mühlbacher, A.C., Kaczynski, A., Zweifel, P., Johnson, F.R., 2016. Experimental
products: case of cotton apparel. J. Consum. Market. 28, 344–353. measurement of preferences in health and healthcare using best-worst scaling: an
Hakovirta, M., Denuwara, N., 2020. How COVID-19 redefines the concept of overview. Health Econ Rev 6, 1–14.
sustainability. In: Sustainability, vol. 12, p. 3727. Nguyen, A.T., Parker, L., Brennan, L., Lockrey, S., 2020. A consumer definition of eco-
Harris, F., Roby, H., Dibb, S., 2016. Sustainable clothing: challenges, barriers and friendly packaging. Journal of Cleaner Production, 252, 119792. J. Clean. Prod. 252,
interventions for encouraging more sustainable consumer behaviour. Int. J. Consum. 119792.
Stud. 40, 309–318. Odom, M.D., Kumar, A., Saunders, L., 2002. Web assurance seals: how and why they
Hiller Connell, K.Y., 2010. Internal and external barriers to eco-conscious apparel influence consumers’ decisions. J. Inf. Syst. 16, 231–250.
acquisition. Int. J. Consum. Stud. 34, 279–286. Oghazi, P., Karlsson, S., Hellström, D., Hjort, K., 2018. Online purchase return policy
Hopfer, E., Istook, C., 2016. The importance of apparel attributes among young Mexican- leniency and purchase decision: mediating role of consumer trust. J. Retailing
American female consumers. Journal of Textile and Apparel, Technology and Consum. Serv. 41, 190–200.
Management 10, 1–15. Paetz, F., Guhl, D., 2017. Understanding differences in segment-specific willingness-to-
Horne, J., Rayner, B., 2013. Does the Analysis of MaxDiff Data Require Separate Scaling pay for the fair trade label. Marketing ZFP 39, 37–46.
Factors? Sawtooth Software Conference Proceedings. Sawtooth Software Inc., Parasuraman, A., Zeithaml, V.A., Berry, L.L., 1985. A conceptual model of service quality
pp. 331–339 and its implications for future research. J. Market. 49, 41–50.
Hustvedt, G., Bernard, J.C., 2008. Consumer willingness to pay for sustainable apparel: Park, D.-H., Lee, J., Han, I., 2007. The effect of on-line consumer reviews on consumer
the influence of labelling for fibre origin and production methods. Int. J. Consum. purchasing intention: the moderating role of involvement. Int. J. Electron. Commer.
Stud. 32, 491–498. 11, 125–148.
Hustvedt, G., Bernard, J.C., 2010. Effects of social responsibility labelling and brand on Paul, J., Modi, A., Patel, J., 2016. Predicting green product consumption using theory of
willingness to pay for apparel. Int. J. Consum. Stud. 34, 619–626. planned behavior and reasoned action. J. Retailing Consum. Serv. 29, 123–134.
Hustvedt, G., Dickson, M.A., 2009. Consumer likelihood of purchasing organic cotton Petersen, J.A., Kumar, V., 2009. Are product returns a necessary evil? Antecedents and
apparel. Jnl of Fashion Mrkting and Mgt 13, 49–65. consequences. J. Market. 73, 35–51.
Ingaldi, M., Ulewicz, R., 2019. How to make E-commerce more successful by use of Pookulangara, S., Shephard, A., 2013. Slow fashion movement: understanding consumer
kano’s model to assess customer satisfaction in terms of sustainable development. perceptions—an exploratory study. J. Retailing Consum. Serv. 20, 200–206.
Sustainability 11, 4830. Prakash, G., Choudhary, S., Kumar, A., Garza-Reyes, J.A., Khan, S.A.R., Panda, T.K.,
Jacoby, J., Berning, C.K., Dietvorst, T.F., 1977. What about disposition? J. Market. 41, 2019. Do altruistic and egoistic values influence consumers’ attitudes and purchase
22–28. intentions towards eco-friendly packaged products? An empirical investigation.
Jegethesan, K., Sneddon, J.N., Soutar, G.N., 2012. Young Australian consumers’ J. Retailing Consum. Serv. 50, 163–169.
preferences for fashion apparel attributes. Jnl of Fashion Mrkting and Mgt 16, Prakash, G., Pathak, P., 2017. Intention to buy eco-friendly packaged products among
275–289. young consumers of India: a study on developing nation. J. Clean. Prod. 141,
Joergens, C., 2006. Ethical fashion: myth or future trend? Jnl of Fashion Mrkting and Mgt 385–393.
10, 360–371. Rausch, T.M., Kopplin, C.S., 2021. Bridge the gap: consumers’ purchase intention and
Kaas, K.P., 1977. Empirische Preisabsatzfunktionen bei Konsumgütern. Springer, Berlin, behavior regarding sustainable clothing. J. Clean. Prod. 278, 1–15.
Heidelberg. Sawtooth Software, 2009a. MaxDiff analysis: simple counting, individual-level logit, and
Kautish, P., Paul, J., Sharma, R., 2019. The moderating influence of environmental HB. https://sawtoothsoftware.com/resources/technical-papers/maxdiff-analysis-s
consciousness and recycling intentions on green purchase behavior. J. Clean. Prod. imple-counting-individual-level-logit-and-hb.
228, 1425–1436. Sawtooth Software, 2009b. The CBC/HB system for hierarchical Bayes estimation version
Kollmuss, A., Agyeman, J., 2002. Mind the Gap: why do people act environmentally and 5.0 technical paper. https://www.sawtoothsoftware.com/support/technical- papers
what are the barriers to pro-environmental behavior? Environ. Educ. Res. 8, /hierarchical-bayes-estimation/cbc-hb-technical-paper-2009.
239–260. Sawtooth Software, 2013. The MaxDiff system – technical paper. https://www.sawtooth
Kovar, S.E., Burke, K.G., Kovar, B.R., 2000. Consumer responses to the CPA software.com/download/techpap/maxdifftech.pdf.
WEBTRUST™ assurance. J. Inf. Syst. 14, 17–35. Scherer, C., Emberger-Klein, A., Menrad, K., 2018. Consumer preferences for outdoor
KPMG, 2019. Sustainable fashion. A survey on global perspectives. https://assets.kpmg sporting equipment made of bio-based plastics: results of a choice-based-conjoint
/content/dam/kpmg/cn/pdf/en/2019/01/sustainable-fashion.pdf. experiment in Germany. J. Clean. Prod. 1085–1094.
Lee, J.A., Holden, S.J.S., 1999. Understanding the determinants of environmentally Schuman, H., Presser, S., 1981. Questions and Answers in Attitude Surveys: Experiments
conscious behavior. Psychol. Market. 16, 373–392. on Question Form, Wording, and Context. Academic Press Inc., Orlando.
Lee, J.A., Soutar, G.N., Louviere, J.J., 2007. Measuring values using best-worst scaling: Schwepker, C.H., Cornwell, T.B., 1991. An examination of ecologically concerned
the LOV example. Psychol. Market. 24, 1043–1058. consumers and their intention to purchase ecologically packaged products. J. Publ.
Lewis, M., 2006. The effect of shipping fees on customer acquisition, customer retention, Pol. Market. 10, 77–101.
and purchase quantities. Journal of Retailing. Journal of Retailing 82, 13–23. Simon, A., 2010. Der Informationsbedarf von Patienten hinsichtlich der
Louviere, J.J., Flynn, T.N., 2010. Using best-worst scaling choice experiments to measure Krankenhausqualität. Gabler Verlag, Wiesbaden.
public perceptions and preferences for healthcare reform in Australia. The patient 3, Smith, M.D., Brynjolfsson, E., 2001. Consumer decision-making at an internet shopbot:
275–283. brand still matters. J. Ind. Econ. 49, 541–558.
Louviere, J.J., Flynn, T.N., Marley, A.A.J., 2015. Best-worst Scaling. Theory, Methods Statista, 2019. Die Top 10 umsatzstärksten Online-Shops im Fashionmarkt in
and Applications. Cambridge University Press. Deutschland 2018. https://de.statista.com/prognosen/646166/top-online-shops-fa
Louviere, J.J., Lings, I., Islam, T., Gudergan, S., Flynn, T., 2013. An introduction to the shion-deutschland-ecommercedb. (Accessed 23 September 2020).
application of (case 1) best–worst scaling in marketing research. Int. J. Res. Market. Statista, 2020. eCommerce - worldwide. https://www.statista.com/outlook/243/100/
30, 292–303. ecommerce/worldwide. (Accessed 15 May 2020).
Louviere, J.J., Woodworth, G.G., 1991. Best-worst Scaling: A Model for the Largest Stöckigt, G., Schiebener, J., Brand, M., 2018. Providing sustainability information in
Difference Judgments (Working Paper). shopping situations contributes to sustainable decision making: an empirical study
Luchs, M.G., Mooradian, T.A., 2012. Sex, personality, and sustainable consumer with choice-based conjoint analyses. J. Retailing Consum. Serv. 43, 188–199.
behaviour: elucidating the gender effect. J. Consum. Pol. 35, 127–144. Taufique, K.M., Vaithianathan, S., 2018. A fresh look at understanding Green consumer
Lundblad, L., Davies, I.A., 2016. The values and motivations behing sustainable fashion behavior among young urban Indian consumers through the lens of Theory of
consumption. J. Consum. Behav. 15, 149–162. Planned Behavior. J. Clean. Prod. 183, 46–55.
Ma, S., 2017. Fast or free shipping options in online and Omni-channel retail? The Thurstone, L.L., 1927. A law of comparative judgment. Psychol. Rev. 34 (4), 273–286.
mediating role of uncertainty on satisfaction and purchase intentions. Int Jrnl Psychological Review 34, 273–286.
Logistics Management 28, 1099–1122. van Loo, E.J., Caputo, V., Nayga, R.M., Seo, H.-S., Zhang, B., Verbeke, W., 2015.
Maichum, K., Parichatnon, S., Peng, K.-C., 2016. Application of the extended theory of Sustainability labels on coffee: consumer preferences, willingness-to-pay and visual
planned behavior model to investigate purchase intention of green products among attention to attributes. Ecol. Econ. 118, 215–225.
Thai consumers. Sustainability 8, 1–20. Viciunaite, V., Alfnes, F., 2020. Informing sustainable business models with a consumer
Marley, A., Louviere, J.J., 2005. Some probabilistic models of best, worst, and best–worst preference perspective. J. Clean. Prod. 242, 1–9.
choices. J. Math. Psychol. 49, 464–480. Wood, S.L., 2001. Remote purchase environments: the influence of return policy leniency
McFadden, D., 1974. Conditional logit analysis of qualitative choice behaviour. In: on two-stage decision processes. J. Market. Res. 38, 157–169.
Zarembka, P. (Ed.), Frontiers in Econometrics. Academic Press, New York, Yadav, R., Pathak, G.S., 2016. Young consumers’ intention towards buying green
pp. 105–142. products in a developing nation: extending the theory of planned behavior. J. Clean.
McKinsey & Company, Global Fashion Agenda, 2020. Fashion on Climate. How the Prod. 135, 732–739.
fashion industry can urgently act to reduce its greenhouse gas emissions. Yadav, R., Pathak, G.S., 2017. Determinants of consumers’ green purchase behavior in a
https://www.globalfashionagenda.com/publications-and-policy/fashion-on-clima developing nation: applying and extending the theory of planned behavior. Ecol.
te/. Econ. 134, 114–122.
15
Young, W., Hwang, K., McDonald, S., Oates, C.J., 2010. Sustainable consumption: green Zhang, Z., Li, Y., Gong, C., Wu, H., 2002. Casual wear product attributes. Jnl of Fashion
consumer behaviour when purchasing products. Sustain. Dev. 18, 20–31. Mrkting and Mgt 6, 53–62.
Zeithaml, V.A., Berry, L.L., Parasuraman, A., 1996. The behavioral consequences of
service quality. J. Market. 60, 31–46.
16

Journal of Retailing and Consumer Services: Theresa Maria Rausch, Daniel Baier, Stefanie Wening

Uploaded by

Copyright:

Available Formats

Journal of Retailing and Consumer Services: Theresa Maria Rausch, Daniel Baier, Stefanie Wening

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Journal of Retailing and Consumer Services: Theresa Maria Rausch, Daniel Baier, Stefanie Wening

Uploaded by

Copyright:

Available Formats

Journal of Retailing and Consumer Services 63 (2021) 102681

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

Does sustainability really matter to consumers? Assessing the importance of

Table 1 Table 1 (continued )

(2016) underpinning Theory higher quality, and

Free shipping Shipping is free of charge (Bower and Maxham, 2012;

Appendix. Descriptive statistics

Demographics/Characteristics Frequency Proportion (in %)

Gender Female 1,770 78.9

Table A.1 (continued )

40–49 years 501 22.3

Demographics/Characteristics Frequency Proportion (in %)

Gender Female 1,678 79.7

Table A.2 (continued )

No child labor 475 22.6

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.