“E-WOM and consumers’ purchase intention: An empirical study on Facebook”

Nowadays, organizations use social media to promote their services and products. At the same time, they use different tools to convey their messages, such as Facebook. Therefore, this study aims to investigate the factors that affect the e-WOM on Jordanian consumers’ purchase intention over Facebook. The study uses the information acceptance model (IAM) to examine the impact of information credibility, information quality, information adoption, and information usefulness over Facebook on Jordanian consumers’ purchase intention. The study uses cross-sectional quantitative research and is conducted online. The questionnaire was distributed through Facebook and WhatsApp, and the people who used only Facebook were allowed to complete the survey. Out of 327 filled questionnaires, only 304 were valid for further analysis. Collected data were coded in SPSS, and after confirming the validity and reliability of the tool, the correlation between variables was checked. In addition, multiple regressions were used to test the hypotheses. Multiple regression results show that the E-WOM can explain 49.2% of the total variation in the consumers’ purchase intention, where R2 = 0.492. Information adoption has the strongest effect on consumers’ purchase intention (β = 0.489), followed by information usefulness (β = 0.204). In contrast, information credibility and information quality do not have a significant effect on customers’ purchase intention (0.189 and 0.312, respectively). This study helps companies and businesses that have pages on Facebook to understand how consumers engage in the e-WOM on business pages and consider the consumers’ reviews, comments, or posts.


INTRODUCTION
Social media channels have been an essential part of individuals' lives. For instance, according to Facebook's fourth-quarter report, Facebook is the most used tool among all social network platforms. It has monthly active users of 2.80 billion (Facebook, 2021). The dramatic growth in Facebook users has empowered them to openly express themselves, interact with each other, and share experiences through e-word of mouth. On the other hand, marketers have made sure to have a presence on Facebook to communicate with their consumers and influence their purchase decision directly (Tien et al., 2019). Furthermore, e-WOM on social networks, including Facebook, directly influences the consumers' purchase intention as it acts as a marketing instrument to communicate and interact with consumers (Alnsour, 2018).
Since E-WOM has a vital role in social media channels and directly impacts consumers' buying intention, this study is valuable for businesses that try to sell their products and services in Jordan via the Facebook platform. It helps them to understand what factors affect the Jordanian consumers' purchase intention, so they give more attention to what customers think of their business. In addition, this study is important to buyers who are Facebook users. It encourages them to express their views about businesses and helps users understand that their opinions matter, especially since Facebook is very popular in Jordan. According to a recent statistic by StatCounter (n.d.), Facebook has the majority of the market share of 74.7% in Jordan among other social media channels, followed by YouTube with 15.43%, and 4.51% of Twitter market share (StatCounter, n.d.). Moreover, this study will assist Facebook users in evaluating the e-WOM message and recognize the power that they got as indirect marketers for businesses. Furthermore, this paper will be helpful for future research about e-WOM and social media in Jordan and understanding of what affects Jordanian consumers' purchase intention on social networks. E-WOM is a trending topic that got many researchers' attention, especially on social networks, where it is noticed that e-WOM has a vital role in consumers' buying intention and has changed the traditional WOM (Elhadidy, 2017). Facebook is a substantial channel of e-WOM, where users have the power as buyers to engage with the business and can openly express their opinions towards products and services and their intentions of purchasing (Wen & Aun, 2020). As a social network channel, Facebook also empowered organizations to interact with buyers and affect their personal opinions without contacting them directly (Bataineh, 2015). However, it was found in previous studies that there is limited research investigating e-WOM in the Middle East including in Jordan (Al-Shibly & Mahadin, 2017). Therefore, this study is dedicated to investigating the effect of e-WOM on consumers' purchase intentions.

LITERATURE REVIEW AND HYPOTHESES
An extensive revision of the recent related studies indicated that the information acceptance model was mainly applied to study e-WOM on consumers' purchase intention. The information acceptance model (IAM) was created in 2003 by Sussman and Siegal (2003 Figure 1. In general, purchase intention is people's probability of buying a service or product (Tien et al., 2019). Purchase intention for e-WOM is described as a process of evaluating word-of-mouth information concerning a product or service. It also eventually leads to the buying decision (Sutanto & Aprianingsih, 2016). When the message of e-WOM on social networks is positive and powerful, it is highly expected to impact consumers' purchase intention (Erkan, 2016). Therefore, purchase intention is a process that may lead to purchase decisions when the e-WOM information is credible, useful, offers acceptable quality to consumers, and in the end, is adopted by consumers.
According to related studies that discussed e-word-of-mouth, it is described as the feedback content from a consumer's viewpoint, which can be negative or positive related to a business, product, or service through different internet platforms ( Erkan (2016), e-WOM is also found to influence consumers to buy their goods and services online, especially on social media.
The increasing use of social media platforms has changed people's lives (Saleem & Ellahi, 2017). It created opportunities for e-WOM conversations that allow users to rate the services and products with their families, friends, and acquaintances, which plays a dynamic role in reshaping and affecting the consumers' purchase intentions (Erkan, 2016; Saleem & Ellahi, 2017). As mentioned before, the e-WOM has an influential role in assessing the products and services from negative and positive comments on social media platforms, which has affected the consumers' purchase intentions (Saleem & Ellahi, 2017

METHODOLOGY
A quantitative method was applied based on related studies (Erkan, 2016). This method was used to determine the opinion and behavior, as well as to generate a result in a numerical statistical form that can test the hypothesis discussed in the pre-vious section. Moreover, it helps to explore the effect of the independent variables (IQ, IC, IU, and IA) on the dependent variable (purchase intention of the Jordanian consumers). The study uses the cross-sectional sampling method to collect responses from as many participants as possible in a short period. The questionnaire was prepared by an expert translator (the official language in Jordan) and was translated back into English (the original questionnaire language) to make sure that the Arabic translation was accurate (Brislin, 1970). The survey was built using the Google forms program in both languages (Arabic and English) and distributed online via several online communication channels like Facebook and WhatsApp, as recommended by many researchers of related studies (Troise et al., 2021). The online survey is the most convenient method to collect responses cost-effectively from an enormous number of people in a limited time, especially during the COVID-19 pandemic, where a face-to-face approach is impossible. Finally, the questionnaire was sent to the professor in marketing for testing purposes.

Measurement, scaling, and data collection
The questionnaire was created based on Erkan The sample size of the survey has targeted all people in Jordan with an active Facebook account and 15 years old and over. The questionnaire was collected from 327 participants; the sample size was chosen through judgmental sampling, so respondents were selected based on their experience (Etikan & Bala, 2017). Out of the 327 participants, 304 answered "Yes" to the filtering question "Do you have an active Facebook account?" and completed the questionnaire. Moreover, the valid data and results were downloaded and analyzed using SPSS (Yee, 2016). Therefore, the sample size is a valid reference to the rule of the minimum acceptable sample size, which is between 30 and 500. Also, according to Hill (1998), in multivariate studies like multiple regression, the minimum acceptable number of respondents must be at least ten times larger than the number of variables.

RESULTS AND DISCUSSION
The online survey was used to collect the data from 304 participants who have an active Facebook account and are willing to share their experience of e-WOM and its influence on their purchase intention. The data from the demographic information is illustrated in Table 1 Table 2 shows the descriptive analysis, which includes the means and standard deviation of the items obtained from 304 active Jordanian Facebook users. The correlation analysis shows a significant positive correlation between variables, as shown in Table 3, where p < 0.01. When correlation coefficients take a value between 0.21 to 0.40, it shows a weak correlation; when it is between 0.41 to 0.60, it shows a moderate correlation, while a strong correlation takes a value from 0.61 to 0.80 (Hair et al., 2010). Table 3 shows that most of the relationships between the variables are moderate, while there is a strong relationship between IU and IQ, IA and IU, and purchase intention and IA. On the other hand, IC and IA have a weak correlation with customers' purchase intentions.
The study uses multiple regressions to examine the relationships between the four independent variables (IC, IQ, IU, and IA) and customers' purchase intention. According to Table 4, the value of R 2 is 49.2%, which indicates that 49.2% of the total variation in the consumers' purchase intention could be described using the four constructs. The outcomes of the multiple regression analysis are shown in Table 4. Table 4 proved that the differences between the variables are statistically significant since R 2 = 0.492, F = 72.302 p-value = 0.000, which shows that e-WOM significantly affects consumers' purchase intention.
As shown in Table 4, the overall significance of e-WOM shows that e-WOM significantly affects consumers' purchase intention (R 2 = 0.492, F = 72.302, Sig. = 0.000), as the significant value for the coefficients of the regression analysis shows whether the relationship is statistically significant or not. Therefore, the significance level should be less than 0.05, as advocated by Sedgwick (2015).
This means that consumers in Jordan are influenced by the online messages and are likely to intend to buy or not because of the e-WOM; thus, the first hypothesis is accepted. Similarly, there is also an insignificant impact of IQ on purchase intention (β = 0.058, t = 1.013, Sig. = 0.312,) which means that the IQ of e-WOM over Facebook does not influence Jordanian consumers' purchase intention. Therefore, the second sub-hypothesis, H 1.2 , is rejected. On the other hand, it is found that there is a significant influence between IU and purchase intention (β = 0.204, t = 3.374, Sig. = 0.001) and between IA and purchase intention (β = 0.489, t = 9.251, Sig. = 0.001). This means that Jordanian consumers' purchase intention is affected by the word-of-mouth message over Facebook; when the message is useful and offers sufficient information, they are willing to engage in this message and adopt it. Thus, the third and fourth sub-hypothesis are both accepted.
This study aims to explore the factors that affect the e-WOM on Jordanian consumers' purchase intention over the Facebook platform. The model was examined quantitatively by using an online survey to collect data. The model of this study is theoretically based on existing literature that has used four variables of the information acceptance model (IAM): IC, IQ, IU, and IA. The find-

CONCLUSIONS
This study aims to explore the effect of e-WOM on Jordanian consumers' purchase intention over Facebook. The study uses the information acceptance model (IAM), which includes the following independent sub-variable: information credibility, information usefulness, information quality, and information adoption, and the dependent variable is customers' purchase intention. Data were collected online using a survey, and 304 responses were used for analysis. The findings show that e-WOM significantly affects Jordanian consumers' purchase intention over Facebook. The information adoption has the strongest influence on Jordanian consumers' purchase intention, then the information usefulness, while the information credibility and the information quality of Facebook were not having a significant effect on Jordanian consumers' purchase intention. Therefore, the marketers have to support their customers to spread positive e-WOM about their brand, service, or product by asking satisfied consumers to write reviews, posts, or comments without pressuring consumers to do so and without any deceptive practice because such unethical acts may lead to further negative e-WOM.
Marketers may also provide channels for unsatisfied consumers to express their experiences and communicate promptly with them to solve their claims and encourage them to change their negative e-WOM. Additionally, the findings of this study show that IC and IQ have no impact on Jordanian consumers' purchase intention. It is recommended that marketers may create forms for customers' attention and educate them on how to use them, so the reviews will have better quality and be perceived as credible information by the readers. This study focuses on e-WOM over Facebook. This can help companies understand how consumers engage in the e-WOM on business pages and take the consumers' reviews, comments, and posts into significant consideration, especially the negative ones.
Finally, despite the above findings, many limitations are recommended to be resolved in future research. First, this study has examined four variables related to IAM and has not examined any external variables. Therefore, it is recommended to include other factors that are related to e-WOM, which may influence the consumers' purchase intention. Second, a cross-sectional approach was applied to understand the consumers' purchase intention at one point in time. Thus, future studies are required to repeat the study by applying a longitudinal approach to explore the effect of e-WOM on purchase intention at multiple points in time. Third, the findings of this study are related to Jordan. Thus, conducting similar research in other settings is recommended to study individual consumers' purchase intentions in different cultures and regions. Finally, this paper explored the impact of e-WOM on all businesses in general. In contrast, in future studies, it is worth focusing on a specific industry to investigate all aspects that may affect e-WOM.