Impact of Online Crisis Response Strategies on Online Purchase Intention: The Roles of Online Brand Attitude and Brand Perceived Usefulness

The present study has concentrated on online crisis response strategies that stimulate online purchase and the acceptance of online purchase intention (OPI) in the marketplace. However, few studies have examined the online crisis response strategies undertaken to alleviate the online crisis. Therefore, on the basis of the Stimulus Organism Response framework and online crisis response strategies, this study offered and tested a model that inspected the effect of online crisis response strategies on OPI through online brand attitude (OBA). Brand perceived usefulness (BPU) was observed as a borderline condition between OBA and OPI. We recruited 392 working professionals undertaking the Master of Business Administration program at a large public university in China. The results suggest that OBA mediates the association between online crisis response strategies and OPI. Similarly, BPU moderates the association between OBA and OPI. The responses with high BPU maximize the effect of online crisis response strategies on OBA. On the basis of the study outcomes, contributions to theory and practice, limitations, and future guidelines are well discussed.


Introduction
Online purchase intention (OPI) is defined as a possibility that lies in the hands of the purchasers who aim to pay for a specific product or service in a specific time (Grewal et al., 1998). Purchase intention is a vital aspect of consumer behavior. Most of previous research scholars have attempted to recognize factors that either encourage consumers to become involved in online purchasing or discourage them (Khwaja et al., 2019;Ruiz-Mafe et al., 2018;Trivedi & Yadav, 2020). The outcomes have shown that positive incentives to purchase online (competitive prices, convenience, and excitement) are common (Dai & Zhang, 2020;Talwar, Dhir, et al., 2020a), while reasons discouraging online purchasing differ and are difficult to recognize (Talwar, Dhir, et al., 2020b;. Among the most investigated factors that may have influence on consumers' acceptance of online purchasing, the perceived risk associated with online purchasing has been of great curiosity among researchers and online retailers alike (Ariffin et al., 2020;Maziriri & Chuchu, 2017). As the acceptance of online purchasing continues, the current research must be updated and extended by investigating specific kinds of predictors that are linked with online purchasing and their influence on OPIs.
Crisis response strategy, which is identified as "immediate activity, elucidating and framing what has happened, who is responsible, and what has been done to address the online crises" (Coombs, 2007), has a significant inference for OPIs (Amaro & Duarte, 2015;Laksamana, 2018). Companies must select a crisis response strategy that is defined by the quantity of crisis responsibility credited to the organization. Scholars have established that in a preventable crisis, crisis executives may practice rebuild tactics (e.g., apology). During a victim crisis, executives should practice defensive denial strategies (e.g., denial and scapegoat) (Coombs & Holladay, 1996;Huang et al., 2006;Torales et al., 2020).  However, despite crisis response strategies being recognized as a vital variable in consumer research, prior studies have hardly begun to reveal the most exciting inquiries about online crisis response strategies, namely, why and under which conditions online crisis response strategies influence OPIs. This path of investigation is particularly significant, as the current work of  has emphasized further studies to inspect the importance of crisis response strategies in consumer literature. This scholarship aims to solve this issue in this research.
By drawing the theory of Mehrabian and Russell (1974), the Stimulus Organism Response (SOR) which states that some environmental factors can perform as stimuli (S) that influence an organism's internal state (O), thus subsequently shaping their behavioral response (R), the current study examines the effect of online crisis response strategies on OPI through online brand attitude (OBA). Crisis is an unexpected event that influences consumer attitudes toward something (Dutta & Pullig, 2011). Online brand attitude is defined as "the total of their perceptions of the brand's personality characteristics" (Chavanat et al., 2009). The advancement and growth of the brands depend on how purchasers consume goods and services in their daily life (Teik, 2018;Yasser, 2016). However, during crisis, consumer purchase intention is likely to decline, as purchasers lose trust toward the brand (Bleier & Eisenbeiss, 2015). OPI goes down and online purchasers reduce purchasing because of the online seller's bad behavior toward their purchasers when selling their products (Amaro & Duarte, 2015). Individual OPI push online brand owners to avoid an online crisis for the development of their brand (Osakwe et al., 2016). Given the subsequent lack of knowledge and the scarcity of research on how to best overcome an online crisis, firms may depend on the perceptions from the general crisis management works, which is welldefined as "a reflective model and centers mainly on image renovation, by taking corrective actions, explaining and framing what has happened, who is responsible, and what has been done to address the online brand crises" (Coombs, 2007). Similarly, on the other extreme, by offering an apology and taking corrective actions, these brands may accept responsibility to rebuild the online brand's image (DiStaso et al., 2015;Ma, 2020).
This study also recommends that the association between OBA and OPI is moderated by brand perceived usefulness (BPU). OBA comprises "the total of purchasers perceptions of the brand's personality characteristics" (Chavanat et al., 2009). Given that OBA leads to counterproductive behaviors (i.e., continue or stop buying from the brands), we then confidently believe that consumers would have a habit of engaging themselves in OPI when their BPU is high (Aditya & Wardhana, 2016;Wang et al., 2019). The previous literature on online crisis response strategies has underexplored the role of BPU in relation to OBA and OPI (Stieglitz et al., 2019;Van der Meer, 2014). Moreover, previous research posits that the impact of OBA may be mitigated by individual characteristics and appearances (Nikolova et al., 2015;Su et al., 2019). Hence, we then suggest that the positive association between OBA and OPI is more likely to be strong in an individual with high BPU perceptions (Coombs, 2007).
In summary, this study makes three novel contributions to the existing literature. First, it exclusively focuses on examining the online crisis response strategies for OPI in a new SOR framework in a developing marketplace. To accomplish this analysis, the embeddedness of concepts into mechanisms of the SOR framework to suggest a reliable and efficient parsimonious theoretical model is reasoned. These concepts comprise online crisis strategies (stimulus component), OBA (organism component), and OPI (response component). Second, by inspecting the sequential nature of this research model embedded into SOR theory, we suggest that an OBA mediates the relationship between online crisis response strategies and OPI. Third, this scholarship inspects the moderating role of BPU on the association between OBA and OPI. By doing so, the research not only theoretically contributes to academics' understanding of SOR theory applications in developing marketplaces but also offers online purchasers in developing marketplaces a model to allow them to foster purchases through convincing online brand perceptions and usefulness. We also contend that purchasers who have had a bad experience with their brands will feel it difficult in helping the brand to grow, ungrateful to benefit it, and will stop purchasing from these brands, which may decline OPI. Hence, OBA can be an important factor for the efficiency of the association between online crisis response strategies and OPI.
The following sections discuss the variables of interest from related literature. A study outline and order of testable hypotheses are well developed. The methodology part is then described, which is followed by an analysis and interpretation of the collected data. Then before we concluded with the possible future research directions, we described the implications and recommendations of our study.

Stimulus Organism Response (SOR) Theory
SOR theory is "a neo-behavioristic style for understanding the process by which people choose to enact positive (approach) or negative (avoidance) behavior in response to a specific stimulus" (Jacoby, 2002, p. 7;Kim & Johnson, 2016). Previous scholars have approved SOR theory in numerous contexts Tandon et al., 2021). The theory clarifies persons' behavioral responses (R) by seeing the impacts of environmental stimuli (S) on their internal states (O) that lead to the activation of cognitive or affective processes. These processes culminate in the growth of attitudes and dispositions and the consequences of information-seeking and decisionmaking (Jacoby, 2002, p. 7).
The value of the Stimulus Organisms Response (SOR) model rests on its holistic consideration of the affective, cognitive, and emotional processes that a person experiences while considering the adoption of a specific behavior. To elucidate the variances in decision-making processes, scholars use SOR theory in numerous settings, including tourism, consumer behavior, and in-service encounters (Gupta & Gunther, 2020;M. J. Kim et al., 2020). Previous scholars have further examined the relations among multiple predictors and consumer reactions toward organic food and other outcomes (Laato et al., 2020). Thus, based on the above, we apply SOR theory to check the effect of online crisis response strategies on OPI through OBA, including BPU as a boundary condition between OBA and OPI.

Online Crisis Response Strategies and Online
Purchase Intention Coombs (2007) distinguished amid three types of crises "(i.e., victim, accidental, and preventable crises, pp. 166-172) based on the quantity of crisis responsibility attributed to the online brand." Online brands faced with a preventable crisis may deny the responsibilities for the crisis. On the other side, these brands may accept responsibility by offering an apology and taking corrective actions to reconstruct the online brand's image (DiStaso et al., 2015). By concentrating on what is going to happen next and how the online brand will be advancing, the renewal paradigm serves as an alternative model (Kelley & Davis, 1994). The brand rebuild strategy is integrally optimistic and pursues to inspire the online customers to remain with the online brand and to continue investing in the brand to a level that exceeds the pre-crisis status (Iglesias et al., 2020;Kaur et al., 2020;Ray et al., 2019). When engaging in the brand rebuild strategy, an online brand will reimburse the affected subscribers and the announcement of a certified declaration of apology . Online crisis response strategies have been regarded as the future online crisis communication management (Bundy et al., 2017;Giotakos & Papadomarkaki, 2016). Although crisis communication strategies have existed for some time, it is still comparatively unused (Bundy et al., 2017), thus opening up opportunities to research into this area.
The importance of OPI has recently been documented by several researchers. For example, , by adapting the theory of consumption values, used a framework for forecasting purchase intention to online travel agencies. Using data from 809 online travel agency users, the authors showed that "quality-of-benefits, social rank, monetary, preference and information standards forecast purchase intention toward online travel agencies, with the main driver being the quality-of-benefits value, followed by the preference value." They further argued that "the strength of the association among these values and purchase intention varies among users in different age groups and users with different levels of privacy and security concerns, hygiene consciousness, and visibility perceptions." In a similar vein, Marutschke et al. (2019) provided evidence that purchaser involvement, restaurant search, ease-of-use, and listing are the important drivers of intentions to use food delivery apps. Likewise,  extending the innovation resistance theory studied the obstacles to positive purchase intentions toward online travel agencies. The authors posited that the benefits obstacle is the main reactant of purchase intentions while security and privacy concerns and the vulnerability obstacle demonstrate a positive association with intentions. The authors also showed visibility prompting the strong power of the relation among the benefits obstacle and intentions to purchase. They discovered that the strong power of this relation differs among young, middle-aged, and previous users. A Japanese research inspecting the association among dissimilar uses and satisfactions of mobile instant messaging apps, extension, and purchase intentions showed that uses and gratifications motivate mobile instant messaging users to consume both positive purchase intentions toward virtual goods, such as stickers, and extension intentions toward mobile instant messaging apps . Scholars who used the novel behavioral reasoning theory found that the attitude-intention gap related to the purchase of organic food can be clarified by incorporating the investigation of values and reasons in consumers' decision-making processes .
Online crisis response strategies must be matched to the online brand crisis. For example, once a preventable crisis happens, for which the online brand is perceived as extremely accountable, the brand should use a rebuilding approach by apologizing to the consumers (Haque et al., 2015;Hoque & Hossan, 2020). As the most appropriate repair effort that results in the most significant post-crisis effect, the rebuilding approach is deemed highly effective in regenerating purchase intention (Crijns, 2017). While the diminishing online crisis response strategies aim at presenting that the crisis is not spoiling the brands and at varying the number of ascriptions created by the investors, the rebuild strategy aims at neutralizing the tarnished name of the brand; although the approach is deemed highly effective, it is also the most expensive approach (Smith & Tyler, 1997), as it could lead to considerable monetary damage for the brand. According to Crijns (2017), when the brand's representative uses a rebuilding approach in which apologies are offered, we assume it to have a positive effect on the customers' OBA and succeeding consumer OPI (Crijns, 2017;Tiruwa et al., 2016).
Nevertheless, once the representative uses a diminishing approach in reaction to a preventable crisis, accountability is abandoned, and they will not offer any apology (Coombs & Holladay, 2009;Ma, 2020). Therefore, consumers will be less motivated to forgive the brand in crisis and would, therefore, have a negative OBA, which would harmfully influence the consumer's OPI (Lis & Fischer, 2020;. Therefore, we expect that the impact of online crisis response strategies will positively predict OPI. Purchase intention is also a probability that lies in the hands of customers who intend to pay for a particular product or service (Grewal et al., 1998;Zhang et al., 2020), which depends mostly on online crisis response strategies (Vignal Lambret & Barki, 2018). Research shows that successful online crisis response strategies may increase OPI (Beldad et al., 2018;Coombs, 2007;Ma, 2020). Some studies also indicate that online crises negatively predict the intention to purchase (Cai & Leung, 2020;Dutta & Pullig, 2011). Therefore, we expect that the impact of an online crisis response strategy will positively predict OPI.
Hypothesis 1 (H1): Online crises response strategies have a positive relationship with online purchase intention.

Mediating Role of Online Brand Attitude
The attitude toward a brand lies on the general assessment of the brand by consumers; it reflects consumer reaction to the brand and can be shaped from beliefs about the intrinsic attributes, functional benefits, and accompanying experience of the brand (Keller & Aaker, 1998). OBA comprises "the total of purchasers perceptions of the brand's personality characteristics" (Chavanat et al., 2009). More specifically, consumer OBA can indicate an online brand's association with subscribers.
Prior studies have considered brand attitude as a mediator Wassler, 2019). For example, Ramesh et al. checked the effect of corporate social responsibility activities on purchase intention. The authors gathered surveys from citizens of India buying products from FMCG brands topping the corporate social responsibility spending list. Their results offer that buyers process corporate social responsibility details instinctively and may not remember explicit details. However, these customers are more likely to comprise the brand in the consideration set induced by positive attitudes trailing behind. Zarantonello and Schmitt (2013) showed that brand attitudes mediate the association among events and brand equity only for certain types of events (e.g., trade and street events). Some scholars have also revealed that event presence can result in more favorable brand attitudes; in turn, brand attitudes regulate stronger purchase intentions (Sun et al., 2020;Sun & Wang, 2019). Other researchers' opinions clarify that a brand with high awareness can endorse brand loyalty to the consumers; higher brand awareness increases the consumers' brand trust and purchase intention (Ramli, 2015;Ramli & Sjahruddin, 2015).
Consumer OBA can function as a good symbol of consumers' feelings toward the online brand, their relationship with it, and how consumers assess online products, such as Pay TV, online games, and online games (Balmer, 2011;Foroudi, 2019). To sustain the consumer's OBA, online brands must align the online crises with their online crisis response strategies. Online crisis response strategies not only minimize damage caused by the crisis but also help to continue a meaningful relationship post-crisis (Barbarossa et al., 2016). Coombs (2007) found that the speed with which online brands respond to crises affects how consumers perceive online brands. Online crisis response strategies are also perceived as a sign of online brand advancement (Coombs, 2007). In view that OBA influences behavioral intention (Ajzen, 1991;Roh, 2017), a person with higher OBA is more likely to continue to purchase online with a positive belief in online crisis response strategies. Thus, they are more likely to develop a positive intention to adopt online crisis response strategies. Therefore, we argue the following hypothesis: Hypothesis 2 (H2): Online brand attitude mediates the relationship between online crisis response strategies and online purchasing intention.

Moderating Effects of Brand Perceived Usefulness
BPU is defined as "the degree to which a prospective adopter believes that using a particular system would improve his or her job performance" (Davis, 1989, p. 3). Following this definition, perceived usefulness is contingent on the purchasers' observation of the outcome of a certain system. If the system is "capable of being used advantageously" (Davis, 1989), then positively perceived usefulness exists. Perceived usefulness originates from what amount users trust that the system or technology will upsurge their job performance (Davis, 1989). Furthermore, perceived usefulness has been revealed to have a stronger effect on the intent to use than the perceived ease of use. This outcome is clarified by the notion that regardless of whether the technology is easy to use or not, a consumer will not accept it if regarded as unusable.
Prior studies have highlighted the moderating effect of BPU as a moderating variable in a number of consumer intention situations. For instance, Mensah (2019, p. 43) in his work used the technology acceptance model (TAM) as a theoretical stance to check the moderating role of perceived usefulness among the trust in the Internet, trust in the government, and the intention to adopt e-government services. The outcomes of their study showed that perceived usefulness had a noteworthy moderation effect on the impact of trust in the government on the intention of using e-government services. However, perceived usefulness did not show a significant moderation effect between trust in the Internet and the intention to adopt e-government services.
In a similar vein, Pitafi et al. (2020) established a theoretical argument to utilize perceived ease of use with habit and psychological dependence to forecast social networking addiction. Using data from the university students in Pakistan the authors verified their model. They further argued that the relation among perceived ease of use, social network addiction, and psychological dependence is moderated by perceived usefulness. Moreover, Sohn (2017, pp. 22-23) explored the roots of purchasers' usefulness insights through the example of mobile online shopping acceptance in Europe. The outcomes showed the importance of consumers' beliefs about the quality of mobile online stores in the development of usefulness perceptions prior to acceptance.
The perceptions of usefulness are crucial factors for forming consumer purchase intentions and behavior. The concept of BPU has not been argued as an easy way to reach, as it is highly related with the contextual use. Thus, researching its influence among individuals' brand attitudes and online consumer intention for purchasing is a crucial aspect in consumer research. OBAs help form consciousness of the values of certain technology, products, or activities, which in turn becomes a valuable factor for online purchasers in the acceptance of products and technology. In the context of online purchasing with technological usage, BPU shows that using technology will improve consumers' performance in learning and acquisition and toward OPI. Therefore, we hypothesize the following: Hypothesis 3 (H3): Brand perceived usefulness positively moderates the relationship between online brand attitude and online purchase intention such that when brand perceived usefulness is high, the relationship of online brand attitude and online purchase intention will be high.

Sampling and Data Collection
As this study emphasis on the adoption of online crisis response strategies, the respondents should have been buying at least in any of the online platforms. Sensibly, these individuals had been exposed to online crisis response strategies. To collect data, the study used an online survey and data were collected from working professionals who were currently undertaking a Master of Business Administration (MBA) degree from a large university in China. The researchers outlined the aim of the study to the respondents before administering the survey so that they could have a clear picture of what we wanted to do. Furthermore, they were told that the survey was not compulsory but voluntary. Then, they were asked to sign the agreement letter that was attached to the survey. This category of purchasers had been known to make most of the purchases online due to convenience (Coombs & Holladay, 2014). They were more likely to identify various strategies implemented by Chinese companies before making decisions that used online platforms to buy goods and services. They also emerged from a diverse purchaser group comprising various ethnicities, ages, and backgrounds, which provided a multifaceted purchasing segmentation. The respondents completed a questionnaire that contained the demographics and online crisis response strategies. They then rated the items related to OBA, BPU, and OPI. The responses constituted a 100% response rate.

Data Analysis
Our analysis fell into three main phases. First, to check the validity of our measurement scales, we conducted a confirmatory factor analysis (CFA). A combination of the chisquare test statistic with consistent degrees of freedom and statistical significance (χ 2 /df, p) was used, comparative fit index (CFI), and the root means the square error of approximation (RMSEA) to assess the fit of our CFA models (Hu & Bentler, 1999). Also to rule out the presence of common method variance in our data, Harman single factor and common latent factor tests were conducted. Finally, we used Process MACRO (Hayes & Scharkow, 2013) to assess the mediation and moderation effect. The said techniques had also been used in recent studies to validate their proposed relationships Kiani, Ali, et al., 2020;Kiani, Liu, et al., 2020;Usman et al., 2020;Zhai et al., 2020). We found no significant difference in study variables concerning age, gender, education, marital status, and experience. Therefore, our results were substantially corresponding with or without the control variables; hence, the results without controls are reported here (Becker, 2005). The respondents consisted of 60.2% males with ages ranging from 25 to 30 years old. Most of the respondents reported having used an online purchasing platform between one and three times a month.

Survey Instruments
The following variables were used to test the proposed hypotheses. The items were initially in English but were later back-translated to Chinese in line with Brislin's (1980) translation-back-translation procedures. To increase consistency and ease of comparison with previous works in this zone, we operationalized each hypothesis with numerous items. The other study variables were measured on a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree).
Online crisis response strategies. The operation of two crisis response strategies, that is, the rebuild and diminish approaches, are involved in this study. The participants were asked to rate the level to which they felt the organization took responsibility and decided to apologize for the occurrence of the crisis. This rating was in line with the Situational Crisis Communication Theory of Coombs (2017), which explains that crisis response strategies are in contrast with the number of responsibilities organizations take and the willingness for the brands to apologize. This study adapted the apology genuineness scale advanced by Bachman and Guerrero (2006). Example items were "Did the online brand's CEO take responsibility for this crisis?" and "Did the online brand's CEO offer you a sincere apology?" Cronbach's alphas were .82.
Online brand attitude. This study adapted Aaker and Biel's (1993) four-item scale. Sample items included, "I like the idea of online brand's services" and "Purchasing online brand's services is a good idea." Cronbach's alphas were .75.
Online purchase intention. Respondents rated five items from Paul et al.'s (2016) purchase intention scale. Sample items included "I will consider paying for online streaming services because they are unlimited" and "I will consider switching to online streaming services for economic reasons." Cronbach's alphas were .72.
Brand perceived usefulness. Respondents were requested to rate the extent of accepting or not accepting the following four reports adapted from Saadé and Bahli's (2005) perceived usefulness scale. Sample items include "Using the online streaming service would make it easier for me to watch all my favorite programs from any device" and "Using the online streaming service would make it easier for me to be more innovative by providing the opportunities to search more information." Cronbach's alphas were .84.

Control variables.
The demographic details of the respondents included gender, age, education, marital status, income, and experience. The demographic details of the respondents are mentioned in Table 1.

Common Method Bias (CMB)
As projected by MacKenzie and Podsakoff (2012) that when collecting the data from a single source, same time, and selfreported, then the CMB can be an issue in the data set. Three approaches were used to address this issue suggested by MacKenzie and Podsakoff (2012). First, Harman's one-factor test was used to test CMV. Four factors were produced in the investigation with eigenvalues > 1.0. The first factor was only 43.48%, which was less than 50%. Second, the common latent factor method was utilized. The standard regression weights were calculated through CFA. Then, included common latent factor in the research model, and reproduced the CFA results. Finally, the regression weights were matched of both the analyses and no dominant factor was found emerged from the results. Moreover, the result in Table 2 also indicates intercorrelation of all the constructs (r < .90), which is another indication of the absence of CMB as suggested by Pavlou and El Sawy (2006). Hence, these results confirmed that in the current study data, CMB was not a substantial issue. Note. N = 392; gender (1 = male, 2 = female), age (1 < 25, 2 = 25-30, and 3 > 30), education (1 = bachelor's, 2 = master's, 3 = PhD), marital status (1 = unmarried and 2 = married), and experience (1 = <1, 2 = 1-5, 3 = 6-10, and 4 = >10).

Reliability and Validity
To check our questionnaire for the reliability and validity of the constructs, we have tried to use different techniques of data analysis. In Table 3, mean scores, standard deviations, and Cronbach's alphas of the scales are all shown. For example, in the table, the Cronbach's alpha, composite reliability (CR), and average variance extracted (AVE) of each construct were found to attain the threshold value of (0.53) and (0.93), respectively. Convergent validity was extracted through the value factor loading of each contract. In Table 3, the factor loading of each item was more significant than the cut-off value of .70, which showed that all constructs exhibited outstanding convergent validity. Moreover, all values for the AVE were less than the maximum shared variance. This result showed that the measurement model had appropriate convergent validity. In addition, the square roots of the AVEs for all variables were more important than the correlations among constructs as shown in Table 3, which confirmed the discriminant validity of the research model. These results recommended a good discriminant validity for the data set. Hence, these results possessed satisfactory convergent and discriminant validity and reliability.

Measurement Model
Following previous studies (Latif et al., 2020), CFA was conducted by AMOS (version 21.0) software. The measurement model included four latent factors, namely, online crisis response strategies, OBA, BPU, and OPI.
As recommended by Hair et al. (2010), the fit of the full measurement model was inspected to validate the association between study variables and goodness-of-fit measures. Hence, the results confirmed that measurement model fit indices were in the given range, χ 2 /df = 0.951, RMSEA = .071, adjusted goodness of fit index (AGFI) = 0.88, standardized root mean square residual (SRMR) = 0.045, normed fit index (NFI) = 0.92, and CFI = 0.95. Therefore, this model was a satisfactory fit to the data as the full measurement model fit revealed.

Hypothesis Testing
The correlation between online crisis response strategies and OPI was estimated at r = .46**. The correlation between online crisis response strategies and OPI when mediated by OBA was estimated at r = .33**. Also, the correlation between OBA and OPI when moderated by BPU was estimated at r = .34**. The correlation was positive and significant. These findings supported our proposed hypothesis H1, which stated that online crises response strategies positively predicted OPI.

Tests of Mediation
To test H1 and H2, we used PROCESS macro suggested by Hayes and Preacher (2010), and Hayes and Scharkow (2013) Note. The bold and italic values mentioned in the parenthesis are the discriminant validity of the constructs. OCRS = online crisis response strategies; OBA = online brand attitude; BPU = brand perceived usefulness; OPI = online purchase intention. *p < .05 level (2-tailed). **p < .01 level (2-tailed). .73 Note. All factor loadings are significant at the p > .05 level. CR = composite reliability; AVE = average variance extracted; OCRS = online crisis response strategies; OBA = online brand attitude; BPU = brand perceived usefulness; OPI = online purchase intention.
in SPSS 22.0 to find the direct effect of online crisis response strategies on OPI and an indirect effect via brand attitude. Table 4 shows that online crisis response strategies were positively and significantly associated with OPI (β = .41, t = 10.36), thereby accepting H1. OBA mediated the relationship between online crisis response strategies and OPI. The confidence interval (CI) [0.07, 0.11] zero were not included, thereby accepting H2.

Moderation Test
The moderating role of BPU between the associations of OBA and OPI was proposed in H3. Table 5 shows the results of the moderating role of BPU between the associations of OBA and OPI. Therefore, the significant interaction (OBA × BPU) value (β = .10, t = 4.98, p< .01) accepted H3. To examine the nature of interaction effects further, an OBA was split into low (−1 SD) and high (+1 SD) levels. In Figure 2, graphical illustrations show the moderating effects of OBA. The positive association in online crisis response strategies and OPI was weaker (β = .20, t = 3.7, CI [.09, .31]) at the low level of OBA, and the same relationship was positive and stronger (β = 39, t = 7.87, CI [.29, .49]) at the high level of OBA. Therefore, these findings provided additional support for H3.

Stimulus Organism Response Theory
The results propose that BPU is a "significant construct in forecasting the intention to adopt SOR. The results verify preceding studies conducted on online crisis response strategies solutions in China" (Amin & Roberts, 2008, p. 43) and on OPI (Amaro & Duarte, 2015). Given that convenience is an important aspect in the adoption of online crisis response strategies (Bigné-Alcañiz et al., 2008), if consumer finds the innovation very useful it will be easy for them to adopt the services that are going to be offered. This assumption is due to the advantages brought about by online crisis response strategies over OPI in terms of quicker and more convenient purchasing. The results will have implications in encouraging usage based on the rewards of online crisis response strategies. Likewise, BPU is a significant factor in predicting the acceptance of online crisis response strategies (Gavilan & Avello, 2020). The result is consistent with a past study on online crisis response strategies in China (e.g., Goodman et al., 2014;Romenti et al., 2016). As online crisis response strategies work with the aid of the CRM technology, we must establish crisis management compared with offline shopping. As such, online crisis response strategies are generally perceived as less complicated to apply, as they only require an online brand owner to admit their faults once a crisis has happened. Horn et al. (2014) suggested that online crisis response strategies will inspire the creation of OPI; the outcome is consistent with the consequences of Davis (1989) throughout the initial adoption where the association was only significant after a certain period of usage.

Theoretical Contributions
Our study contributes very significant implications to the theory. First, it integrates research on online crisis response strategies. It offers empirical support to fill a significant gap in our understanding of how the association between online crisis response strategies and OPI is mediated through OBA and how individual behavior, such as BPU, moderates the relationship of OBA and OPI. On the basis of the SOR framework in parallel with a previous scholarship, our result Note. Bootstrap sample size = 5,000. OPI = online purchase intention; OBA = online brand attitude; LL = lower limit; UL = upper limit; CI = confidence interval. **p < .01. Note. Bootstrap sample size = 5,000. OPI = online purchase intention; OBA = online brand attitude; BPU = brand perceived usefulness. **p < .01.
explores the effect of online crisis response strategies on OPI in the online brand service context. As a result, and in line with previous scholarships, our results suggest that online crisis response strategies act as the main predictor, which further influence the OPI performance of purchasers. Previous research in developing economies has paid attention to identifying the antecedents of online purchase attention, such as decision-making styles (Ashfaq et al., 2019;Prakash, 2018) and purchase barriers (Kushwah et al., 2019;Shen et al., 2020). This research broadens academic knowledge by recognizing the significance of online crisis strategies as a value that critically impacts buyers' brand attitudes for online purchasing. Second, this study extends the geographic possibility of previous scholarly inquiries into OPIs, especially in the context of Chinese markets. In addition to the same line of research, the outcomes of this study conclude that online crisis response strategies positively predict OPI in a developing economy (e.g., China) and that this association is mediated by the OBA. By using SOR theory, this research contributes to the development of preceding theoretic foundation applied to investigate Chinese consumers. In general, three hypotheses were inspected in this study, and the results mostly support our hypothesized associations. And our results vividly approve that individuals with high BPU perceptions strengthen the association between OBA and OPI. We prolonged this line of study by approving that BPU increases the purchasing outcome (i.e., OPI) of online crisis response strategies. This result determines the theoretical generalizability of scholarships on online crisis response strategies beyond the Chinese setting in which it has been mostly tested.
Third, the excessive use of online crisis response strategies automatically results in OPI (Coombs, 2007). In this study, the inspected variables, namely, online crisis response strategies, exert a positive influence on positive consequences credited to OPI. Therefore, academics must justify contextual issues in theories or frameworks focused toward online product purchase intentions. The current research reveals that the boundary between online branding and online crisis literature is valuable in further understanding post-crisis customer evaluation of online brands. Previous studies have explored various communication strategies with online crisis response strategies. Scholars have also investigated how online crisis response strategies result in consumers' OPI. By extension, this scholarship contributes to SOR theory by inspecting the role of online crisis strategies on consumers' OPI. Our results are in line with earlier studies on the positive consequences of online crisis response strategies.

Practical Implications
This study offers a number of managerial implications. Evidently, online crisis response strategies prevail in online platforms (Choy et al., 2018;Lee et al., 2021), and their damaging outcomes have been noted in the existing literature. The suggested findings reported in this study advised online brands to introduce certain programs for the owners of the brands to decrease online crisis behaviors. For instance, the owners of the brands may be advised to participate in the established programs or specialized training with a focus on crisis management and interpersonal skills improvement (Goodman et al., 2014;Romenti et al., 2016). Furthermore, through human resource practices, such as training, the owners of the online brands can aim at diminishing the occurrences of online crisis.
Online crisis response strategies are the tools for OBA (Coombs & Tachkova, 2019;Othman & Yusoff, 2020). OPI behaviors increase purchaser trust, and the performance of an online brand also enhances the brand's aptitude to fetch innovation, which, in turn, poses to the planned goals of the organization (Khatoon et al., 2020). However, our research suggests that the higher-ups in the business may control OPI behaviors by encouraging online crisis response strategies. One way to increase online crisis response strategies is to diminish its online crisis and contextual causes (Coombs & Holladay, 1996;Ma, 2020).
Online brands should be attentive to the cause of the online crisis among online brand owners. Online crisis is more likely to occur when online brand owners are not careful enough. In addition, online brands should ask directly or indirectly with the consumers about their changing expectations. In light of the findings on online crisis response strategies, the providers of these strategies should separate the marketplace and modify to the exact needs of this niche market with the suitable services needed. They must do so since the required services by innovative customers might be different from those of non-innovative customers.
Finally, our research findings may also be a silver lining for managers and customers. Given that BPU is "an important factor in this study, online brands could deliberate employing opinion leaders, celebrity endorsements or WOM effects, as eliminating online crisis completely is often difficult" (Lin et al., 2017, p. 2) As mentioned previously, our findings suggest the significance of image; campaigns advertisement should factor into consideration a positioning strategy that could reflect the upper-class status of accepting SOR theory.

Limitations and Directions for Future Research
The results should be understood by considering a number of limitations that need to be addressed by future scholars. The generalizability of the results is limited because the sample only included to students who were working professionals undertaking the Master of Business Administration (MBA) program. Hence, this model could be tested in another social/ group setting to see whether the outcomes are alike. Furthermore, forming the generalizability of existing findings would also help greatly. Therefore, we invite future scholars to adopt empirical models by including additional crisis response strategies. For example, we have taken two strategies, namely, diminish and rebuild strategies, which were appropriate in our context. However, a broad strategy may be appropriate and should be inspected in other settings.
The test of causality was not allowed in our self-reported survey data. They may be subject to confounded constructs and do not give a clear way to the effects of the tested variables. Scholars are encouraged to conduct another technique to gather data to get precise evidence about the causality and direction of effects among different variables. Possible scholarships can also accept dissimilar scale properties and practice other possible corrective measures, like balancing positive and negative items. This scholarship focused on the young population group segment within the same country with different traditional backgrounds. Culture is extremely inspiring individuals to purchase online. Hence, scholars are encouraged to consider this possible gap and other disturbances associated with the communication crisis (Aghekyan-Simonian et al., 2012).
Given that the respondents are online customers, concentrating exclusively on this group may pose potential unfairness, as behavioral differences may be important between those who use and those who don't use it (Ashfaq et al., 2020;Patsiotis et al., 2012). Therefore, we invite future scholars to consider those who don't use it as well so that judgment can be made. Furthermore, our study deliberately only takes into consideration the customers' viewpoint. The acceptance rate of online crisis response strategies also depends on the availability of an online crisis. Therefore, we suggest that future studies gather the merchants' perspectives. As most of the online crisis response strategies have the same functions, the setting of this study refers to the general and widespread OPI in China. Types of online crisis response strategies may differ between merchants. The upcoming studies should motivate different online crisis response strategies for various purposes. Prospective research should obtain data from all cultural contexts to control the shared method technique.

Conclusion
The results of the present study are encouraging because they provide additional empirical evidence to the notable role of online crisis response strategies in relation to OPI. Also, our investigation plays a big and important role in contributing to the literature by testing an underlying instrument, that is, BPU, in the online crisis response strategies. This study observed individual BPU as a borderline condition for online crisis response strategies. Ultimately, we hope these results will shed light on both theoretical contributions and also deliver valuable information for online brand owners and offline brands, online brand decision-makers, merchants, software designers, governments, and practitioners when expressing their communication and business policies related to SOR framework acceptance.

Authors' Note
Usman Ghani is now affiliated to School of Management, Iqra University, Karachi, Pakistan and College of Education, Zhejiang University, Hangzhou, China.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Professor Yongjun Li would like to thank the National Science Foundation of China (Nos. 72071192, 71671172, 71631006) and the Fundamental Research Funds for the Central Universities (WK2040000027).