Incorporating new variables into a model of brand extension in fast fashion

“Incorporating new variables into a model of brand extension in fast fashion” AUTHORS Jitka Burešová Roman Vavrek ARTICLE INFO Jitka Burešová and Roman Vavrek (2024). Incorporating new variables into a model of brand extension in fast fashion. Innovative Marketing , 20(2), 240-253. doi:10.21511/im.20(2).2024.20 DOI http://dx.doi.org/10.21511/im.20(2).2024.20 RELEASED ON Thursday, 13 June 2024 RECEIVED ON Monday, 22 January 2024 ACCEPTED ON Monday, 20 May 2024 LICENSE This work is licensed under a Creative Commons Attribution 4.0 International License JOURNAL "Innovative Marketing " ISSN PRINT 1814-2427 ISSN ONLINE 1816-6326 PUBLISHER LLC “Consulting Publishing Company “Business Perspectives” FOUNDER LLC “Consulting Publishing Company “Business Perspectives” NUMBER OF REFERENCES 66 NUMBER OF FIGURES 1 NUMBER OF TABLES 6 © The author(s) 2024. This publication is an open access article. businessperspectives.org 240 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 Abstract This study tests a brand extension in fast fashion to explore the extension’s effect on the parent brand. It investigates whether extensions to varyingly distant product classes modify customers’ attitudes toward the parent brand. University students from the Technical University of Liberec, the Faculty of Economics (Czech Republic), aged 2225 years, participated in an online survey for this study. The number of respondents was 310. The outcomes are relevant for this segment of customers. The model with classic brand extension factors (perceived fit (FIT), attitudes toward the brand extension (ATE), parent brand attitude change (PBCH)) was constructed. Factors of fashion leaders and emotional variables (e.g., trust and loyalty) were added to the model. The model was tested using structural equation modeling (SEM) in AMOS software and was statistically significant (Chi-squared value of 6.402, p = 0.171). A positive relationship was observed between FIT and ATE (β = 0.534, p-value = 0.000), the same as trust and ATE (β = 0.693, p-value = 0.000). Equally, ATE had a significant positive impact on PBCH (β = 0.722, p-value = 0.000) and trust and loyalty (β = 0.649, p-value = 0.000). Loyalty negatively affects ATE (β = -0.126, p-value = 0.010), indicating that these customers may have problems with brand extension, similar to a fashion leader (β = -0.126, p-value = 0.010). TRUST has a negative effect on the PBCH (β = -0.338, p-value = 0.000). Insights derived from this study hold substantial relevance for marketers in fast fashion aiming to prepare brand extensions effectively. Jitka Burešová (Czech Republic), Roman Vavrek (Czech Republic) Incorporating new variables Incorporating new variables into a model of brand into a model of brand extension in fast fashionextension in fast fashion Received on: 22nd of January, 2024 Accepted on: 20th of May, 2024 Published on: 13th of June, 2024 INTRODUCTION This article focuses on brand extension. However, care must be taken with the extension to avoid affecting the image of the original brand as perceived by customers. In this study, the fashion segment, specifically the fast fashion category, was selected to identify a parent brand. The market is expected to reach USD 39.84 billion by 2025. These numbers show that fast fashion is a large market with the potential to grow. Therefore, it is essential to conduct research in this area. Fast fashion is not a typical category of brand extension research. Brands in this segment do not meet the commonly accepted premise of brand testing, which is that they generally have a good reputation and are perceived positively by customers. Therefore, a brand extension model has not yet been developed or tested for fast fashion, a gap that this study fills. Specifically, it is assumed that the model will show other relationships between latent variables compared with brands that are perceived positively by customers. From this perspective, this study yields new findings regarding the theory of brand extension. To date, the topic of fast fashion has only been covered by Hill and Lee © Jitka Burešová, Roman Vavrek, 2024 Jitka Burešová, Ing., Ph.D., Assistant Professor, Department of Marketing and Trade, Technical University of Liberec, Czech Republic. (Сorresponding author) Roman Vavrek, Ph.D., Assistant Professor, Department of Public Economics, Technical University of Ostrava, Czech Republic. JEL Classification M31, C38 Keywords brand extension, structural equation modeling, fast fashion, parent brand, attitudes toward the brand extension, fit, parent brand attitude change, loyalty, trust, fashion leader LLC “СPС “Business Perspectives” Hryhorii Skovoroda lane, 10, Sumy, 40022, Ukraine This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited. www.businessperspectives.org BUSINESS PERSPECTIVES Conflict of interest statement: Author(s) reported no conflict of interest 241 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 (2015) who measured the level of ecological knowledge and the influence of the perceived importance of ecological issues on attitudes toward extension. A fast fashion brand, H&M, was used to measure brand extension from fast to sustainable fashion. 1. LITERATURE REVIEW AND HYPOTHESES Brand extension refers to the application of an established brand to a new product (Keller & Aaker, 1990). Whether customers positively perceive the brand extension depends on several factors, such as the segment to which the parent brand belongs and how distantly the brand segment extends. How well-known the parent brand is, what emotions it evokes in customers, the customer’s country of origin (Keller & Aaker, 1990; Kim et al., 2003; Pourazad et al., 2019), and the segment in which the brand operates are also important. Therefore, various brand extension models have been created, differing in the combination of various factors and applied in several sectors (e.g., luxury fashion (Pourazad et al., 2019), men’s fashion accessories (Dwivedi et al., 2010), groceries (Joshi & Yadav, 2017), and tourism (del Barrio-García & Prados-Peña, 2019)). No model has yet been created for fast fashion. In the opinion of Lei et al. (2008, p. 113), “perceived fit refers to the degree of proximity between the parent brand and the extension product”. The perceived FIT plays a significant role in brand extension. An important question is how far the brand segment should extend? The key is whether customers would perceive the extension as meaningful and feasible (Keller, 2003). Customers view brand extension positively when they have positive associations with the parent brand and believe that the brand can successfully place a new product in the market (Bridges et al., 2000). However, a negatively perceived FIT can cause undesirable associations with an extended brand because of inconsistencies between the parent and the new brand (Bhat & Reddy, 2001; Nguyen et al., 2018). The perceived FIT mainly influences the variable attitude toward brand extension (Zhu et al., 2023), which represents a customer’s relationship with the extended brand (Bhat & Reddy, 2001; Sattler et al., 2010). Thus, the FIT is a common component of practically all brand extension models. Generally, customers assign the parent brand to a certain product category and have specific associations with it. When a brand is extended to a new product, customers connect these associations to the new product (Bhat & Reddy, 2001). Brand extension always carries the risk that negative associations with the parent brand will be transferred to the new product. Therefore, attitude toward brand extension (ATE) acts as a dependent variable in some attitude-toward-the-brand-extension models, and it is determined by factors influencing the success of the extension (Joshi & Yadav, 2017). However, this does not always concern only the risk of poor perception of the new brand but also a change in the perception of the parent brand. This effect can be understood using schema change theories, which claim that attitudes and beliefs held about the parent brand in memory change in response to brand extensions (Chen & Liu, 2004). Consequently, some models address how brand extension affects parent brands (Dwivedi et al., 2010). Thus, the success or failure of an extended brand is positively or negatively affected by its association with the parent brand (Martínez & de Chernatony, 2004). This is because extending the brand to a new product sends new information about the parent brand to customers (Dens & De Pelsmacker, 2010). When customers’ attitudes toward the extension are positive, the parent brand is also rated positively; conversely, when the attitude toward the extension is negative, the parent brand is also perceived negatively (Dwivedi et al., 2010). Trust is an important attribute in brand theory (Alhaddad, 2015) and plays an important role in the fashion industry (Stathopoulou & Balabanis, 2016). Brand trust is the average customer’s willingness to rely on a brand’s ability to perform its functions (Chaudhuri & Holbrook, 2001). Trust is commonly perceived as a separate factor firmly tied to a brand that affects the value of the brand as perceived by customers (Nadeem et al., 2015; Ruiz-Mafe et al., 2014). Trust reduces the uncertainty associated with buying unknown products. If a product is sold under a well-known 242 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 brand that customers trust, they are likely to trust the product (Agustin & Singh, 2005; Mabkhot et al., 2017). In this case, customers generally rate the brand’s extension to a new product more positively. Chaudhuri and Holbrook (2001) perceived trust as part of the attitude toward a brand, which is measured through trust and brand effect. Wu and Yen (2007) demonstrated that trust and brand affect the attitude toward brand extension. The stronger the effect of trust and brand, the better the perceived brand extension. In this study, only the effect of trust on the (ATE) was determined. When customers encounter an extended brand from a trusted parent brand, they perceive it as the brand’s effort to strengthen its relationship with them (Davis & Halligan, 2002). Dwivedi et al. (2010) stated that trust in the parent brand plays an important role in brand extension and can even influence how customers’ attitudes toward the parent brand change after extension to the new segment. If the extension is rated positively, the attitude toward the parent brand may improve; however, if the extension is rated negatively, the attitude may worsen. In 2013, the same authors identified trust in the parent brand as the strongest factor affecting changes in the attitude toward the parent brand. Therefore, it is favorable for marketers to build trust in the parent brand before extending it to new product lines. If customers trust a parent brand, they will see a new product line in an equally positive light after its extension (Dwivedi & Merrilees, 2013). Trust is an important attribute that influences customer loyalty to a brand (Alhaddad, 2015). Several theoretical models have already emerged that have tested this relationship using SEM (Kudeshia et al., 2016; Nadeem et al., 2015; Ruiz-Mafe et al., 2014). However, it is uncommon to test the relationship between trust and loyalty in a model aimed at extending a brand. This study subsequently determines whether this classic positive linear relationship between trust and loyalty can be confirmed in a model with typical variables for the theory of brand extension. Fashion leaders are categorized into two types of customers: fashion innovators and fashion opinion leaders. Fashion innovators are customers who buy new outfits before others do. Fashion opinion leaders can influence followers’ shopping decisions in the fashion segment. They are fashion enthusiasts aware of the latest fashion trends, which they recommend to others (Martinez & Kim, 2012). These customer categories are often identified using scale items from Goldsmith et al. (1991) for fashion innovators and Flynn et al. (1996) for fashion opinion leaders. However, some authors have found that fashion innovators and opinion leaders are often the same people who share a positive attitude toward fashion, which they understand much more than average customers (Martinez & Kim, 2012; Novotová, 2018). Referring to Novotová (2018), four scale items were selected for this study: two measuring fashion innovators and two measuring fashion opinion leaders, which were jointly used to measure the aggregated fashion leader category. The theory of brand extension mainly involves the category of consumer innovators, that is, innovative customers regardless of the segment they are buying. Rahman et al. (2020) found that consumer innovators generally have a positive attitude toward new products that they want to try immediately, especially if the new product has a modern design. However, these customers are not interested in where the product is made or if it is made by a company that is socially responsible (Rahman et al., 2020). These customers are generally more positive about extending the brand to a new product line because they are not afraid of change (Hirschman, 1980). This assertion was confirmed by Joshi and Yadav (2017), who determined that consumer innovativeness is an important factor that influences attitudes toward brand extension. This study verifies whether such findings also apply to fashion and the customer group referred to as fashion leaders. Yoo and Donthu (2001, p. 8) define loyalty as “the tendency to be loyal to a focal brand, which is demonstrated by the intention to buy the brand as a primary choice.” According to many authors, loyalty consists of two dimensions: attitude loyalty (positive word of mouth – WOM) and behavioral loyalty (repeat purchases) (Chaudhuri & Holbrook, 2001; Gounaris & Stathakopoulos, 2004; Kim et al., 2003). Loyalty, like trust, can be measured using various methods. The composite approach, which combines both dimensions into one latent variable (Bowen & Chen, 2001), was used in this study. 243 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 It can be assumed that loyalty, as an important factor associated with a strong emotional connection to the brand and purchasing decisions, also influences attitudes toward the extension of a popular brand. This assumption was confirmed by Völckner and Sattler (2006) who stated that the (ATE) tends to be better among customers who regularly buy and relate to the brand. Martinelli et al. (2015) stated that brand loyalty affects the purchase of hitherto unknown products of a specific brand. It was confirmed through SEM that the degree of loyalty positively affects the attitudes toward brand extension. In other words, loyal customers have a more positive attitude toward the extension of a brand (Ozretic-Dosen et al., 2018). He et al. (2016) even claim that loyalty to a brand is a basic prerequisite to extending a brand. The presence of customer loyalty indicates a well-built brand with high value (brand equity) and implies that the brand can be extended. Loyalty refers to customers’ emotional connection to a brand. A high level of positive emotion evokes a sense of brand belonging among customers, thus cultivating loyalty among customers. Consequently, they ignore other brands and focus solely on the selected brand (Loureiro & Kaufmann, 2012). Pourazad et al. (2019) stated that positive emotions toward the parent brand also affect the frequency of buying extended brands. This means that emotions that promote customer loyalty can influence customer opinions of the products to which the brand has been extended. However, the attitude toward the parent brand should be maintained given the strong emotional connections. By contrast, if loyalty to the parent brand is low, customers’ attitudes and intentions in relation to the original brand are not necessarily transferred to the extended brand (Hesse et al., 2022). This variable was chosen for the model as a dependent variable because of the need to measure any changes in customers’ perception of the parent brand (Hem et al., 2014; Ramanathan & Velayudhan, 2015). Dwivedi et al. (2010) showed that extending the parent brand to new products changes the perceived value of the parent brand. This effect is explained by Keller (2003) as follows: When consumers are exposed to a brand extension, relevant memory nodes are activated, and various parent brand associations, beliefs, and attitudes are retrieved in memory. Therefore, extending a brand to a new product clearly influences customer perception of the parent brand. For dominant brands, defined as brands that are major players in a particular segment and are generally well-known, a horizontal extension to a new brand does not lead to brand dilution. The dilution of the parent brand refers to its devaluation due to a negatively perceived extension to a new product. This is due to their strong position in the segment, familiarity, and embedment in customers’ minds. Therefore, these brands can be extended without major problems (Leong, 1997). A dominant brand was selected for this study; hence, it can be assumed that the brand will not suffer by extending it to new products. Leong (1997) found that the distance of extension does not affect a dominant brand. However, Ping and Lei (2010) claim that extending a brand to a segment that is too close can dilute it. This is because of the excessive similarity between the new and existing products such that customers do not understand why the new product was launched in the market in the first place. Extensions to a closely related product (a towel) and a distantly related product (a sleeping bag) were selected for this study. The towel extension was deemed as not too close. The literature review shows that brand extension is an important topic. Determining which factors influence ATE and PBCH for fast fashion is necessary. As part of the literature review, perceived fit, trust, loyalty, and fashion leader were identified as the most important factors. The study aims to ascertain whether brands in fast fashion can extend their product lines with new products under the parent brand. The proposed hypotheses, in accordance with the conceptual framework depicted in Figure 1, are as follows: H1: Perceived fit has a significant positive impact on attitude toward brand extension. H2: Attitude toward brand extension positively affects parent brand attitude change. H3: Parent brand trust positively affects attitude toward brand extension. 244 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 H4: Parent brand trust positively affects parent brand attitude change. H5: Parent brand trust positively affects brand loyalty. H6: Fashion leaders positively affect (customer) attitude toward brand extension. H7: Loyalty positively affects attitude toward brand extension. H8: Loyalty positively affects parent brand attitude change. H9: Extending the H&M brand to new products will not change customers’ attitudes toward the parent brand. H10: Different distances between the extended products and the parent brand, H&M, do not result in varying effects on customers’ attitudes toward the parent brand. 2. METHODOLOGY In brand extension research, the assumption is that research will be conducted on brands that respondents know. Therefore, the most important brands in a given segment, the so-called dominant brands, are usually selected (Aaker, 2003). The H&M brand was chosen for this study because it ranks among the best-known fast fashion brands worldwide and because, according to Novotová (2018), it is the most purchased fast fashion brand in the Czech Republic. This brand is present in most shopping centers in the Czech Republic and is also available on the Internet. This real brand was hypothetically extended horizontally into two new product classes for research purposes: one product class could be described as closely extended, and the other distantly extended. A similar approach was taken by Pourazad et al. (2019) and Martínez et al. (2009), where one group of respondents was presented with a questionnaire concerning a closely extended product class and another concerning a distantly extended product class. This means that the items in the questionnaire (using a seven-point Likert scale ranging from strongly agree (1) to strongly disagree (7)) for the FIT variable were evenly represented because one group evaluated FIT as feasible for a given brand, and the other group evaluated it as less feasible (Prados-Peňa & Del Barrio-Garcia, 2020). Implementing this approach yielded a normal distribution. A towel was chosen as the closely extended product class (FIT: mean = 3.94, std. deviation = 1.71), and a sleeping bag was selected as the Note: FIT – Perceived fit, ATE – Attitude towards the brand extension, TRUST – Trust in the parent brand, PBCH – Parent brand attitude change, FASHION – Fashion leaders, LOYALTY – Loyalty to the parent brand. Figure 1. Tested model 245 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 distantly extended product class (FIT: mean = 5.9, std. deviation = 1.56). An unpaired t-test was conducted to verify the selection of product classes that are differently spaced from the parent brand. The unpaired t-test showed statistically different means (t = 11.51, p < 0.001); therefore, the product classes – towels and sleeping bags – were considered differently spaced from the parent brand. This study used a product-to-product extension, which is usually better perceived by customers than a product-to-service extension (Sarasvuo et al., 2023; Rahman et al., 2020; Ramanathan et al., 2015). Both extended products are made of fabric and can thus be feasibly produced by H&M. The most common customers of the H&M brand are young people aged below 24 years, students, and women (Burešová, 2018; Joung, 2014). Therefore, students aged below 24 years from the Faculty of Economics at the Technical University of Liberec were selected for this study. The sample comprised 70% women and 30% men. Approximately 8% of them bought from the H&M brand at least once a month, 59% several times a year, and 32% less often. Respondents who did not buy the H&M brand at all (1%) were excluded from this study. The sample size was determined considering the issues of too small and too large a set when using the SEM statistics. A very small sample presents an increased risk that the statistical tests will be weak. By contrast, in a large sample, a significant value for the test criterion usually emerges (p < 0.05) (Greenland et al., 2016). It is generally recommended that researchers recruit ten samples per path. The analytic model contains 23 items and 8 paths among latent factors. Therefore, at least 310 participants were required to fit the model. Therefore, the sample size was set to 310 respondents. Respondents were divided into two groups: 160 received a questionnaire concerning the closely extended new product (towel), and 150 received a questionnaire concerning the distantly extended new product (sleeping bag). Data were collected using an online questionnaire from Google in October 2022. Although the sample consisted of respondents from only one European country, the results are nonetheless deemed applicable to customers aged 22-25 from other European countries and the United States. Based on the results of a study by Holden and Barwise (1995) who stated that European and US customers perceive brand extension with no major differences, it is possible to generalize research results from one country to another. Table 1. List of items used Construct Items (measured on a 7-point Likert scale) Authors Cronbach’s alpha (CR) Perceived fit (FIT) The towel/sleeping bag extension by H&M makes sense Dwivedi et al. (2010) 0.859 H&M brand has the skills to launch the brand extension I find H&M’s decision to launch a towel/sleeping bag logical The launch of towels/sleeping bags by H&M in the market was expected Attitude-toward-thebrand extension (ATE) My attitude toward H&M towels/sleeping bags is very positive Dwivedi et al. (2010) 0.897 I am very favorably disposed toward H&M towels/sleeping bags I find that H&M towels/sleeping bags are great I admire H&M towels/sleeping bags a lot Parent brand attitude change (PBCH) My attitude toward H&M would become… Dwivedi et al. (2010) 0.884 My disposition toward H&M and its products would become… My admiration toward H&M would become… My opinion regarding H&M as having a great reputation would become… Trust H&M never disappoints me Munuera-Aleman et al. (2003) 0.959 H&M guarantees satisfaction I could rely on the H&M brand to solve the problem H&M will compensate me if problems arise Fashion leader (Fashion) I know new fashion trends before other people do Novotová (2018) 0.786 I often persuade other people to buy the fashion products that I like If I heard that a new H&M collection was available in the store, I would be interested enough to buy it I often influence people’s opinions about clothing Loyalty I consider myself to be loyal to the H&M brand Martínez et al. (2009) 0.862I typically choose the H&M brand at the store I recommend the H&M brand to other consumers 246 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 The individual scale items used in the model are presented in Table 1. Each latent variable was measured using at least three items from previous studies. Fashion leader is a specific latent variable that does not normally appear in brand extension models. The term fashion leader is a collective term for two types of customers: fashion innovators (Goldsmith & Hofacker, 1991) and fashion opinion leaders (Flynn et al., 1996). In this study, both types of customers were measured together using four items from Novotová (2018). The measures for all the constructs were based on a seven-point Likert scale. Responses ranged from strongly agree (1) to strongly disagree (7). The items used to measure the change in attitude toward the parent brand as a consequence of brand extension were anchored as a more strongly negative attitude to H&M as before (-3) and a more strongly positive attitude toward H&M as before (+3), with a mid-point of the same attitude toward H&M as before (0) (Dwivedi et al., 2010). The answers were transcribed from negative and positive values to the classic scale (-3 = 7, +3 = 1) to allow the use of SEM. 3. RESULTS Missing data were excluded from the data matrix. This was achieved using a Google questionnaire containing only the mandatory items. Respondents had to fill in all the items to complete the questionnaire. Significant deviations from the norm were found for some scale items, and some skewness and sharpness values were out of the ±1 range (Ranaweera & Prabhu, 2003). Nevertheless, the maximum likelihood test method generally yields good results even when using data that do not correspond to a normal distribution (Marsh et al., 2004). Therefore, the normality of the data was not addressed. Given that common method bias exists when a measurement technique introduces systematic variance, a careful questionnaire design was implemented prior to data collection to reduce common method bias. Besides, a single-factor post hoc test for common method variance suggested no “same-source” factors in the data (Podsakoff et al., 2003). Thus, common method bias was not considered a threat. Confirmatory factor analysis (CFA) was performed using the AMOS 24 software. All constructs presented in Table 1 were entered into the analysis and showed good reliability – Cronbach’s alpha (CR); that is, all items measured the latent variable well. The problem with some latent variables, such as loyalty and trust, is that the individual scale items are somewhat similar. This is an obstacle to CFA, where some latent variables are too interdependent, and the scale items of these latent variables may be correlated, thus making the model not statistically significant. Therefore, several scale items were excluded from the model. According to Škapa (2012), the latent variable must be measured using at least two scale items; hence, at least two items were always preserved in the model. The scale items preserved in the model, their factor loadings (CFA), means, and standard deviations are shown in Table 2. After these modifications, the entire model was tested using commonly used indices: the Chisquared/degrees of freedom ratio (ꭓ2/d. f.), goodness-of-fit index (GFI), comparative fit index (CFI), root means square error of approximation (RMSEA), normal Chi-squared (CMIN/df), adjusted goodness of fit index (AGFI), and p close (Casaló et al., 2010). Owing to the smaller sample size, the chi-square test was insignificant (64.052, p = 0.073), CFI (0.983); GFI (0.942); RMSEA (0.046); CMIN/df (1.307); AGFI (0.892); and p close (0.570). Therefore, the construct measures were deemed valid. Furthermore, it was necessary to verify the reliability and validity of the model. The reliability and validity of the resulting model were determined after performing CFA. Reliability was measured using Cronbach’s alpha (CR), and the resulting value must be higher than 0.7 for each construct (Cronbach, 1951). Table 4 shows that the TRUST construct marginally meets the value of 0.7 while all other values are above 0.8, and are, therefore, in order. Convergent validity was measured using Cronbach’s alpha (CR) and average variance explained (AVE), provided that the conditions CR > 0.7, CR > AVE, and AVE > 0.5 are met (Hair, 2010). The TRUST construct has a limited value, whereas the other values are very good. Discriminatory validity was assessed using AVE and maximum shared variance (MSV) for which an MSV < AVE relationship should exist 247 Innovative Marketing, Volume 20, Issue 2, 2024 http://dx.doi.org/10.21511/im.20(2).2024.20 and using the Fornell and Larcker (1981) test for which the square root of AVE for a given construct should be greater than the absolute value of the standardized correlation of a given construct with all other constructs. This was also observed in the present study. Table 3 presents the inter-construct correlations and square roots of the AVE for each construct (in bold). The discriminatory validity was ordered according to the results. Then covariance-based SEM to test the hypothesized relationships were deployed and yielded a Chi-squared value of 6.402, p = 0.171; CFI (0.992); GFI (0.986); RMSEA (0.064); CMIN/df (1.601); AGFI (0.926); and p close (0.326). All results met the requirements for structural modelling (Casaló et al., 2010), thus indicating an acceptable model fit. Regarding hypotheses testing (Table 4), the empirical analysis returned the following outcomes. Path analysis standardized coefficients and t-statistics illustrated that FIT (β = 0.534, CR = 11.114, p-value = 0.000) had a significant positive impact on ATE, thus confirming H1. ATE (β = 0.722, CR Table 2. Items, factor loadings (CFA), mean, and standard deviation Construct Items (measured on a seven-point Likert scale) Factor loading Mean Std. deviation FIT I find H&M’s decision to launch towels/sleeping bags logical 0.975 4.97 1.81 The launch of towels/sleeping bags by H&M in the market was expected 0.854 ATE I find that H&M towels/sleeping bags are great 0.885 5.31 1.53 I admire H&M towels/sleeping bags a lot 0.801 PBCH My attitude toward H&M would become … 0.932 4.08 1.47 My opinion regarding H&M as having a great reputation would become… 0.785 Trust H&M guarantees satisfaction 0.914 4.18 1.27I could rely on the H&M brand to solve the problem 0.641 H&M will compensate me if problems arise 0.505 Fashion I often persuade other people to buy the fashion products that I like 0.809 4.98 1.77 I often influence people’s opinions about clothing 0.835 Loyalty I consider myself to be loyal to the H&M brand 0.961 4.75 1.78 I recommend the H&M brand to other consumers 0.785 Note: FIT – Perceived fit, ATE – Attitude towards the brand extension, TRUST – Trust in the parent brand, PBCH – Parent brand attitude change, FASHION – Fashion leaders, LOYALTY – Loyalty to the parent brand. Table 3. Reliability and validity CR AVE MSV MaxR(H) TRUST Fashion LOYAL FIT ATE PBCH TRUST 0.739 0.500 0.355 0.859 0.707 Fashion 0.807 0.676 0.030 0.808 0.172 0.822 LOYAL 0.869 0.770 0.355 0.932 0.596 0.170 0.877 FIT 0.913 0.840 0.310 0.956 0.080 –0.044 –0.048 0.916 ATE 0.832 0.712 0.310 0.844 0.514 –0.055 0.154 0.557 0.844 PBCH 0.851 0.742 0.235 0.892 0.062 –0.011 –0.019 0.373 0.485 0.862 Note: The square root of the AVE of the constructs is in bold along the diagonal. FIT – Perceived fit, ATE – Attitude towards the brand extension, TRUST – Trust into the parent brand, PBCH – Parent brand attitude change, FASHION – Fashion leaders, LOYALTY – Loyalty to the parent brand. Table 4. Results of testing the hypotheses Hypothesis Latent variable path Standardized estimates Critical ratios Sig. level hypotheses support H1 FIT →ATE 0.534 11.114 0.000 Supported H2 ATE → PBCH 0.722 8.988 0.000 Supported H3 TRUST → ATE 0.693 10.876 0.000 Supported H4 TRUST → PBCH –0.338 –4.210 0.000 Not supported H5 TRUST→ LOYALTY 0.649 10.385 0.000 Supported H6 FASHION → ATE –0.126 –2.577 0.010 Not supported H7 LOYALTY → ATE –0.223 –3.550 0.000 Not supported H8 LOYALTY → PBCH 0.137 1.572 0.116 Not supported Note: FIT – Perceived fit, ATE – Attitude towards the brand extension, TRUST – Trust into the parent brand, PBCH – Parent brand attitude change, FASHION – Fashion leaders, LOYALTY – Loyalty to the parent brand.