Factors affecting the credibility of online reviews on TIKI: An assessment study in Vietnam

Article history: Received: January 18, 2020 Received in revised format: February 29, 2020 Accepted: February 29, 2020 Available online: February 29, 2020 The purpose of this study was to test the determinants affecting the credibility of online reviews. In this model, the relationships between argument quality and peripheral cue and credibility of online reviews were evaluated. The measurement model and the conceptual model depicting hypothesized relationships were evaluated based on responses from 460 customers using multiple linear regression analysis, accordingly. After the data collection process was completed, all data were put into SPSS 20. The process of checking and analyzing the data sequentially follows: reliability test, EFA factor analysis, Correlation analysis and Multiple linear regression analysis. Initially, the scale was assessed by Cronbach's Alpha and EFA. Then, the correlation coefficient test was used to check the linear relationship between the independent and dependent variables. The results from the study indicated the accuracy, completeness, review quantity, review consistency, product/ service rating, and website reputation were accepted and all of them had positive effects on review credibility. This shows, shoppers are very interested in completeness, the full information as well as necessary evidence to help shoppers assess the truthfulness and accuracy of the information more easily. At the same time, the website reputation also helps increase the credibility of the reviews and buyers tend to believe in online reviews if the website has a good reputation. Basically, the research completed its original purpose of identifying factors that affect the credibility of online reviews. © 2020 by the authors; licensee Growing Science, Canada.

The purpose of this study was to test the determinants affecting the credibility of online reviews. In this model, the relationships between argument quality and peripheral cue and credibility of online reviews were evaluated. The measurement model and the conceptual model depicting hypothesized relationships were evaluated based on responses from 460 customers using multiple linear regression analysis, accordingly. After the data collection process was completed, all data were put into SPSS 20. The process of checking and analyzing the data sequentially follows: reliability test, EFA factor analysis, Correlation analysis and Multiple linear regression analysis. Initially, the scale was assessed by Cronbach's Alpha and EFA. Then, the correlation coefficient test was used to check the linear relationship between the independent and dependent variables. The results from the study indicated the accuracy, completeness, review quantity, review consistency, product/ service rating, and website reputation were accepted and all of them had positive effects on review credibility. This shows, shoppers are very interested in completeness, the full information as well as necessary evidence to help shoppers assess the truthfulness and accuracy of the information more easily. At the same time, the website reputation also helps increase the credibility of the reviews and buyers tend to believe in online reviews if the website has a good reputation. Basically, the research completed its original purpose of identifying factors that affect the credibility of online reviews.

Introduction
Today, marketers consider social media to promote their products. Social media has become a very important tool that enables marketers to communicate with their customers. After the significant development of the internet and online social media, people have started to share their opinions about products through online reviews. Reimer and Benkenstein (2016) indicated that the information generated by consumers and the information created by the marketer do not have a vested interest and were therefore more independent and reliable. According to Baek et al. (2015) credibility is the most important factor in electronic word-of-mouth adoption. For this reason, the credibility of online reviews seems crucial when making purchase decisions based on those reviews. Given the importance of credibility in the context of online reviews and associated purchase decisions, it stands to reason that review credibility is of great interest to both consumers and marketers. Today, as online reviews become more accessible to internet users, managers need to not only promote positive online information about their products but also need to do reduce the unwanted impact of negative online information on their brands/products. Therefore, the reputation and use of online review and review sites can be threatened in the long term, if consumer concerns and uncertainties continue to spread and consolidate. In order to counteract this development initiated by those rogue firms, it is important for reliable companies to understand how consumers perceive and assess the credibility of online reviews, and particularly, to know what factors determine review credibility from the consumers' point of view. However, up to now, there is very little research in Vietnam on this topic. Therefore, this research aims to find out factors affecting the credibility of online reviews on the e-commerce website TIKI. The paper aims to propose and tests the factors effect to test the determinants affecting the credibility of online reviews. The results of this paper show that there are six factors effect to credibility of online reviews.
The structure of this paper is as follows. Section 2 presents the theoretical background and a review of previous studies. Section 3 discusses the data and methodology. Section 4 summarizes the empirical results. Section 5 offers a discussion of results. The article ends with the sections outlining the research limitations and the conclusion 2. Literature review

Accuracy
According to Jamil and Hasnu (2013) accuracy primarily refers to the reliability and correctness of online reviews and represents one of its main influencing factors. Daft and Lengel (1986) showed that media richness theory, quality, accuracy, and reliability of the information exchanged are important across a medium. The more accurate the message, the higher the usefulness of the information received. Therefore, Accuracy plays an important role when consumers deliberately deal with the content of online reviews, particularly in connection with credibility.

Completeness
According to Luo et al. (2013), completeness refers to the degree to which an online review is comprehensive and provides satisfactory information. Cheung et al. (2009) found that online reviews have valid arguments, they can consider online reviews to be positive and reliable information. For example, a full online review of a hotel should include core aspects such as location, cleanliness, equipment, room rates, breakfast quality and choice, staff friendliness (Filieri & McLeay, 2013).

Timeliness
In the process of seeking information, consumers may encounter large amounts of relevant information that are related to a specific timestamp, leading to the concept of timeliness research. Timeliness relates to the novelty and updating of online reviews, thus reflecting the current state of the product or service (Jamil & Hasnu, 2013). Since online reviews are available at any time and are often one of the first sources of information on new products and services, they have a decisive advantage in time overprint media or traditional word of mouth (Filieri & McLeay, 2013). Madu and Madu (2002) have shown that a website needs to be constantly updated to provide value-added information to users. From a consumer's point of view, as time goes by, the average usefulness of reviews decreases.

Review quantity
This factor refers to the number of online reviews available for a specific product or service on the review website (Filieri & McLeay, 2013). The quantity of online reviews is usually much higher on review sites than in offline environments, because only review websites store online reviews and make them available to recipients (Zhang et al., 2014). A large quantity makes online reviews more observable (Cheung & Thadani, 2010) and contributes to verifying individual online reviews, constituting an important factor in the context of peripheral cues (Zhang et al., 2014).

Review consistency
According to Chang et al. (2015) review consistency relates to the extent to which information in a review is consistent with information in other reviews. By contrast, consumers are very likely to be more skeptical towards reviews that are not consistent with most of the other reviews. In other words, consumers consider a review as a credible if it is consistent with other reviews (Cheung et al., 2009. In addition, Cheung et al. (2012) show that recipients perceive those online reviews as more likely and useful that assess the same products or services similar over time. Godes and Silva (2012) note that the others opinion on reviews have significant influence on the credibility evaluation of online reviews. Fang (2014) shows that this factor describes the extent to which recipients perceive reviewers as competent with regard to the online review of products and services. Park and Nicolau (2015) show that it is a very important aspect to recipients of an online review, particularly when the information searched shall support them in the decision-making processes. Cheng and Ho (2015) confirm that recipients perceive online reviews written by experts as more useful and that such reviews have a greater impact on recipients' purchase intention than those of reviewers without expert knowledge. In this connection, the reviewers have the ability and qualifications to write useful online reviews that constitute their expertise, which reflects the amount of the reviewers' knowledge about the respective product or service.

Product or service rating
The rating of products and services is an assessment based on pictograms (Filieri & McLeay, 2013). In this way, online reviews on e-commerce websites, such as TIKI are complemented by a star rating. Star ratings of products or services can be considered as a general assessment of conclusions. Review websites aggregate all individual ratings of online reviews with regard to a product and service to an overall rating, thus allowing statements about the average rating of all online reviews. In this way, online reviews on review websites, the higher the number of stars in a star rating, the better the rating of the product or service.

Website reputation
According to Lin (2007), the main variables of website that compose this element are: Information Quality and Reputation/Brand. The reputation of a website depends heavily on its popularity and integrity. The more known and reputable a website is, the more people will accept it (Park & Lee, 2009b) and that the latter often rely on reputations when assessing the credibility of presented information (Chih et al., 2013). According to Chang and Chen (2008) Website image is defined instead as, the perceptions about a website name as reflected by the website associations held in customer memory. Website brand has in fact an impact on consumer's trust then on their purchase intentions (Fam et al., 2004) 2.9 Credible online reviews Cheung et al. (2009) empirically proves that if people perceive a particular review as credible then they are likely to persuade that particular review. According to Sen and Lerman (2007), online consumer reviews are a type of word of mouth involving positive or negative statements given to the products by the consumers in online shopping malls. Moreover, Viglia et al. (2016) showed that consumers often rely on outside sources such as customer reviews online due to limited direct information about the quality of service provided. In other words, credibility can be considered as the opinion of the people on the reality of the assessment (Erkan & Evans, 2016). Credibility of online reviews can be described as a process by which consumers assess the accuracy of online reviews. Consumers are likely to believe the reviews if they consider the source of the reviews as credible. Cheung et al. (2009) conceptualized accuracy within the central route of persuasion and confirmed it as a meaningful factor of argument quality, especially in the context of social media. Accuracy plays an important role when consumers deliberately deal with the content of online reviews, especially regarding credibility. According to Jamil and Hasnu (2013), if those aspects already known to the consumer are accurately represented in the online review, consumers will most likely also assume unknown aspects of online reviews are accurate. proposing the following hypothesis:

Hypotheses
Hypothesis 1 (H1): The accuracy of an online review positively influences its credibility. Cheung et al. (2009) found that online reviews have valid arguments, they are likely to regard online reviews to be positive and reliable information. Although this effect relationship has not been empirically confirmed so far, researchers urge the demand for further research on this issue (Luo et al., 2013). Therefore, investigates the influence of completeness on the credibility of online reviews as a dimension of argument quality, proposing the following hypothesis: Hypothesis 2 (H2): The completeness of an online review positively influences its credibility.
It is regarded as a major factor of information quality or, in this context, argument quality (Cheung et al., 2008). Abdulla et al. (2002) have shown that timeliness plays an important role in the credibility of the information, especially in the context of the Internet. Similar to the completeness, timeliness improves the quality of online reviews and thus also supports the strength of argument, thus potentially enabling a more favorable attitude to online reviews, including a more positive perception of credibility. Therefore, proposing the following hypothesis: Hypothesis 3 (H3): The timeliness of an online review positively influences its credibility.
Review quantity has been frequently applied as a peripheral cue and considered the influencing factor with regard to online reviews in the social media and e-commerce literature (Zhang et al., 2014). Previous research also provides empirical evidence that review quantity has a positive impact on the credibility of online reviews (Fan et al., 2013). Therefore, proposing the following hypothesis: Hypothesis 4 (H4): Review quantity positively influences the credibility of an online review.
According to Munzel (2016) review consistency has already been subject to studies in social media and e-commerce literature and confirmed as an important peripheral cue related to online reviews. When comparing online reviews, consumers are more likely to perceive those reviews as more credible that show a high consistency with most of the other related reviews. The previous study provides empirical evidence that review consistency has a positive impact on review credibility (Cheung et al., 2012;Luo et al., 2015). Therefore, proposing the following hypothesis: Hypothesis 5 (H5): Review consistency positively influences the credibility of an online review.
Reviewer expertise represents another element of online reviews that have been analyzed specifically in the context of social media research (Cheng & Ho, 2015;Fang, 2014). Previous research has shown that experts are more reliable than laypersons. The reviewers' expertise is very important for recipients, especially when they search for the information that is supposed to assist them in the decision-making process (Gilly et al., 1998). According to Cheung and Thadani (2012) the reviewers' expertise is not only an important aspect of their credibility but also have a positive impact on the credibility of online reviews. Therefore, proposing the following hypothesis: Hypothesis 6 (H6): Reviewer expertise positively influences the credibility of an online review. Park (2015) showed that the rating of products or services is another factor of online reviews that have been often examined within social media and e-commerce research. Moreover, Fang (2014) show that such rankings can affect the way consumers perceive the credibility of online reviews. Therefore, proposing the following hypothesis: Hypothesis 7 (H7): A product or service rating positively influences the credibility of an online review.
Website reputation is a well-established structure in e-commerce research and social media and describes the popularity of a review site among online review recipients (Lee & Shin, 2014). The more known and reputable a website is, the more people will accept it (Park & Lee, 2009). Chih et al. (2013) demonstrate that website reputation has a decisive influence on the credibility of online reviews.
Hypothesis 8 (H8): Website reputation positively influences the credibility of an online review. Online reviews on Tiki contain very detailed information about the reviewed products/services. Cheung et al., 2008 COM3 Online reviews on Tiki contain a variety of information about the reviewed products/services. Luo et al., 2013 COM4 Overall, online reviews on Tiki are comprehensive. Timeliness TIM1 Online reviews on Tiki are current. Somers et al., 2003 TIM2 Online reviews on Tiki are timely. Cheung et al., 2008 TIM3 Online reviews on Tiki are up to date. Cheung, 2014 Review quantity RQ1 There is a great number of reviews from different authors about many products/services on Tiki. Kang & Kim, 2006 RQ2 There is a variety of reviews about many products/services on Tiki. Zhang et al., 2014 RQ3 There is a multitude of information about many products/services on Tiki.
Filieri, 2015 RQ4 The number of online reviews about products/services on Tiki is high. Review consistency

RC1
Different online reviews about a product/service on Tiki are often consistent with each other in terms of content.

RC2
Different online reviews about a product/service on Tiki overlap to some extent with each other in terms of content.

RC3
There are a number of overlaps among different online reviews about a product/service on Tiki. Luo et al., 2015 Reviewer Expertise RE1 Reviewers of online reviews on Tiki seem to possess sufficient knowledge. Ohanian, 1990 RE2 Reviewers of online reviews on Tiki seem to have enough insights to make an assessment. Folse et al., 2013 RE3 Reviewers of online reviews on Tiki seem to be competent. Fang, 2014 Product/Service rating PSR1 The rating of products/services on Tiki by means of stars has narrowed down the number of alternative prod-

PSR2
The rating of products/services on Tiki by means of stars has allowed me to find products/services that satisfy my needs.

PSR3
The rating of products/services on Tiki by means of stars has allowed me to find well-rated products/services. Filieri 2015

PSR4
The rating of products/services on Tiki by means of stars has allowed to me get a quick overview of products/services. Website reputation WR1 The website of Tiki is very popular. Hsiao et al., 2010 WR2 The website of Tiki has a high reputation with good reason. Tams, 2012 WR3 The website of Tiki is known for its high reputation. Chih et al., 2013 WR4 The website of Tiki is trustworthy in my opinion Review credibility on TIKI RCRE1 I think online reviews on Tiki are credible.
Cheung & Thadani., 2012 RCRE2 I think online reviews on Tiki are factual.
Qiu et al., 2012 RCRE3 I think online reviews on Tiki are reliable.
Luo et al., 2013 RCRE4 I think online reviews on Tiki are trustworthy.

Research design
This study consists of two main steps: Step 1: Preliminary research Conducting qualitative research through discussion techniques with instructors and hand-to-hand discussions with subjects selected by the convenient method that young people have used online shopping services to evaluate and adjust the questionnaire. In addition, qualitative research is also aimed at considering the appropriate scale and adjusting the scale to suit the research topic.
Step 2: Official research This is done through quantitative methods. Aims to retest the research model and research hypotheses. The data is cleaned and processed on SPSS 20.0 software. These observed variables were assessed by two criteria: (i) assessing the reliability of the scale (Reliability Analysis) through Cronbach's alpha coefficient and (ii) explaining factor of EFA (Exploratory Factor Analysis). Observed variables that satisfy the conditions will be used for formal research (official scales). Then, conduct model tests and research hypotheses using linear regression models.

Data collection
The authors selected the research subjects who have been using the Tiki e-commerce site, used a convenient sampling method but still ensure the full range of gender, age, income and frequency purchase. There are two common methods of collecting samples: direct survey and online survey (via social networks, email, google form tool ...). For this study, in order to ensure high quality, reliable data, high response rate, the author sent the survey to friends, relatives and social relationships in Vietnam that they have been using TIKI. At the same time, the author also conducted the survey directly through going to universities and cafe shops to ask the subjects to conduct surveys.

Scale development
According to Likert (1932), it is a kind of question form that respondents rate the level of their agreement on statements or questions. Likert's scale has many different types of scales; among those, there are two scales which are the most common in quantitative analysis that are questions structured with five-point or seven-point response scale. Likert surveys are quick, efficient and inexpensive methods for data collection. Therefore, to be easier for collecting and having more accurate data, the questionnaire will be designed in the form of Likert 5-point-scale, which is a scale from 1 (Strongly Disagree) to 5 (Strongly Agree). Includes questions designed to measure independent variables including accuracy, completeness, timeliness, review quantity, review consistency, reviewer expertise, product/service rating and web reputation.

Sample Demographic characteristics of respondents
Demographic data are the characteristics of human population. In this study, demographic questions are used to determine gender, age, job monthly income of respondents so that we can find out and compare the differences in experience and opinions between subgroups. The sample collected consisted of 460 surveys. Of which, 11 votes were invalid due to omitting the answer. Therefore, the final exact number of valid samples selected was 449 and was imported into SPSS 20 software to perform quantitative analysis.

Reliability test
Cronbach's coefficient alpha provides an indication of the average correlation between all items that make up the scale. Testing the reliability of the scale through Cronbach's Alpha coefficient and the scale is accepted as Cronbach's alpha reliability coefficient greater than 0.7 (Nunnally, 1978). Elimination of observed variables have Item-Total correlation of less than 0.4 (Nurosis, 1994). In addition, Cronbach's alpha values for accuracy, completeness, timeliness, review quantity, review expertise, review consistency, service rating, website reputation and review credibility on Tiki are 0.797, 0.729, 0.820, 0.839, 0.790, 0.833,0.779, 0.720 and 0.829 respectively, Cronbach's Alpha coefficient > 0.7 and the Corrected Item-Total Correlation > 0.4. Therefore, the scale of subjective variables is reliable.

Exploratory Factor Analysis -EFA
Factor analysis is a statistical method used to describe variability between observed, correlated variables in terms of a potentially lower number of unobserved variables called factors (Hair et al., 2010). The step of factor analysis of independent variables is recap in one round because KMO and Bartlett's Test, Total Variance Explained are satisfied as the theoretical analyses. The results of factor analysis show that the KMO index is 0.781 > 0.5, which proves that the data used for factor analysis is perfectly appropriate. Bartlett's test result is 528 with Sig = 0.000 < 0.05. Therefore, reject the hypothesis H0: the observed variables have no correlation with each other in the whole. So, variables are correlated with each other and satisfy factor analysis conditions. Besides, the results showed that 33 observed variables were grouped into 9 groups. Value of the total variance extracted = 66,350 > 50%: satisfactory; then it can be said that these 9 factors explain 66,350% of the data variability. Factor loadings are greater than 0.5, and there is no case in which the upload of both factors at the same time has a close load factor. So, factors that ensure the convergence and differentiation when analyzing EFA. In addition, there is no disturbance of factors, meaning that one question cannot be confused with another. So, after factor analysis, these independent factors remain the same, without being added or reduced.

Regression Analysis
The coefficient of determination (R-squared) is a statistical metric that is used to measure how much of the variation in outcome can be explained by the variation in the independent variables. R2 always increases as more predictors are added to the MLR model even though the predictors may not be related to the outcome variable. R2 by itself can't thus be used to identify which predictors should be included in a model and which should be excluded. R2 can only be between 0 and 1, where 0 indicates that the outcome cannot be predicted by any of the independent variables and 1 indicates that the outcome can be predicted without error from the independent variables. Adjusted R-square shows 49.3% of the complete variance of the dependent variable (Review Credibility) that can be interpreted by eight independent factors (Accuracy, Completeness, Timeliness, Review Quantity, Review Consistency, Reviewer Expertise, Product/Service Rating, and Website Reputation). The remaining is 50.7% are due to factors outside the model and random errors. For n = 33, k = 8, look up Durbin-Watson table, dL = 0.757, dU = 1,874. Attached to the DW value bar, 0.757 < 1.802 < 1.874. Thus, there is no superlative sequence correlation in the model. Table 2 shows R Square value at 0.717 above 0.7 and below 0.9. According to Guilford's (1973) this means regression analysis results have high correlations which means the five independent variables explain 71.7% of variations of the de-pendent variables -Intention to purchase. The ANOVA table given in Table 3 refers to the applied regression model's statistical significance.
To check the understanding of the result, according to Table 3 which present the results of ANOVA test, the F value of 53.490 shows the model is significant (at an observed sig. = 0.000). Hence, it can be cased that the miniature of this research has gained statistical understanding.  The results of the path coefficients and t-values were itemized as outlined in Table 4. Whereby accuracy positioning is seen to have a significant and positive link with review credibility, which is well within expectations. Beta standardized coefficients, Beta coefficient of accuracy variable is 0.147 with accreditation t = 4.080, sig = 0.000 < 0.05. Hence, H1 is therefore supported. In a similar vein, completeness had a significant influence on review credibility. Beta coefficient of completeness variable is 0.378 with accreditation t = 9.622, sig = 0.000 < 0.05, inferring that H2 is also retained. Further examination of the path coefficient shows that timeliness is not significant to review credibility. Beta coefficient of timeliness variable is -0.047 with accreditation t = -1.357, sig = 0.174 > 0.05, timeliness variable is excluded from the model. Thus, H3 is not supported. Besides, H4 examines whether or not review quality has a significant effect on review credibility. Beta standardized coefficients are reported as beta coefficient of review quality variable is 0.097 with accreditation t = 2.807, sig = 0.005 < 0.05, review quality variable is statically significant. Thus, H4 is maintained. H5, whereby review consistency positioning is seen to have a significant and positive link with review credibility, which is well within expectations. Beta standardized coefficients, (Beta coefficient of review consistency variable is 0.077 with accreditation t = 2.140, sig = 0.033 < 0.05). Hence, H5 is therefore supported. Beta coefficient of review expertise variable is -0.017 with accreditation t = -0.506, sig = 0.613 > 0.05, review expertise variable is excluded from the model. H6 is not supported. Besides, H7 examines whether or not product/service has a significant effect on review credibility. Beta coefficient of product/service variable is 0.177 with accreditation t = 4.839, sig = 0.000 < 0.05, product/service variable is statically significant. Thus, H7 is maintained. Finally, whereby web reputation positioning is seen to have a significant and positive link with review credibility, which is well within expectations. Beta coefficient of web reputation variable is 0.243 with accreditation t = 6.075, sig = 0.000 < 0.05, web reputation variable is statically significant. It is clear that Accuracy, Completeness, Review Quantity, Review Consistency, Product/Service Rating, and Website Reputation variables are making significant matchless donation to the present of the dependent variable (review credibility on Tiki).

Table 5
Hypotheses testing result

Discussions
Due to increasing consumer skepticism about the credibility of online reviews and less knowledge of consumer cognitive determinants, the purpose of this study is to identify and test the determinants of online review credibility and thereby build the basis for behavioral purchase research. So basically, the research has completed the initial goal. The research was conducted based on the application of the research model of Thomas et al. (2019). For the previous study, all 8 factors mentioned were significant in influencing the reputation of online reviews. However, the results in this study only have 6 factors: completeness (0.378), website reputation (0.243), product / service rating (0.177), accuracy (0.147), review quantity (0.097), and review consistency (0.077). And the model only explained 49.3%. This may stem from differences in the research context as well as the subject of the survey, so more testing is needed. According to this, consumers perceive a higher number of online reviews for a certain product or service as less credible in contrast to the study results and also contrary to the initial hypothesized direction of effect and prior research findings (Fan et al., 2013). At the same time, timeliness also has a negligible influence. An interesting result that there is some kind of spillover effect between website reputation and reviews credibility. In the previous study (Thomas et al., 2019), website reputation had the strongest impact on the credibility of online reviews and at the same time, in this research, this factor was also the second in its influence on online reviews credibility. According to which consumers more likely to trust an online review when they trust the respective website. These findings are most widely consistent with those of prior studies examining these factors in the context of online reviews (Chih et al., 2013;Fang, 2014).

Conclusion
The study has identified factors that affect the reputation of online reviews on e-commerce website TIKI including: Accuracy, Completeness, Review Quantity, Review Consistency, Product / Service Rating, and Website Reputation. The two factors that have the most influence on the reliability of online assessments are completeness (0.378) and website reputation (0.243), which show that these factors are especially relevant in determining the credibility of the online reviews. In addition, product/service rating (0.177), accuracy (0.147), review quantity (0.097) and review consistency (0.077) are also important determinants of review credibility. In the research, there are two factors that are eliminated: timeliness and reviewer expertise. For reviewer expertise, the reason may come from the fact that TIKI is an e-commerce site, not a pure website that provides online reviews. User accounts, as well as the reviews submitted, do not represent the assessor's expertise. As for timeliness, this factor should be further tested, although timeliness has been shown to play an important role in the credibility of information in the internet context (Abdulla et al., 2002). However, it had not found a significant effect in this research. Besides, the explanatory power of the model is not high at only 49.3%. The cause for this may stem from the problem of research data or because the research model has many points that are not suitable to the research environment and context.

Managerial implications
To begin with, marketing managers should consider the importance of factors of argument quality, in particular, the completeness of online reviews. For instance, marketing managers could offer customers of a certain product or service an incentive for writing an online review that meets certain requirements regarding accuracy and completeness. Moreover, they can implement a monitoring system that can detect inaccurate or incomplete assessments and respond to them, for instance, by posting a comment to the respective review that serves as a corrective and provides the accurate or complete information. That means that person actually purchased the item online. In this way, online reviews become more credibility. In addition, it is also advisable to regularly check negative reviews and quick responses to protect the brand when judgments are false or deliberately badly played by competitors. According to this, marketing managers should also especially be aware of the dominant role of peripheral cues. In particular, they should promote website reputation when it has a great impact on online reviews credibility. In addition, the more prominent use or display of product/service ratings seems to promise to increase review credibility among consumers. The most groundbreaking significance of this study refers to the conflicting impact of the number of reviews. Accordingly, marketing managers should note that too many reviews can be counterproductive and cause doubts for consumers. Once a large number of critical reviews have been achieved, they no longer actively encourage consumers to contribute online reviews, but must ensure a sufficient level of quality, especially in terms of accuracy and completeness of online reviews.

Limitations
This study is subject to some important limitations to consider, but at the same time can represent promising starting points for future research. This pertains particularly to those findings that conflict with previous research. Specifically, the elimination of timeliness and reviewer expertise as well as factors only explained 48.4% of the model. In addition, this study focuses solely on users and online reviews on the e-commerce platform TIKI and therefore may not represent other review platforms. Perhaps, there are differences with regard to the importance of influencing factors of online reviews. In particular, since TIKI is primarily focused on trading, the consumer assessment issues are less encouraged. Accordingly, the recipient may consider more of the other influencing factors of online reviews. This topic has not yet been thoroughly researched and requires further testing, particularly from a provider perspective and perhaps also a more technical perspective when it comes to forgery and fraud detection techniques. Because our research can provide insight into the factors that promote credibility from a consumer perspective and therefore, the factors towards which these measures should be targeted from the consumer's point of view. Besides that the age group of the research subject is limited and the study was conducted only Vietnam. Therefore, it needs to be expanded to get a more comprehensive perspective.

Future study
First of all, as mentioned above, this research purpose becomes the premise for the study of buying behavior based on the reference of online reviews and look for other factors. That is a future research direction. In addition, future research can address the opposite of the impact of review quantity. Some other interesting research directions such as could include moderators considering more strongly typical aspects of social networking websites. In this connection, it would be interesting to investigate how the reviewer's identity influences recipients. For instance, future studies could examine the impact of usefulness of online reviews on recipients' electronic word-of-mouth behavior. Finally, the generalization of the results is limited by the research context of Vietnam and the website is the e-commerce website TIKI. Therefore, future research may investigate characteristics across different cultural contexts and assessment platforms.