Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6, 2629-2645 2025 Publisher: Learning Gate DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate © 2025 by the authors; licensee Learning Gate History: Received: 11 April 2025; Revised: 13 June 2025; Accepted: 16 June 2025; Published: 30 June 2025 * Correspondence: hafizhekinaufal@student.telkomuniversity.ac.id Influence attitude homophily and social attractiveness to purchase intention with parasocial relationship as an intervening variable in Erigo Hafizh Eki Naufal1*, Citra Kusuma Dewi2 1,2School of Economics and Business, Telkom University, Indonesia; hafizhekinaufal@student.telkomuniversity.ac.id (H.E.N.) citrakusumadewi@telkomuniversity.ac.id (C.K.D.) Abstract: The purpose of this study is to examine the impact of attitude homophily and social attractiveness on purchase intention, with parasocial relationships serving as an intervening variable in the local fashion brand Erigo. This study employs a quantitative approach with an explanatory design to investigate the causal relationships between attitude homophily, social attractiveness, parasocial relationships, and purchase intention among Erigo customers via an online poll of @erigostore Instagram followers. A purposive sampling strategy was used to select 400 respondents based on specific criteria, and data were collected via a Likert scale-based questionnaire. Data were analyzed using the PLS-SEM approach with SmartPLS, which included measurement model testing (validity and reliability) and structural model testing. The results showed that attitude homophily and social attractiveness of influencers significantly influence the formation of parasocial relationships with audiences. Strong parasocial relationships are shown to increase consumers' purchase intentions toward recommended products. Similarity of values and social attractiveness strengthen emotional attachments that drive trust and purchase decisions. The practical implication is that companies need to select influencers who are able to build emotional closeness and have social attractiveness that matches the target audience to optimize digital marketing strategies. Keywords: Attitude homophily, Erigo, Parasocial relationship, Purchase intention, Social attractiveness. 1. Introduction In today's increasingly competitive business era, many new companies have emerged and enlivened market competition. This also happens in the retail business which is growing rapidly along with Indonesia's economic growth. To survive, companies must implement effective marketing strategies to achieve targets and obtain the desired profits [1]. This intense competition means that consumers no longer only play a passive role, but also become more critical decision makers in choosing products among the many choices available [2]. Indonesia is now in the industrial era 4.0, where information spreads quickly, including in the fashion sector, which has experienced rapid progress. Indonesian designers also contribute to creating innovative works that follow global trends, increasing the selling value of fashion products. Fashion is not just clothing, but also a means for individuals to express their self-identity, increasing self-confidence [3]. Industry data indicates that the textile industry's growth has also increased significantly between 2022 and 2023. 2630 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Figure 1. Growth of the Textile and Apparel Industry 2010 – 2023. Based on Figure 1, the growth of the textile and clothing industry in Indonesia shows an increase in 2022 to 2023, which indicates that the fashion business continues to grow. One contributing element to this upward trend is the widespread access to information through digital platforms [4]. Technological advancements, particularly the internet, have become key in driving greater business productivity [5]. In the Indonesian context, internet adoption continues to rise, with user numbers in January 2023 showing a 5.44% increase compared to the previous year. This growth indicates the increasing access of Indonesian people to the internet and digital technology. Business people realize that through the internet and social media they can develop their business which can help market the products they own. Then you can also provide information about the products you own to consumers via the internet and social media and these fashion products dominate sales in e- commerce in Indonesia. This is also supported by the growth and emergence of local fashion products that utilize online media. There are several local Indonesian fashion products that use social media to market their products, including Erigo, Thanksinsomnia, Roughneck 1991, 3Second and so on. One of the local fashion brands that attracts people's attention is Erigo products, where Erigo has a theme street style & traveler, which is currently popular with the public. Table 1. Number of followers Brand Fashion local. No. Nama Brand Instagram account Number of Followers Number of Posts 1 Erigo Apparel @erigostore 2.5 M 7.794 2 Thanksinsomnia @thanksinsomnia 707 K 6.38 3 Roughneck 1991 @roughneck1991 1 M 547 4 3Second @its3second 1.1 M 340 Table 1 shows that Erigo is ranked first among local fashion brands with the number of Instagram followers reaching 2.5 million. This shows the public's interest in Erigo. Erigo, which was founded by 2631 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Muhammad Sadad, a businessman from Aceh, initially appeared in 2010 under the name Selected & Co. At that time, Sadad produced fashion products with a casual batik concept. However, these products are less popular due to rapidly changing fashion trends. In 2015, Sadad decided to rebrand with a focus on style street style and traveling. Since then, Erigo sales have succeeded in dominating the market compared to other local competitors, public interest in brand Local fashion in Indonesia shows that Erigo has a lot of buying interest (purchase intention) which is higher than competitors such as Thanksinsomnia, Roughneck, and 3Second. Several factors that support this include unique and relevant product designs, reflecting casual and youth fashion trends stylish. Erigo also took advantage brand ambassador effective, improving product quality, as well as carrying out innovative promotional strategies through social media and influencer. Compared to its competitors who are more focused on niche In particular, Erigo is able to offer wider appeal with a balance of modern design, quality and competitive prices. Dewi, et al. [6] stated that online businesses can increase purchase intention by reducing the effort required to use the system, as effort expectancy significantly influences online buying intentions. According to Nagori [7] purchase intention involves a comprehensive cognitive and behavioral process in considering or buying a product or service, which includes the willingness to try, purchase, and use it. Many people prefer Erigo over its competitors for several key reasons. Erigo's product design is unique and trends in accordance with the style of today's young people, with a combination of casual but still style stylish. Apart from that, their promotional strategy is through collaboration with influencer popular and brand ambassador which effectively strengthens the brand image in the eyes of consumers. The quality of Erigo products which are considered good and long-lasting also provides added value for consumers who are looking for clothing at affordable prices but still of high quality. Erigo is rated more trends compared to its competitors such as Thanksinsomnia, Roughneck, and 3Second for several key reasons. Erigo focuses on simple but elegant designs, following international fashion trends that are casual, but still look luxurious. This style is very suitable for young urban consumers who want clothes that are comfortable but still fashionable. Meanwhile competitors like Thanksinsomnia tend to focus on design streetwear which is more niche, Roughneck with a more "look"gritty" And sporty, as well as 3Second which is more towards everyday casual wear, they may not be as flexible as Erigo in reaching a wider audience. Erigo has succeeded in balancing a design that is acceptable to many people with an affordable price, and this gives it greater appeal compared to other competitors who are more focused on certain segments. Erigo is also more intensive in digital campaigns, especially through collaboration with influencers and celebrities who increase their visibility significantly. Competitors may not do as much marketing as Erigo, so they lose out in terms of audience reach and popularity. Sales of fashion products in Indonesia fell 10% in the first half of this year due to declining people's purchasing power, competition with imported goods, and the existence of illegal products sold online. The General Chairperson of Hippindo, Budihardjo Iduansjah, said that consumers tend to choose imported goods because the prices are cheaper and the models are more up-to-date compared to local products. Apart from that, delays in stock and the latest models from global brands in the local market are a major challenge. Illegal imported goods also disrupt the competitiveness of local products because they do not pay taxes. Hippindo asks the government to increase supervision and support policies that encourage the domestic retail industry to be more competitive [8]. Apart from producing quality products, Erigo has also succeeded in attracting consumer attention through a digital marketing approach and collaboration with various companies’ influencer famous ones that are socially and culturally relevant. In its marketing strategy, Erigo effectively utilizes roles influencer to build closeness with consumers through digital campaigns, utilizing strong social ties through collaboration with influential figures on social media such as Oza Rangkuti. Influencer such as Oza Rangkuti plays an important role in creating intimate and authentic relationships with his followers, thereby influencing consumer perceptions and purchasing interest. Phenomenon parasocial relationship (PSR) as built by influencer Oza Rangkuti describes the false relationship between consumers and influencer also contribute to influence purchase intention [9]. PSR allows the audience 2632 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate to feel close to influencer without direct interaction, through a feeling of intimacy that builds with repeated exposure. Influencer like Oza Rangkuti is able to build PSR with his followers, creating an emotional closeness that often increases consumer loyalty and encourages them to buy recommended products. This strong presence of PSR shows that effective marketing strategies no longer depend solely on popularity influencer, However, it also depends on how close the audience feels and how emotionally connected they are. According to Masuda, et al. [9] parasocial relationship influenced by many factors, personal attributes, among others attitude homophily and social attractiveness. In the context of modern marketing, personal characteristics such as attitude homophily and social attractiveness are vital in developing relationships between influencers and audiences. Attitude homophily reflects similarities in attitudes or values between influencer and their followers, which can increase the appeal of the message conveyed. On the other hand, social attractiveness, or social attractiveness, also has a significant effect in shaping good attitudes about influencers. The higher the social attractiveness possessed by influencer, Consumers are more likely to be intrigued and persuaded by their recommendations. Oza Rangkuti is influencer who has the ability to attract the interest of his audience through the similar values and lifestyle he displays, which in this research is referred to as attitude homophily. For example, Oza often shares views and opinions that are relevant to the younger generation, such as his views on current issues and contemporary lifestyles, so that his audience feels that their values are aligned. This creates an emotional connection between Oza and his followers, which ultimately influences their perception of the messages or products that Oza recommends. When customers believe that influencers understand their values or viewpoints, they are more inclined to embrace the contents transmitted, including product suggestions. Additionally, social attractiveness (social attractiveness) Oza, which includes personality and the ability to build warm and engaging relationships with audiences, also enhances his effectiveness as influencer. Oza's relaxed but honest delivery style and non-distant attitude make him more easily accepted by the audience. This charisma and social charm helps him build interactions that feel more personal, like a relationship between friends. This appeal makes it easier for his followers to trust his recommendations, thereby encouraging their intention to buy the products he promotes. These two aspects, namely similar attitudes and social attraction, make the relationship formed between Oza and his audience very strong and have great potential in building purchase intentions for the Erigo products he is promoting. This study seeks to measure how attitude homophily and social attractiveness from influencer Oza Rangkuti had an influence on purchase intention Erigo consumers, with PSR as a variable that bridges this influence. Furthermore, this study aims to provide a better understanding of how businesses can use influencer characteristics to optimize the effectiveness of social media marketing campaigns, particularly on platforms that allow for two-way interactions, such as Instagram or YouTube [10]. As the role increases social media influencer as a strategic marketing asset, This study also contributes significantly to our understanding of how PSR influences consumer behavior and improves marketing effectiveness in the context of social media marketing in Indonesia. Therefore, further research will be carried out to see the impact of attitude homophily, social attractiveness, and parasocial relationship to purchase intention on Erigo products, with the title “Influence Attitude Homophily and Social Attractiveness to Purchase Intention with Parasocial Relationship as an Intervening Variable in Erigo.” 2. Literature Review 2.1. Attitude Homophily Attitude homophily is the tendency for people to create relationships with others who share their attitudes, values, and views [11]. Attitude homophily is important in developing emotional ties between consumers and influencers when using social media for commercial communication. Consumers are more likely to connect with and trust influencers who share similar beliefs or perspectives, 2633 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate increasing the impact of marketing communications [12]. Other studies have also found that attitude similarity between endorsers and audiences significantly strengthens psychological bonds, which ultimately influence purchasing behavior [13]. In today’s digital era, attitude homophily has become increasingly important as consumers are more selective in choosing trusted sources of information. Research by Wang, et al. [14] shows that online consumers are more attracted to and loyal to brands or public figures whom they perceive to share similar values and beliefs. This similarity not only reinforces parasocial relationships but also directly impacts the increase in purchase intention. Therefore, in influencer-based marketing strategies, it is crucial for companies to consider the level of attitude homophily between influencers and the target audience to maximize campaign effectiveness. 2.2. Social Attractiveness Social attractiveness is an individual's perceived warmth, friendliness, and ability to build enjoyable relationships [15]. In the context of digital marketing, influencers who demonstrate high levels of social attractiveness are seen as more capable of forming emotional bonds with their audiences, ultimately increasing trust and purchase intention [16]. Influencers who are approachable, communicative, and personable are considered more effective in influencing consumer behavior compared to those who rely solely on physical appearance [17]. Similarly, Lou and Kim [10] discovered that social attractiveness improves parasocial interactions, resulting in increased audience loyalty and purchase intentions for endorsed products. As a result, in the age of social media marketing, social attractiveness has emerged as a significant criterion in identifying influencers who can establish emotional ties and drive customer purchases. 2.3. Purchase Intention Purchase intention is an important indication in understanding consumer behavior since it reflects the possibility that a consumer will buy a product or service in the future. According to Shmueli [18] Theory of Planned action, intention is the primary predictor of actual action, including purchase decisions. Several studies have found that perceived quality, brand trust, price, and promotions all have a major impact on purchase intention Kotler and Keller [19]. Bilal [20] discovered that social media interactions can boost consumer purchase intentions by enhancing brand engagement. Additionally, Chaudhuri [21] emphasized that purchase intention often serves as a strong predictor of actual buying behavior, although external factors such as economic conditions may modify this relationship. Therefore, understanding purchase intention is crucial for companies in designing effective marketing strategies. 2.4. Parasocial Relationship Parasocial relationships (PSRs) are one-sided emotional links developed by individuals (typically audiences) with media figures like as celebrities, influencers, or fictitious characters, in which the audience feels intimate despite the absence of genuine reciprocal engagement [22]. PSRs have grown in popularity in the digital age, as social media sites allow users to feel more connected to public figures. Compared to traditional media like television or radio, modern PSRs are strengthened by interactive features on Instagram, YouTube, and TikTok, creating an even greater illusion of closeness. Bhattacharya [23] found that perceived intimacy in social media-based parasocial relationships can influence consumer behavior and support for influencers, in ways like real-life relationships. Thus, understanding the dynamics of PSRs is crucial not only for advancing communication theory but also for strategic applications in entertainment industries, marketing, and mental health interventions. 2634 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate 3. Research Method 3.1. Research Design This study takes a quantitative method, which is commonly utilized to obtain precise measurements of individuals' behaviors, perceptions, knowledge, or attitudes [24]. Rooted in the philosophy of positivism, this method serves to explain phenomena and test predefined hypotheses [25]. It involves examining a specific population or sample through structured instruments, with the gathered data being analyzed using statistical techniques. 3.2. Population and Sampling Technique In this research, the population in focus is a number of 2,400,000 followers of the Erigo Instagram account. These followers cover various groups, such as individuals who are interested in local fashion products to active consumers who often interact with content or buy Erigo products. The source of information about this population was obtained from the latest data on the number of followers of Erigo's official Instagram account. Therefore, the population that will be the basis for sampling in this research is 2,400,000 followers of the Erigo Instagram account. The Slovin formula was employed in this research to ensure the selected sample accurately represents the overall population. Based on the calculation, the result was 399.93, which was then rounded up to 400 for practical purposes. Therefore, a minimum sample size of 400 respondents was used in this investigation. 3.3. Data Collection Method This research will collect data via a questionnaire with a variety of questions that will be distributed to respondents, namely Erigo consumers who are familiar with the product, are interested in buying it, or who already have one. 3.4. Data Analysis Technique This study used the SmartPLS software to analyze data using the Partial Least Squares Structural Equation Modeling (PLS-SEM) technique. As stated by Widodo and Yusiana [26] PLS is advantageous due to its capability to handle complex model structures, accommodate multiple independent variables, and function effectively regardless of data distribution or sample size constraints. The measurement model (outer model) and the structural model (inner model) were the two main phases of the analytical process. Convergent validity, discriminant validity, and reliability were tested as part of the evaluation of the outer model using metrics including Cronbach's alpha, composite reliability, and Average Variance Extracted (AVE). The R-Square, Q-Square, F-Square, and path coefficients were examined in the inner model assessment to determine the model's capacity for explanation and prediction. Using the bootstrapping technique, the significance of the proposed connections between variables was examined. 3.5. Institutional Review Board Statement This study followed established ethical guidelines for the protection of human participants. Prior to data collection, all participants were given clear information regarding the purpose of the study, procedures, and potential risks, and were given the opportunity to give informed consent. Participants in this study were able to leave at any moment without facing any repercussions because participation was entirely voluntary. The data collected was kept confidential and used only for the purpose of this study. The researcher ensured that only participants who met the predetermined criteria were involved in the study, and the data obtained were verified for eligibility and validity. 4. Results 4.1. Measurement Model Test (Outer Model) In the Structural Equation Modeling–Partial Least Squares (SEM-PLS) approach, the 2635 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate measurement model—also referred to as the outer model—serves to evaluate the validity and reliability of the constructs. The primary goal of this analysis is to confirm that each indicator effectively represents the variable it is intended to measure. Validity and reliability assessments of the indicators were carried out using SmartPLS 4 software to ensure precise computation. This research involves four constructs measured by a total of 16 indicators. Based on the estimation results from the PLS method, a complete model path diagram was produced. In this diagram, the indicators are illustrated with yellow rectangles, while the constructs are marked with blue circles. The numerical values along the arrows indicate the factor loading for each indicator. An indicator is deemed valid if its loading factor exceeds 0.50. Figure 2. Outer Model. 4.1.1. Validity Test Validity assessment ensures that the study instrument accurately measures the target notion. In this work, two types of validity assessments are used: convergent and discriminant validity. Convergent validity is assessed using the Average Variance Extracted (AVE) and factor loadings, which measure how frequently the indicators indicate the same concept. Meanwhile, discriminant validity assesses the degree to which each construct is distinct from others. This is measured using cross loadings, the Fornell-Larcker criterion, and the heterotrait-monotrait (HTMT) ratio. 4.1.2. Convergent Validity Convergent validity assesses how well a construct correlates with theoretically related constructs. It is evaluated using factor loadings and Average Variance Extracted (AVE). An AVE value above 0.50 indicates that over half of the indicator variance is explained by the construct, suggesting good convergent validity. In general, the higher the AVE, the better the indicators represent the same construct. 4.1.2.1. Loading Factor The following table displays the loading factor values for each indication used in this investigation: 2636 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Table 2. Loading Factor Score. Variable Indicator Loading Factor Conclusion AH1 0,815 Valid Attitude Homophily AH2 0,859 Valid AH3 0,847 Valid As shown in the table, all indicators exhibit factor loading values above 0.70. Therefore, they fulfill the validity criteria and are deemed appropriate for inclusion in the analysis without requiring elimination. 4.1.2.2. Average Variance Extracted (AVE) A statistical measure called Average Variance Extracted (AVE) assesses the proportion of variance in observed indicators that can be explained by the underlying latent construct that they reflect. Table 3. AVE value. Variable Average Variance Extracted (AVE) Attitude Homophily 0.706 Social Attractiveness 0.763 Parasocial Relationship 0.763 Purchase Intention 0.744 According to the table provided, all variables in this study have Average Variance Extracted (AVE) values greater than 0.5. This indicates that each variable effectively captures the latent construct it is intended to measure. As a result, all the indicators included are valid and can remain in the analysis without the need for any exclusions. 4.1.3. Discriminant Validity The Heterotrait-Monotrait (HTMT) ratio, cross loadings, and the Fornell-Larcker Criterion are frequently used in SmartPLS to assess discriminant validity. 4.1.3.1. Fornell-Larcker Criterion A construct's ability to explain more variance with its own indicators than with other constructs is evaluated using the Fornell-Larcker Criterion. This is accomplished by contrasting the correlations between constructs with the square root of the AVE. When a construct's AVE is higher than its correlations with other constructs, discriminant validity is proven. Table 4. Fornell-Larcker Criterion Test. Variable Attitude Homophily Purchase Intention Parasocial Relationship Social Attractiveness Attitude Homophily 0.840 Purchase Intention 0.576 0.874 Parasocial Relationship 0.626 0.619 0.874 0.863 Social Attractiveness 0.574 0.543 0.572 0.863 It is evident from the above table that there is a comparatively high connection between the variables and other constructs. Thus, it can be concluded that the Fornell-Larcker criterion has been satisfied. 2637 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate 4.1.3.2. Heterotrait-Monotrait (HTMT) Table 5. Heterotrait-Monotrait (HTMT). Variable Attitude Homophily Purchase Intention Parasocial Relationship Social Attractiveness Attitude Homophily Purchase Intention 0.674 Parasocial Relationship 0.695 0.688 Social Attractiveness 0.680 0.653 0.648 0.653 Referring to the table above, each variable satisfies the HTMT criterion and meets the conditions for discriminant validity since its HTMT values are less than 0.9. 4.1.3.3. Cross Loading The following table displays each indicator's cross loading values. Table 6. Cross loading values of each indicator. Indicator Attitude Homophily Purchase Intention Parasocial Relationship Social Attractiveness AH1 0.815 0.511 0.511 0.459 AH2 0.859 0.452 0.530 0.499 AH3 0.847 0.475 0.513 0.494 AH4 0.839 0.497 0.547 0.476 PI1 0.494 0.848 0.475 0.511 PI2 0.524 0.897 0.608 0.447 PI3 0.490 0.875 0.525 0.476 PR1 0.543 0.533 0.857 0.529 PR2 0.539 0.520 0.881 0.516 PR3 0.521 0.504 0.880 0.474 PR4 0.558 0.568 0.900 0.492 PR5 0.532 0.541 0.859 0.458 PR6 0.582 0.572 0.863 0.522 SA1 0.507 0.475 0.496 0.872 SA2 0.495 0.450 0.486 0.890 SA3 0.482 0.479 0.496 0.825 According to the cross-loading values of the indicators in this study, all indicators show a value greater than 0.700 and exhibit the strongest correlation with their respective latent variables. Therefore, none of the indicators need to be excluded from the analysis. 4.1.4. Reliability Test The reliability test's objective is to assess how well the employed indicators represent the construct being examined. 4.1.4.1. Cronbach’s Alpha The following table displays the Cronbach's Alpha values for each of the study's variables: 2638 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Table 7. Values Cronbach's Alpha Every Variable. Variable Cronbach’s Alpha Attitude Homophily 0.906 Purchase Intention 0.906 Parasocial Relationship 0.951 Social Attractiveness 0.897 The table above indicates that all variables have values exceeding 0.7. This indicates that each measurement of the variables employed in this study has a high degree of dependability. Because of this, every sign can be used in the study and doesn't need to be left out of the analysis. 4.1.4.2. Composite Reliability The table below displays the Composite Reliability values for each variable in this study: Table 8. Composite Reliability Results for Each Variable Variable Composite Reliability Attitude Homophily 0.862 Purchase Intention 0.859 Parasocial Relationship 0.939 Social Attractiveness 0.827 According to the Composite Reliability values, all variables have values greater than 0.700, which suggests that each variable in this study meets the required criteria. Therefore, all indicators are valid for inclusion and do not need to be removed from the analysis. 4.2. Structural Model (Inner Model) Examining the relationships between constructs within a study framework is the main goal of the inner model. It is employed to test theories regarding the links between latent variables and assess their importance. Figure 3. Inner Model. 2639 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate 4.3. Path Coefficient In a research model, the path coefficient is a number that indicates the direction and intensity of the association between latent variables. Table 9. Path Coefficient. Construct Path Coefficient Attitude Homophily -> Parasocial Relationship 0.444 Parasocial Relationship -> Purchase Intention 0.619 Social Attractiveness -> Parasocial Relationship 0.317 Based on the table above, it can be concluded that: 1. Variable Attitude Homophily has a significant positive relationship with the variable Purchase Intention with Path Coefficient of 0.275. 2. Variable Attitude Homophily has a significant positive relationship with the variable Parasocial Relationship with Path Coefficient of 0.444. 3. Variable Parasocial Relationship has a significant positive relationship with the variable Purchase Intention with Path Coefficient of 0.619. 4. Variable Social Attractiveness has a significant positive relationship with the variable Purchase Intention with Path Coefficient of 0.196. 5. Variable Social Attractiveness has a significant positive relationship with the variable Parasocial Relationship with Path Coefficient of 0.317. 4.4. R Square R Square is a measure of how well the independent variable can explain the fluctuations in the dependent variable in a research model [27]. Table 10. Test Results R Square. Variable R Square R Square Adjusted Purchase Intention 0.383 0.382 Parasocial Relationship 0.459 0.456 The table above allows us to derive the following conclusions: 1. At 0.383, or 38.3%, the independent variable influences the Purchase Intention variable. Other factors outside the purview of this study have an impact on the remaining 61.7%. 2. At 0.459, or 45.9%, the independent variable has an impact on the Parasocial Relationship variable. Other factors not included in this study have an impact on the remaining 54.1%. 4.5. F Square The effect size (f²) value in this study is explained in the following table: Table 11. F Square (f²) Test Result. Variable F Square Attitude Homophily -> Parasocial Relationship 0.244 Parasocial Relationship -> Purchase Intention 0.621 Social Attractiveness -> Parasocial Relationship 0.124 In this study, the F Square test yielded the following results: • The F Square value for Attitude Homophily towards Parasocial Relationship is 0.244, which 2640 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate indicates a moderate influence because the f2 value is between 0.15 to 0.35. This implies that Parasocial Relationships are moderately impacted by Attitude Homophily. • The F Square value for the Parasocial Relationship towards Purchase Intention is 0.621, which indicates a significant effect because the f2 value is more than 0.35. This suggests that purchase intention is significantly impacted by parasocial relationships. • The F Square value for Social Attractiveness towards Parasocial Relationship is 0.124, which indicates a small effect because the f2 value is between 0.02 to 0.15. This suggests that parasocial relationships are marginally impacted by social attractiveness. 4.6. Q Square The results of the Q Square test are as follows: Table 12. Q Square test result. Variable Q Square Purchase Intention 0.349 Parasocial Relationship 0.451 Source: Information handled using SmartPLS (2025) The following is evident from the Q Square test results: • The Q Square value of 0.349 for the Purchase Intention variable indicates that it is greater than 0. This implies that the Purchase Intention variable can be explained by the independent variables in this study. • The Q Square value of 0.451 for the Parasocial Relationship variable is likewise higher than 0. This suggests that the parasocial relationship variable in this study can be explained by the independent factors. 4.7. Model Fit The outcomes of the model fit test are as follows: Table 13. Output Model Fit. Indicator Saturated Model Estimated Model SRMR 0.051 0.075 d_ULS 0.349 0.764 d_G 0.228 0.256 Chi-Square 566.855 613.851 NFI 0.876 0.865 According to the above table, the estimated model's SRMR value is 0.075, which is likewise below 0.08, and the saturated model's is 0.051, which is less than 0.08. Consequently, it can be said that the model satisfies the requirements for model adequacy. 4.8. Hypothesis Test The examination of the bootstrapping path coefficients yielded the following findings: 2641 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Table 14. Results of the Direct Influence Hypothesis Test Construct Original Sample (O) T Statistics (|O/STDEV|) P Values Hypothesis Description Attitude Homophily -> Parasocial Relationship 0.444 8.592 0.000 H1 Accepted Social Attractiveness -> Relationship 0.317 5.363 0.000 H2 Accepted Parasocial Relationship-> Purchase Intention 0.619 6.166 0.000 H3 Accepted Based on the data above, we can derive the following conclusions: 1. Homophily has a strong positive impact on parasocial relationships, with an original sample value of 0.444, T Statistics of 8.592 (higher than 1.96), and P Value of 0.000 (less than 0.05). Thus, Hypothesis H1 is supported. 2. Social attractiveness has a strong and favorable impact on parasocial relationships, with an Original Sample value of 0.317, T Statistics of 5.363 (higher than 1.96), and P Value of 0.000 (less than 0.05). As a result, hypothesis H2 has been accepted. 3. Parasocial relationships had a strong positive impact on purchase intention, with an Original Sample value of 0.619, T Statistics of 6.166 (higher than 1.96), and P Value of 0.000 (less than 0.05). Therefore, Hypothesis H3 is validated. Table 15. Results of Indirect Effect Hypothesis. Construct Original Sample (O) T Statistics (|O/STDEV|) P Values Hypothesis Description Attitude Homophily -> Parasocial Relationship -> Purchase Intention 0.196 5.029 0.000 H4 Accepted Social Attractiveness -> Parasocial Relationship -> Purchase Intention 0.275 7.079 0.000 H5 Accepted Source: Data that has been processed with SmartPLS (2025). The table above allows for the following deductions to be made: 1. With a P Value of 0.000 (less than 0.05), a T Statistics of 5.029 (higher than 1.96), and an Original Sample value of 0.196, the Parasocial link mediates the link between Attitude Homophily and Purchase Intention. This suggests that through parasocial relationships, attitude homophily has a large impact on purchase intention, supporting hypothesis H4. 2. With a P Value of 0.000 (less than 0.05), a T Statistics of 7.079 (higher than 1.96), and an Original Sample value of 0.275, the Parasocial Relationship mediates the influence of Social Attractiveness on Purchase Intention. Hypothesis H5 is validated since it implies that social attractiveness has a major influence on purchase intention through parasocial relationships. Based on the results of the hypothesis tests, all hypotheses in this study were supported. Both direct and indirect relationships were found to be significant, suggesting that the mediating variable plays a partial mediating role. This is evidenced by the presence of both a direct relationship between the independent and dependent variables, as well as an indirect relationship mediated by the intervening variable. The significance of these relationships, as indicated by the T-statistics and P-values, confirms that the mediation is partial. 5. Discussion 5.1. Attitude Homophily Influence on Parasocial Relationship The variable Attitude Homophily has a path coefficient of 0.444, which indicates a substantial positive link with Parasocial link, according to the results of the hypothesis test. This implies that the parasocial relationship between customers and influencers gets stronger as the degree of 2642 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate attitude homophily rises. Hypothesis H1 is accepted since the P value of 0.000 is less than 0.05 and the T Statistics value of 8.592 is more than 1.96. Previous studies have highlighted that influencers who establish a strong connection with their audience tend to be more effective in marketing strategies [28]. This result is in line with the phenomenon where consumers tend to feel closer to influencer when they have similar attitudes, values, or thought patterns. This similarity creates a feeling of closeness that strengthens parasocial relationships. Previous research also shows that similarities in attitudes and values between audiences and influencer can strengthen parasocial relationships which influence emotional attachment [16, 29, 30]. 5.2. Social Attractiveness influence on Parasocial Relationship The hypothesis test's findings indicate that Social Attractiveness has a path coefficient of 0.317 with a significant positive relationship to Parasocial Relationship. This means the more socially attractive a person becomes influencer, the higher the level Parasocial Relationship formed with its audience. Mark T Statistics of 5.363 is greater than 1.96 and P Value of 0.000 is smaller than 0.05, so hypothesis H2 is accepted. This is in accordance with the reality that a person has social attractiveness influencer such as communication skills, a pleasant personality, and closeness to the audience can strengthen parasocial relationships. Previous research also supports these findings, where social appeal plays a role in increasing emotional attachment between audiences and influencer [16]. 5.3. Parasocial Relationship influence on Purchase Intention The variable Parasocial link has a path coefficient of 0.619, which indicates a substantial positive link with Purchase Intention, according to the results of the hypothesis test. This implies that consumers are more likely to buy the product that an influencer recommends if they have a greater parasocial relationship with the influencer. The T Statistics value of 6.166 exceeds 1.96, and the P Value of 0.000 is below 0.05, leading to the acceptance of hypothesis H3. acquire intention, according to Rahmi, et al. [31] indicates a customer's intention to acquire a product at a specific moment. Because it comes before real purchase behavior and has a big impact on customer decisions, it is an essential predictor of the effectiveness of marketing strategies [32]. After consumers process information during their purchase decision-making, they develop a path that may influence their considerations, ultimately leading them to make a decision to buy [33]. This result is in accordance with the conditions in which consumers have an emotional attachment to influencer more likely to trust the product recommendations given. This parasocial relationship can enhance positive perceptions of the promoted product, build trust, and ultimately drive purchase intention [34]. Strong parasocial relationships can create feelings of closeness, as it were influencer are their own friends, thus influencing purchasing decisions. Additionally, prior studies indicate that the more strongly an audience and influencer have a parasocial bond, the more probable it is that the audience would heed the influencer's advice on what to buy [10, 35]. 5.4. Attitude Homophily Influence Indirectly through Parasocial Relationship to Purchase Intention With a path coefficient of 0.196, the variable Parasocial Relationship is found to moderate the effect of Attitude Homophily on Purchase Intention, according to the results of the hypothesis test. This link is very positive, suggesting that the parasocial interaction between customers and influencers grows stronger as their attitudes and beliefs become more similar. This, in turn, raises the likelihood that consumers will make a purchase. The acceptance of hypothesis H4 is confirmed by the T Statistics value of 5.029 exceeding 1.96 and the P Value of 0.000 being below 0.05. This finding is in line with the phenomenon where consumers feel they have similar attitudes and values with influencer more likely to trust and feel close to him. This closeness then contributes to shaping purchasing decisions because consumers consider product recommendations from influencer as 2643 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate advice from someone they trust. Previous research also states that Attitude Homophily can strengthen Parasocial Relationship, which ultimately impacts purchasing decisions [36]. 5.5. Social Attractiveness Influence Indirectly through Parasocial Relationship to Purchase Intention The results of the hypothesis test show that, with a path coefficient value of 0.275, Parasocial Relationship likewise mediates the impact of Social Attractiveness on Purchase Intention. This relationship is significantly positive, which means the more socially attractive a person is influencer, the more likely the audience is to feel an emotional attachment to it, which then increases purchasing interest (Purchase Intention). Research indicates that consumers tend to feel more connected and comfortable with influencers who possess strong social attractiveness, such as a pleasant personality, friendly attitude, and effective communication skills [37]. The T Statistics value of 7.079 exceeds 1.96, and the P Value of 0.000 is less than 0.05, confirming the acceptance of hypothesis H5. According to these results, building a parasocial bond with the audience depends heavily on the influencer's social desirability. The more attractive and approachable a person becomes influencer, the more likely the audience is to feel emotionally connected, so they are more open to product recommendations. Previous research also supports this finding, where Social Attractiveness can strengthen Parasocial Relationship, which ultimately influences purchasing decisions [16]. 6. Conclusion According to the findings and analysis conducted in this study using techniques Structural Equation Modeling (WHICH) via software SmartPLS 4.1.0.9, it can be concluded that several main findings answer the research questions as follows: 1. Attitude Homophily significantly impacts Parasocial Relationship. Consumers who perceive a shared alignment in attitudes and values with an influencer are more likely to form strong parasocial bonds. As the similarity between the audience and influencer increases, so does the emotional connection felt by the audience. 2. Social Attractiveness has a significant influence on Parasocial Relationship. Influencer those who have high social attractiveness, such as good communication skills, a friendly personality, and close interaction with the audience, tend to form stronger parasocial relationships. 3. Parasocial Relationship proven to have a positive effect on Purchase Intention. Close parasocial relationships make consumers more confident in product recommendations provided by influencer, thereby increasing the likelihood of purchase. 4. Attitude Homophily influence on Purchase Intention through Parasocial Relationship. This shows that the similarities between the audience and influencer creating parasocial relationships that strengthen purchase intentions. 5. Social Attractiveness influence on Purchase Intention through Parasocial Relationship. The more socially attractive a person becomes influencer, the more likely the audience is to develop a parasocial relationship and trust the product recommendations provided. Overall, this study found that similar attitudes (Attitude Homophily) and social appeal (Social Attractiveness) from influencer play a role in forming Parasocial Relationship, which ultimately has an impact on Purchase Intention. Transparency: The authors confirm that the manuscript is an honest, accurate, and transparent account of the study; that no vital features of the study have been omitted; and that any discrepancies from the study as planned have been explained. This study followed all ethical practices during writing. 2644 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate Copyright: © 2025 by the authors. This open-access article is distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). References [1] S. G. Patianugra and A. Suyanto, "The influence of brand image on purchase decision for Dettol hand sanitizer products during the corona pandemic March 2020 period in Indonesia," Jurnal Manajemen Indonesia, vol. 8, no. 2, pp. 1–23, 2021. [2] A. D. Basallama and M. Ariyanti, "The influence of reference group on purchase decision of Iphone products with lifestyle as a moderating variable (Case Study: Iphone product users in Indonesia)," Jurnal Manajemen Indonesia, vol. 23, no. 1, pp. 73-82, 2023. https://doi.org/10.25124/jmi.v23i1.4178 [3] Y. Chetioui, H. Benlafqih, and H. Lebdaoui, "How fashion influencers contribute to consumers' purchase intention," Journal of Fashion Marketing and Management: An International Journal, vol. 24, no. 3, pp. 361-380, 2020. https://doi.org/10.1108/JFMM-08-2019-0157 [4] L. M. Hong, W. F. W. Zulkiffli, and C. A. C. Amran, "Online feedback impact online shoppers’ impulse purchases in Malaysia," Jurnal Manajemen dan Kewirausahaan, vol. 23, no. 2, pp. 197-203, 2021. https://doi.org/10.9744/jmk.23.2.197-203 [5] R. Hendayani and M. Y. Febrianta, "Technology as a driver to achieve the performance of family businesses supply chain," Journal of Family Business Management, vol. 10, no. 4, pp. 361-371, 2020. https://doi.org/10.1108/JFBM-10- 2019-0070 [6] C. K. Dewi, Z. Mohaidin, and M. A. Murshid, "Determinants of online purchase intention: A PLS-SEM approach: Evidence from Indonesia," Journal of Asia Business Studies, vol. 14, no. 3, pp. 281-306, 2020. https://doi.org/10.1108/JABS-03-2019-0086 [7] A. Nagori, "Impact of influencer marketing on purchase intention with specific reference to health and beauty products," International Journal of Creative Research Thoughts, vol. 8, no. 3, pp. 3157-3170, 2020. [8] Business Universe. (2024) Challenges facing Indonesia’s local fashion industry: Competition, imports, and consumer trends. Business Universe Magazine. [9] H. Masuda, S. H. Han, and J. Lee, "Impacts of influencer attributes on purchase intentions in social media influencer marketing: Mediating roles of characterizations," Technological Forecasting and Social Change, vol. 174, p. 121246, 2022. https://doi.org/10.1016/j.techfore.2021.121246 [10] C. Lou and H. K. Kim, "Fancying the new rich and famous? Explicating the roles of influencer content, credibility, and parental mediation in adolescents’ parasocial relationship, materialism, and purchase intentions," Frontiers in Psychology, vol. 10, p. 2567, 2019. https://doi.org/10.3389/fpsyg.2019.02567 [11] G. Ertug, J. Brennecke, B. Kovács, and T. Zou, "What does homophily do? A review of the consequences of homophily," Academy of Management Annals, vol. 16, no. 1, pp. 38-69, 2022. https://doi.org/10.5465/annals.2020.0230 [12] T. A. Abdulaziz, A. M. E. Mohamed Omar, M. M. Farrag, and W. M. Elwardany, "The attitudes homophily between influencers and customers and the intention to use the brand in social media marketing within the Tourism and Hospitality Industry: The role of trust," The International Journal of Tourism and Hospitality Studies, vol. 6, no. 2, pp. 153-171, 2024. https://doi.org/10.21608/ijthsx.2024.275366.1082 [13] W. Wang, "How has the fashion industry reacted to Gen z’s completely different consumption concepts and shopping habits?," Academic Journal of Humanities & Social Sciences, vol. 4, no. 10, pp. 72-80, 2021. https://doi.org/10.25236/AJHSS.2021.041015 [14] T. Wang, T. D.-H. Thai, P. T. M. Ly, and T. P. Chi, "Turning social endorsement into brand passion," Journal of Business Research, vol. 126, pp. 429-439, 2021. https://doi.org/10.1016/j.jbusres.2021.01.011 [15] K.-P. Wiedmann and L. Mettenheim, "The role of social attractiveness in digital influencer marketing," Journal of Marketing Insights, vol. 15, no. 3, pp. 45–60, 2020. [16] K. Sokolova and H. Kefi, "Instagram and YouTube bloggers promote it, why should I buy? How credibility and parasocial interaction influence purchase intentions," Journal of Retailing and Consumer Services, vol. 53, p. 101742, 2020. https://doi.org/10.1016/j.jretconser.2019.01.011 [17] M. Garg and A. Bakshi, "Exploring the impact of beauty vloggers’ credible attributes, parasocial interaction, and trust on consumer purchase intention in influencer marketing," Humanities and Social Sciences Communications, vol. 11, no. 1, pp. 1-14, 2024. [18] L. Shmueli, "Predicting intention to receive COVID-19 vaccine among the general population using the health belief model and the theory of planned behavior model," BMC Public Health, vol. 21, p. 804, 2021. https://doi.org/10.1186/s12889-021-10816-7 [19] P. Kotler and K. L. Keller, Marketing management, 15th ed. Upper Saddle River, NJ: Pearson Education, Inc, 2016. [20] A. Bilal, "Impact of social media interactions on consumer purchase intentions: The mediating role of brand engagement," Journal of Marketing Research, vol. 58, no. 4, pp. 654–672, 2021. https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.25124/jmi.v23i1.4178 https://doi.org/10.1108/JFMM-08-2019-0157 https://doi.org/10.9744/jmk.23.2.197-203 https://doi.org/10.1108/JFBM-10-2019-0070 https://doi.org/10.1108/JFBM-10-2019-0070 https://doi.org/10.1108/JABS-03-2019-0086 https://doi.org/10.1016/j.techfore.2021.121246 https://doi.org/10.3389/fpsyg.2019.02567 https://doi.org/10.5465/annals.2020.0230 https://doi.org/10.21608/ijthsx.2024.275366.1082 https://doi.org/10.25236/AJHSS.2021.041015 https://doi.org/10.1016/j.jbusres.2021.01.011 https://doi.org/10.1016/j.jretconser.2019.01.011 https://doi.org/10.1186/s12889-021-10816-7 2645 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 6: 2629-2645, 2025 DOI: 10.55214/25768484.v9i6.8458 © 2025 by the authors; licensee Learning Gate [21] A. Chaudhuri, "Purchase intention and actual buying behavior: Effects of external factors," Journal of Consumer Psychology, vol. 31, no. 2, pp. 234–247, 2021. [22] G. Kamenova, "Exploring parasocial relationships formed in isolation," Master’s Thesis. Linnaeus University, Faculty of Arts and Humanities, Department of Cultural Sciences, 2024. [23] A. Bhattacharya, "Parasocial interaction in social media influencer-based marketing: An SEM approach," Journal of Internet Commerce, vol. 22, no. 2, pp. 272-292, 2023. https://doi.org/10.1080/15332861.2022.2049112 [24] P. D. Indrawati, Research methods of management and business convergence of communication and information technology. Bandung: PT Refika Aditama, 2015. [25] Sugiyono, Quantitative research methods. Bandung, Indonesia: Alfabeta, 2022. [26] A. Widodo and R. Yusiana, Research methodology–method determination using partial least square approach–structural equation modeling (Pls-Sem). Bandung: Refika, 2021. [27] J. F. Hair, J. J. Risher, M. Sarstedt, and C. M. Ringle, "When to use and how to report the results of PLS-SEM," European Business Review, vol. 31, no. 1, pp. 2–24, 2019. [28] K. McCormick, "Celebrity endorsements: Influence of a product-endorser match on Millennials attitudes and purchase intentions," Journal of Retailing and Consumer Services, vol. 32, pp. 39-45, 2016. https://doi.org/10.1016/j.jretconser.2016.05.012 [29] F. A. Al-Emadi and I. Ben Yahia, "Ordinary celebrities related criteria to harvest fame and influence on social media," Journal of Research in Interactive Marketing, vol. 14, no. 2, pp. 195-213, 2020. https://doi.org/10.1108/JRIM-02-2018- 0031 [30] R. Ladhari, E. Massa, and H. Skandrani, "YouTube vloggers’ popularity and influence: The roles of homophily, emotional attachment, and expertise," Journal of Retailing and Consumer Services, vol. 54, p. 102027, 2020. [31] Y. Rahmi, L. Sekarasih, and B. Sjabadhyni, "The influence of beauty vlog on perceived source credibility and purchase intention," Makara Human Behavior Studies in Asia, vol. 21, no. 1, pp. 13-23, 2017. https://doi.org/10.7454/mssh.v21i1.3496 [32] N. Peña-García, I. Gil-Saura, A. Rodríguez-Orejuela, and J. R. Siqueira-Junior, "Purchase intention and purchase behavior online: A cross-cultural approach," Heliyon, vol. 6, no. 6, p. e04284, 2020. https://doi.org/10.1016/j.heliyon.2020.e04284 [33] Indrawati, P. C. P. Yones, and S. Muthaiyah, "eWOM via the TikTok application and its influence on the purchase intention of Somethinc products," Asia Pacific Management Review, vol. 28, no. 2, pp. 174–184, 2023. https://doi.org/10.1016/j.apmrv.2022.07.007 [34] G. D. Razan and A. Suyanto, "Analysis of the influence of online customer reviews, influencer reviews and trust on purchase interest in Elsheskin Serum skincare products," eProceedings of Management, vol. 9, no. 4, pp. 1781–1791, 2022. [35] K. Hwang and Q. Zhang, "Influence of parasocial relationship between digital celebrities and their followers on followers’ purchase and electronic word-of-mouth intentions, and persuasion knowledge," Computers in Human Behavior, vol. 87, pp. 155-173, 2018. https://doi.org/10.1016/j.chb.2018.05.029 [36] J. E. Lee and B. Watkins, "YouTube vloggers' influence on consumer luxury brand perceptions and intentions," Journal of Business Research, vol. 69, no. 12, pp. 5753-5760, 2016. https://doi.org/10.1016/j.jbusres.2016.04.171 [37] E. Djafarova and C. Rushworth, "Exploring the credibility of online celebrities' Instagram profiles in influencing the purchase decisions of young female users," Computers in Human Behavior, vol. 68, pp. 1-7, 2017. https://doi.org/10.1016/j.chb.2016.11.009 https://doi.org/10.1080/15332861.2022.2049112 https://doi.org/10.1016/j.jretconser.2016.05.012 https://doi.org/10.1108/JRIM-02-2018-0031 https://doi.org/10.1108/JRIM-02-2018-0031 https://doi.org/10.7454/mssh.v21i1.3496 https://doi.org/10.1016/j.heliyon.2020.e04284 https://doi.org/10.1016/j.apmrv.2022.07.007 https://doi.org/10.1016/j.chb.2018.05.029 https://doi.org/10.1016/j.jbusres.2016.04.171 https://doi.org/10.1016/j.chb.2016.11.009