Antecedents of Generation Y consumers ’ usage frequency of online consumer reviews Antecedents of frequency of

Purpose – The purpose of this paper is to examine the in ﬂ uence of perceived information andentertainment value, perceived credibility and perceived value on Generation Y consumers ’ usage frequency of online consumer reviews. Design/methodology/approach – The paper proposes and tests, with structural equation modelling analysis ofmoment structures, a research model usingdata from a large sample of Generation Y consumers. Findings – The results con ﬁ rm that Generation Y consumers perceive online reviews to be informative, entertaining, credible and valuable, and that they frequently consult such reviews. More speci ﬁ cally, the empirical analysis con ﬁ rms that perceived information value, perceived entertainment value and perceived credibility signi ﬁ cantly in ﬂ uenced the perceived value that Generation Y attach to online consumer reviews, which,inturn,wasa signi ﬁ cant predictor oftheir usage frequency of such reviews. Practical implications – The results highlight the strategic importance of integrating online consumer reviews into the marketing communication mix when targeting Generation Y, together with the necessity of having ﬁ ltering mechanisms to ensure that only authentic reviews are published and the need to implement tactics to ensure that such reviewsareinformative andentertaining and, consequently,of value. Originality/value – This study contributes to marketers ’ comprehension of strategically using online consumer reviewswhen targetingtheGeneration Y segment.


Introduction
The purpose of this paper is to ascertain the influence of perceived information and entertainment value, perceived credibility and perceived value on Generation Y consumers' usage frequency of online consumer reviews. As a particular form of electronic word-ofmouth communication, online consumer reviews are becoming popular amongst an increasing number of consumers as a way of assessing the quality and performance of products and services across a range of industries prior to purchase (Filieria et al., 2018). Word-of-mouth communication or vica voce, has been around, as the dawn of human existence and refers to the informal exchange of information between individuals or groups of individuals. Consumption-related activities are a frequent topic in such social communication exchanges and this type of vica voce refers to consumers informally giving or seeking advice amongst each other concerning consumption-related behaviour (Flynn et al., 1996), leading to the behavioural consequence of product/service information sharing (Sun et al., 2006). In the fields of marketing and consumer behaviour, vica voce takes on a special relevance as it has a powerful influence on consumers' purchase behaviour in that it is the most persuasive form consumption-related communication that exists (Schiffman et al., 2010;Kotler, 2003).
In the pre-digital age, the spread of consumption-related vica voce was slow, confined to two or a few individuals (Cheung and Lee, 2012), occurred in a spontaneous fashion and lacked longevity (Stern, 1994). However, the commercialisation of the internet in 1991 and the rapid advances in the World Wide Web changed the fundamental characteristics of word-of-mouth communication. These digital technologies have resulted in electronic consumption-related vica voce, including online consumer reviews, having significantly greater reach (Kucukemiroglu and Kara, 2015) by extending consumers' options of gathering and disseminating unbiased consumption-related experiences amongst each other on a global scale and across time zones (Park and Lee, 2009). As a manifestation of digital word-of-mouth communication, online consumer reviews (Filieria et al., 2018) can be dispersed at an accelerated velocity, have longevity beyond a single conversation as they can be archived indefinitely and are accessible 24/7 to anyone with an internet connection, regardless of their geographical location (Cheung and Lee, 2012;Park and Lee, 2009).
Debuting in the 1990s, with Amazon.com (Ante, 2009), rateitall.com, deja.com and Epinions.com (Notess, 2000) being amongst the first, online consumer review sites have grown exponentially (Filieria et al., 2018). Today, such reviews encompass an ever-growing range of industries (Wang et al., 2018) and have become relevant to organisations regardless of size (Bowman, 2019). The abundance of online consumer-generated reviews available for a diverse range of products, services and organisations, coupled with their accessibility has contributed to their significant and growing popularity , and accounts for them fast becoming the most important source of consumption-related information for many contemporary consumers (Bowman, 2019;Kim et al., 2018;. This is particularly true amongst the Youth (Fertik, 2019;Hall, 2018;Kats, 2018), who are currently classified as being members of Generation Y (individuals born between 1986 and 2005) (Markert, 2004).
As the first generation to grow up in the digital age of internet connectivity, mobile telephony and, for the later cohort of Generation Y, social media (Taylor and Keeter, 2010), these technologically-astute individuals harness the power that their digital devices give them as consumers and regularly share their consumption-related experiences across digital platforms, thereby relying on each other to make informed purchase decisions (Gailewicz, 2014). A study conducted in Portugal concluded that, in comparison to the previous generation, Generation Y individuals are more likely to read consumer-generated product reviews and comments on company Facebook pages (Bento et al., 2018). In a 2016 study across 50 countries, KPMG (2017) confirms that Generation Y individuals have a marked tendency towards consulting online consumer reviews while shopping. A study in the UK found that eight out of ten members of Generation Y never make a purchase without first reading online consumer reviews (Hall, 2018). Another study in the USA indicates that 37.3% of Generation Y individuals almost always consult online consumer reviews prior to purchase and 42.9% often consult such reviews (Kats, 2018). A more recent survey in the USA suggests that 50% of Generation Y always read online consumer-generated reviews of local businesses prior to purchasing and that 91% of them trust those reviews as much as they do personal recommendations (Murphy, 2018). A similar picture emerges in Africa, with reports indicating that online consumer reviews are Generation Y individuals' most trusted source of consumption-related information (Zinkevych, 2018).
This generation's propensity to consult and contribute to online consumer reviews (Fertik, 2019;Hall, 2018;Kats, 2018;Zinkevych, 2018) highlights the value of marketers integrating such reviews into their marketing communication strategy, especially when targeting Generation Y. While there is an abundance of research on online consumer reviews (Srivastava and Kalro, 2019;Filieria et al., 2018;Elwalda et al., 2016;Liu and Park, 2015;Mudambi and Schuff, 2010;Park and Lee, 2009), few studies seek to understand a specific generation's usage frequency of such reviews in terms of the uses and gratifications (U&G) theory. The benefit of the U&G theory is that it takes a consumer-level approach in determining media usage (Lin et al., 2017). Specifically, this study applies the advertising value model developed by Ducoffe (1996), which is derived from the U&G theory and extends it to include perceived credibility to model the antecedents of Generation Y consumers' usage frequency of online consumer reviews. The strategic success of online consumer reviews as part of a marketing communication strategy targeting generation Y consumers depends on their perceived value and usage frequency of such reviews, which, in turn, are dependent on certain factors that influence that perceived value. Understanding the factors that influence Generation Y consumers' perceived value and Antecedents of frequency of reviews usage frequency of online consumer reviews will contribute to marketers' ability to leverage such reviews to their advantage. In light of this, this study focussed on determining Generation Y university students' usage frequency of online consumer reviews and the extent to which they perceive such reviews as being informative, entertaining, credible and of value. In addition, the study sought to determine the influence of perceived information and entertainment value, perceived credibility and perceived value on Generation Y students' usage frequency of online consumer reviews. The focus on university students as the study's target population is deliberate and based on the assumption that a tertiary qualification is typically associated with a higher future earning potential and a higher social standing, thereby rendering graduates as trend setters and opinion leaders amongst their peers (Bevan-Dye and Akpojivi, 2016).

Theoretical background
Online consumer reviews refer to consumer-generated product or service evaluations that are electronically uploaded to a company or third-party website (Mudambi and Schuff, 2010). Online consumer review content may be in the form of text, images, videos or a combination thereof, and ranges from informal discussions about products, services and/or organisations to more structured reviews in text or on video (Berthon et al., 2012). Examples of websites that publish more structured consumer reviews include platform-specific sites such as product review videos on YouTube.com, general reviews on Google.com, travel reviews on industry-specific sites such as on TripAdvisor.com and restaurant reviews on Zomato.com. Online consumer reviews are also available on retail store-specific sites such as Amazon.com, product category-specific sites such as fragrance reviews on Fragrantica.com and service category-specific review sites such as medical practitioner reviews on RateMDs. com. Generally, the text-based online consumer reviews take the format of open-ended evaluations of a product, service or organisation, together with a numerical star rating that usually ranges from one to five stars (Mudambi and Schuff, 2010).
The usage frequency of consumption-related communication, whether formal or informal, is generally dependent on the extent to which such communication is perceived to be of value in consumption-related decision-making (Park and Lee, 2009). This perceived value refers to the overall representation of the worth of consumption-related communication (Zeng et al., 2009;Ducoffe, 1996), where consumption-related communication that is perceived as being of value is associated with a positive attitude towards that communication (Bakr et al., 2019), and therefore, the active use of that communication (Bailey et al., 2018;Park and Lee, 2009). A potentially useful model to help understand the usage frequency of online consumer reviews is the Ducoffe (1996) advertising value model. This model is derived from the U&G theory that seeks to explain individuals' media consumption in terms of why they use certain media and the gratifications they derive from that consumption (Katz et al., 1973). The Ducoffe model has been used to explain consumers' perceived value of web advertising (Zha et al., 2015;Brackett and Carr, 2001;Ducoffe, 1996), email advertising (Chang et al., 2013), mobile advertising (Martins et al., 2019;Kim and Han, 2014;Tsang et al., 2004) and social media advertising (Shareef et al., 2019;Saxena and Khanna, 2013). Ducoffe (1995) explains the perceived value of consumption-related communication as being consumers' subjective evaluation of the relative worth or utility of that communication. The U&G theory suggests that communication is goal-directed behaviour and that individuals' choice to engage with a particular media, as well as their level of participation with that media depends on the degree to which it satisfies their needs; that is, their perceived value of that media (Rubin, 1993).
In accordance with the U&G theory, Ducoffe (1996) identifies perceived informativeness and entertainment as salient factors that account for how consumers appraise the value of consumption-related communication. More recent studies confirm the positive influence that perceived information and entertainment value have on the overall perceived value of online advertisements (Saxena and Khanna, 2013;Wang and Sun, 2010), mobile advertising (Wang and Genç, 2019), advertising on social networking sites (Hassan et al., 2013) and electronic word-of-mouth communication (Lien and Cao, 2014). In replicating the Ducoffe (1996) study, Brackett and Carr (2001) added another salient predictor of advertising value, namely, perceived credibility, which, in addition to perceived informativeness and entertainment, they found to be a significant positive antecedent of the perceived value of web advertisements. Subsequent studies have confirmed the role of credibility in accounting for a target audience's assessment of the value of advertisements (Martins et al., 2019;Zha et al., 2015;Kim and Han, 2014). Ducoffe (1996) explains the perceived information value of consumption-related information as being the completeness, currency and quality of the information provided. In terms of online consumer reviews, the quality of the information provided relates to the cogency of the review content, where rational, concrete and objective reviews are perceived as being more informative than abstract ones of an emotional nature (Park and Lee, 2009). This infers that reviews that indicate both the pros and cons of the consumption-related experience will be viewed as being more complete, and thus, of a better quality (Yang et al., 2016). Flavi an et al. (2009) add that the format in which the information is presented on websites also influences the perceived quality and usability thereof. To improve the cogency, and hence, the quality of the information offered up in the review,  suggest providing a review template that has demarcated sections requesting specific usage experience, and a section encouraging reviewers to mention not only their positive experience with the product or service under review but also to mention any negative experiences or views.

Perceived information value
Typically, consumers do not perceive reviews older than three months as being particularly relevant; something that highlights the importance of marketers encouraging a constant stream of up-to-date reviews (Rose, 2017). Possibly the most intuitive way of encouraging customers to write reviews is to make it easy for them to do so. This entails providing easy to locate review buttons on the organisation's site that direct the customer to the review section or to a third-party review site (Millwood, 2018;Akalp, 2014). Alternatively, set up a system that directs the customer to the review section/site immediately after purchase or a system whereby an automatic email gets sent out after a purchase inviting the customer to provide a review (Greene, 2019). Regardless of the system, it is essential that these review features are easy to use on either a computer or a mobile device (Millwood, 2018;Akalp, 2014). While care should be taken not to create the impression of buying positive reviews, it is acceptable to ask a customer to leave an authentic review in exchange for coupons or discount codes (Hawlk, 2017). Rather than financial incentives, sites such as TripAdvisor.com incentivise reviews using a badge system, where a greater number of reviews by an individual earns a higher badge status.

Perceived entertainment value
According to the U&G theory, perceived enjoyment/entertainment refers to the intrinsic or hedonic motive of consuming media content (Luo et al., 2011). In terms of online consumer reviews, the dimension of perceived enjoyment relates to the pleasure derived from perusing Antecedents of frequency of reviews such content (Elwalda et al., 2016). Mathwick and Rigdon (2004) posit that for many consumers, online product information search is a leisure activity, associated with the intrinsic benefits of pleasure and enjoyment. Flavi an-Blanco et al. (2011) indicate that this affective state of enjoyment is heightened when an online product information search is perceived as being successful in terms of being pertinent to the consumption-related decision-making process. An important aspect of the perceived enjoyment of reading online consumer reviews is the readability of such reviews, where understandable rather than complicated and excessively long reviews are viewed as being more pleasurable (Liu and Park, 2015). Grammar and spelling errors have empirically been proven to hamper the readability of online consumer reviews (Ghose and Ipeirotis, 2011), thereby detracting from their reading enjoyment. Furthermore, lengthy and overly complicated reviews have also been found to influence the enjoyment of reading such reviews negatively (Park and Lee, 2009). This suggests that it is advisable for marketers to put a word-count limit on online consumer reviews to prevent information overload (Kim et al., 2018) and to provide a spelland-grammar check tool on the review site.

Perceived credibility
Despite not having being part of Ducoffe's (1996) original model, the perceived credibility of consumption-related communication is possibly the most important predictor of its value and usage frequency. The persuasiveness of consumption-related vica voce is rooted in its inherent credibility; that is, its quality of being both trustworthy and believable (Cheung et al., 2009;Bone, 1995). In a seminal qualitative study, Dichter (1966) uncovered two primary factors that account for the perceived credibility of consumption-related word-ofmouth communication, namely, the opinion receiver perceiving the opinion giver as being motivated by a genuine desire to help rather than by any material gain, and the opinion receiver perceiving the opinion giver as having a certain level of experience with and knowledge of the product category being recommended.
In the case of online consumer reviews, this perceived credibility becomes even more important given that such reviews are typically provided by anonymous consumers (Park and Lee, 2009). Unfortunately, the illegal practice of organisations' posting fake positive reviews to promote their own product/service/organisation or fake negative reviews aimed at sullying their competitors' reputations (Elliott, 2018;Zhang et al., 2016), together with the legal but unethical practice of only publishing positive reviews have brought the authenticity of online consumer reviews into question (Kim et al., 2018;O'Neil, 2015). As a case in point, Fakespot.com, an artificial intelligence platform that analyses online reviews, concluded that up to 70% of the reviews on Amazon.com are fake (Elliott, 2018). There are strong motives that drive sellers to engage in the practice of posting fake reviews and/or only posting reviews that show their goods in a positive light. Basically, the incentive for sellers to cheat by manipulating consumer reviews is high, while the probability of getting caught doing so is low (Vedantam, 2012).
Owing to website algorithms, products and pages with positive reviews show up at the top of search results, while those with negative reviews get pushed further down the search results. Furthermore, positive reviews have a powerful influence on consumers' purchasing behaviour and are an important part of the formula in driving sales (Heinzman, 2019). Sadly, unless consumer review sites take swift action to root out these illegal and unethical practices, the very thing that makes online consumer reviews so persuasive and valuable is likely to be the very thing that will lead to their demise and render them obsolete. Possible actions include only allowing identified customer to post a review immediately after a purchase, using software such as that used by Fakespot.com to identify perpetrators, releasing the names of consumer review sites that only post positive reviews to the media and for third-party review sites to impose harsh penalties on any of their vendors caught engaging in such behaviour.
Moreover, when confronted by negative reviews, rather than hide them, organisations are encouraged to offer a timely response, whereby they acknowledge the complaint, express empathy with the complainant, thank the reviewer and, where possible, describe the actions to be taken to rectify the problemthis will enhance the credibility of the organisation (Ho, 2017).
To this end, whether the reviews are positive or negative, it is advisable to have a dedicated reputation marketer(s), whose sole focus is on monitoring reviews and responding accordingly, whether it is to thank someone for a positive review or to address any issues in a negative review (Pitman, 2019).
In accordance with the literature, this study theorises that Generation Y students' perceived information and entertainment value, together with their perceived credibility of online consumer reviews contributes to their overall perceived value and usage frequency of such reviews.

Methodology
A descriptive research design, using the single cross-sectional sampling approach informed the research methodology applied in this study.

Sampling and data collection
As per the primary objective of the study, the target population was defined as Generation Y university students registered at South African public higher education institutions (HEIs), who were years of age between 18 and 24 years. The sampling frame was limited to HEI campuses in South Africa's Gauteng province and judgement sampling was used to select one campus from a traditional university, one from a university of technology and one from a comprehensive university, thereby ensuring that the sample comprised participants from each of South Africa's three types of public HEIs. The choice of these specific institutions' campuses was based on previous research findings that indicate that these three HEIs typically attract students from across South Africa's nine provinces (Synodinos, 2017;Bevan-Dye and Akpojivi, 2016). Data collection took place at the three selected universities' campuses. Using the mall-intercept survey approach, fieldworkers distributed 600 questionnaires across the selected three campuses (200 per campus) to a convenience sample of students.

Research instrument
The self-reporting survey questionnaire used to gather the required data included a cover letter describing the aim of the study and providing a guarantee of the anonymity of the sample participants. This was followed by a section requesting demographic information and a section containing scaled-response items that were adapted from published studies.
Perceived information value included the seven items of online consumer reviews "are a good source of product information", "supply relevant product information", "provide timely product information", "are a good source of up-to-date product information", "make product information immediately accessible", "are a convenient source of product information" and "supply complete product information". Perceived entertainment value consisted of the five items of online consumer reviews are "entertaining", "enjoyable", "pleasing", "fun to use" and "exciting". Perceived value was measured using the three items of online consumer Antecedents of frequency of reviews reviews are "useful", "valuable" and "important". These three scales were adapted from the scales developed by Ducoffe (1996). The perceived credibility of online reviews was measured using a scale developed by Ohanian (1990) and comprised five items, namely online consumer reviews are "dependable", "honest", "reliable", "sincere" and "trustworthy".
Usage frequency was measured using a scale developed by Park and Lee (2009), which included the three items of "I read online consumer reviews frequently", "I often search consumer reviews on the internet" and "I refer to online consumer reviews whenever I need information on companies or goods". A six-point Likert-type scale, ranging from strongly disagree (1) to strongly agree (6) was used to measure responses to these 23 scaled items.

Data analysis
The gathered data were analysed using of IBM's Statistical Package for Social Sciences and analysis of moment structures (AMOS), Versions 26. Data analysis included descriptive statistics and a one-sample t-test, collinearity diagnostics, structural equation modelling, with confirmatory factor analysis, including internal-consistency and composite reliability analysis, nomological, convergent and discriminant validity analysis and path analysis.
As a starting point, the descriptive statistics, including the means and standard deviations were computed, along with a one-sample t-test. Given that scaled-responses were recorded on a six-point Likert-type scale, the expected mean for the t-test was set at 3.5.
Next, to assess nomological validity, a matrix of Pearson's product-moment correlation coefficients was calculated. Nomological validity is inferred when there are statistically significant relationships in the theoretically correct direction between pairs of latent factors proposed for inclusion in a model (Hair et al., 2010). As excessively high correlation coefficients between latent factors may cause problems in interpreting the results of multivariate statistical analysis methods, a screening for multi-collinearity was undertaken by computing the tolerance values (TV) and the variance inflation factors (VIF). TV less than 0.10 and an average VIF above 10 are warning indications of multi-collinearity (Pallant, 2010).
Following this, SEM, including confirmatory factor analysis and path analysis was carried out using the maximum likelihood method. For the confirmatory factor analysis, a five-factor measurement model was specified for testing. As per convention, the first loading on each of the five latent factors was fixed at 1.0, which resulted in 276 distinct sample moments and 56 distinct parameters to be estimated, which equates to 220 degrees of freedom (df) based on an over-identified model and a x 2 value of 539.274, with a probability level equal to 0.000. While a statistically significant x 2 is indicative of poor model fit, the literature indicates that this statistic is vulnerable to large sample sizes, which is why the use of other model fit induces is advised (Byrne, 2010). In this study, the model fit indices computed included the goodness-of-fit index (GFI), the incremental-fit index (IFI), the Tucker-Lewis index (TLI), the standardised root mean square residual (SRMR) and the root mean square error of approximation (RMSEA), where GFI, IFI and TLI values above 0.90, together with SRMR and RMSEA values below 0.08 are indicative of acceptable model fit (Malhotra, 2010).
Composite reliability (CR) and average variance extracted (AVE) were computed using the formulae advocated by Fornell and Larcker (1981). Standardised loading estimates and an AVE value of 0.50 or higher, together with a CR value of 0.70 or above are required to conclude convergent validity for a latent factor. Discriminant validity necessitates that the square root of the AVE (HAVE) value of the latent factor exceeds the correlation estimates between the relevant latent factors (Hair et al., 2010;Fornell and Larcker, 1981).
An additional measure of discriminant validity proposed by Henseler et al. (2015), namely, the Heterotrait-Monotrait ratio of correlations (HTMT), was also used, whereby a HTMT value below 0.85 is recommended. While HTMT was originally intended as a criterion for assessing discriminant validity in variance-based SEM, its superior performance and the fact that its calculation is reliant on inter-item correlations rather than model estimates has led to some advocating its universal use across latent variable methods (Franke and Sarstedt, 2019;Voorhees et al., 2016). Reliability was tested using Cronbach's alpha (a) and CR, where values of 0.70 and above are indicative of acceptable reliability (Malhotra, 2010). A structural model was then specified based on the validated measurement model to test the theorised paths that Generation Y students' perceived information and entertainment value, and perceived credibility of online consumer reviews have a direct positive influence on their perceived value of such reviews, which, in turn, was theorised to have a direct positive influence on their usage frequency of such reviews. The level of statistical significance was set at p # 0.01 throughout.

Results
Following data collection, 538 complete questionnaires were received back of the 600 distributed (90 % response rate). This sample size of 538 was deemed adequate for conducting SEM, given that the proposed measurement model comprised five constructs, each with three or more observed variables (Malhotra, 2010), that there was a 23:1 ratio of cases to observed variables, a 9:1 ratio of cases to estimated parameters (Ullman, 2014) and 220 df (MacCallum et al., 1996). A description of the sample participants is summarised in Table 1.
Descriptive statistics, together with a one sample t-test, were computed to assess the extent to which Generation Y students perceive online consumer reviews as being informative, entertaining, credible and valuable, as well as their usage frequency of such reviews. The means, standard deviations, t-values and p-values for the five latent factors are reported in Table 2.
As indicated in Table 2, the means of the responses recorded on the six-point Likert-type scale were all significantly (p # 0.01) above the expected mean of 3.5, thereby suggesting that Generation Y students perceive online consumer reviews favourably. The highest means were returned for perceived value (mean = 4.47), perceived information value (mean = 4.33) and perceived entertainment value (mean = 4.17), which infers that Generation Y students view online consumer reviews as important, useful and valuable, as a convenient and good source of relevant, timely and up-to-date information and as pleasing and fun to use. While still significantly in the agreement area of the scale, lower means were computed on usage frequency (mean = 3.87) and perceived credibility (mean = 3.61), which is somewhat surprising given the high mean recorded for perceived value. Marketers need to remain vigilant regarding the authenticity of their online consumer reviews, as this will affect both Generation Y members' perceived value and usage frequency of such reviews. Moreover, the statistically significant means recorded on all of the latent factors emphasises the importance of marketers integrating online consumer reviews into their marketing communication strategy when targeting Generation Y consumers.
Prior to conducting confirmatory factor analysis on the proposed model, pairwise correlation analysis was carried out to assess the nomological validity of the latent variable planned for inclusion in the model. Thereafter, collinearity diagnostics were run to ascertain if the proposed model had any serious multi-collinearity issues. The correlation matrix and collinearity diagnostic results are reported in Table 3. As is shown in Table 3, the correlation coefficients between each of the pairs of latent factors proposed for inclusion in the model were in the correct direction and were statistically significant (p # 0.01), which suggests nomological validity. Concerning the collinearity diagnostics run, with tolerance values ranging from 0.559 to 0.766 and an average VIF 1.486 there were no multi-collinearity issues worth noting. Following on from this, confirmatory factor analysis of the measurement model was run using AMOS. Table 4 reports on the estimates for the measurement model, including the standardised loading estimates and their corresponding error variance estimates (R 2 ), Cronbach's alpha (a), CR, AVE and the HAVE values.

Antecedents of frequency of reviews
The results outlined in Table 4 indicate that the Cronbach's alpha (a) and CR values for each of the latent factors exceed 0.70, thereby indicating both internal-consistency and CR. Furthermore, with standardised loading estimates all exceeding 0.50 and CR values exceeding 0.70 there are indications of convergent validity. With the exception of one latent factor, namely, perceived information value, the AVE values exceed 0.50, thereby providing further evidence of the convergent validity of four of the five latent factors. Similarly, discriminant validity is evident for all but one latent factor in Table 4, in that the HAVE values for four of the five latent factors exceed their respective correlation coefficients. The low AVE of 0.34 and HAVE of 0.58 computed for the perceived information value latent factor was of serious concern for both convergent and discriminant validity. In an effort to ascertain where the problem with this latent factor lay, the mean inter-item correlation coefficient was computed and the item-total statistics were inspected. No underlying problem was evident. The computed mean inter-item correlation coefficient of 0.33 suggests the homogeneity of the seven items in the scale (Briggs and Cheek, 1986;Clark and Watson, 1995). Furthermore, the corrected item-total correlation values where all above 0.30, ranging from 0.45 to 0.58, suggesting that each of the seven indicators correlate well with the information value latent factor (Pallant, 2010). Moreover, deletion of any of the seven items would result in the Cronbach's alpha value for that latent factor decreasing rather than increasing. Therefore, the decision was taken to conclude convergent validity tentatively for the perceived information value latent factor, based on the standardised loading estimates, the mean inter-item correlation coefficient, the corrected item-total correlation values, the Cronbach's alpha value and the CR value.
Given the importance of establishing discriminant validity (Franke and Sarstedt, 2019) and in light of the HAVE of 0.58 for the perceived information value latent factor being below two of its relevant correlation coefficients, the HTMT ratio values were also calculated, as reported in Table 5.
As is evident in Table 5, the computed HTMT values range from 0.362 to 0.643, thereby providing evidence of discriminant validity between the latent factors in the measurement model. As such, discriminant validity was concluded between all of the five latent factors proposed for inclusion in the measurement model, including between perceived information value and the other four latent factors. Having established the reliability and construct validity of the model, attention was then turned to evaluating the model fit indices returned by AMOS.
The computed model fit indices all suggested acceptable model fit with a GFI of 0.919, an IFI of 0.943, a TLI of 0.934, a SRMR of 0.047 and a RMSEA of 0.052. Taken together, this suggests that usage frequency of online consumer reviews amongst Generation Y students is a five-factor model that exhibits construct validity, reliability and acceptable model fit.
Based on the validation of this measurement model, a structural model was specified to test the theorised paths that Generation Y students' perceived information and entertainment value and perceived credibility of online consumer reviews have a direct positive influence on their perceived value of such reviews, which, in turn, has a direct positive influence on their usage frequency of online consumer reviews. This structural model is illustrated in Figure 1.
With both the SRMR (0.06) and the RMSEA (0.05) below 0.08, together with a GFI (0.914), an IFI (0.937) and a TLI (0.928) above 0.90, the structural model in Figure 1 exhibited acceptable model fit. Table 6 presents the un-standardised and standardised regression coefficients, standard error estimates and p-values estimated by AMOS for the structural model.
The estimates reported in Table 6 indicate that all of the regression paths tested were positive and statistically significant. In terms of the standardised regression estimates, perceived information value (b = 0.37, p < 0.01), perceived credibility (b = 0.35, p < 0.01) and, to a lesser extent, perceived entertainment value (b = 0.17, p < 0.01) are statistically significant predictors of Generation Y students' perceived value of online consumer reviews. With a squared multiple correlation coefficient (R 2 ) of 0.54, these three factors explain 54% of the variance in Generation Y students' perceived value of online consumer reviews. Perceived value (b = 0.50, p < 0.01), in turn, is a statistically significant predictor of Generation Y students' usage frequency of online consumer reviews and, together with its predictors, explains 25% of the variance in their usage frequency of such reviews.

Discussion and implications
The study contributes to the literature by applying the U&G approach to explaining the usage frequency on online consumer reviews. The study empirically determined the factors that contribute to the usage frequency of online consumer reviews, with specific reference to Generation Y consumers. Following the U&G approach, the Ducoffe (1996) advertising value model comprising perceived information, entertainment and overall value was extended to include perceived credibility as predictors of Generation Y consumers' usage frequency of online consumer reviews. The findings of this study confirm that Generation Y students' usage frequency of online consumer reviews is a five-factor model that exhibits acceptable model fit, construct validity, internal-consistency reliability and CR. This model, which comprised the latent factors of perceived information and entertainment value, perceived credibility, perceived value and usage frequency of online consumer reviews, provides marketing academics and practitioners with a starting point for better understanding Generation Y's use of online consumer reviews. Based on the evidence in the sample, the results intimate that Generation Y individuals view online consumer reviews as important, useful and valuable, as a convenient and good source of relevant, timely and up-to-date information, as pleasing and fun to use, and as trustworthy. Furthermore, there are indications that they frequently consult such reviews. These findings, in conjunction with reports in the media, underscore the importance of marketers integrating online consumer reviews into their marketing communication strategy when targeting Generation Y consumers.
The results of the path analysis infer that perceived information and entertainment value, and perceived credibility have a significant positive influence on Generation Y students' perceived value of online consumer reviews, which, in turn, has a significant positive influence on their usage frequency of such reviews. The implications of these findings for marketers is that to leverage the persuasiveness of online consumer reviews to their strategic advantage when targeting Generation Y necessitates using tactics that will ensure that these reviews are comprehensive yet readable and authentic. This will contribute to their perceived value and usage frequency amongst the Generation Y segment. To encourage the cogency of consumer-generated review content, a template with demarcated sections requesting specific usage experience and a section requesting reviewers to mention both their positive experience with the product or service under review and any negative experiences or views they may have should be provided. The entertainment value of online consumer reviews relates to their readability and, as such, suggests that there should be a word-count limit set in the review template and that a spelland-grammar check tool should be provided on the review site. Arguably, the most important aspect that relates to the persuasive power of online consumer reviews is their perceived credibility, which highlights the importance of review sites earning a reputation for only publishing authentic customer reviews. For third-party review sites, this necessitates using software tools to monitor and identify fake reviews, and to use punitive measures against those vendors caught engaging in such practices. Company-owned consumer review sites need to be educated as to the long-term negative consequences of indulging in the practice of cherry-picking reviews or disguising paid-for product endorsements as real consumer reviews. Here, mass media has an important role to play by publishing reports on sites that engage in unethical consumer review practices.
Given the important role that online consumer reviews play in the consumer decisionmaking process, companies, whether large, medium, small or micro, are advised to use the services of a dedicated reputation marketer(s) with strong written communication skills whose sole focus is on monitoring reviews and responding accordingly, whether it be to thank someone for a positive review or to address any issues in a negative review.
The findings of this study also have implications for consumers, including Generation Y consumers. Blind trust in online consumer reviews is not prudent. Online consumer review sites that only have positive reviews are in all probability either censoring negative reviews or disguising paid-for product endorsements as authentic consumer reviews. The same is likely to hold true for vendors on third-party review sites that attract only positive reviews. Reviews that are overly enthusiastic and use superlatives should also be viewed with a degree of scepticism. It is advisable for consumers to consult multiple review sites before making a decision. An internet browser search, which may turn up contradictory consumer reports, is also advised.

Limitations and suggestions for future research
In interpreting the findings of this study, marketing academics and practitioners are advised to take cognisance of certain methodological limitations in the study. The most important limitation is the use of convenience sampling, which hampers the extent to which the results can be generalised to the target population. Unfortunately, the use of a convenience sample was unavoidable as the South African Protection of Personal Information Act 4 of 2013 prevents HEIs from supplying certain information, including lists of registered student names.
Another noteworthy limitation is that the data were collected using a cross-sectional approach, which means that the perceptions captured in the findings are dependent on the mood of the sample participants at that specific point in time. As mass media publishes more reports on the extent to which certain sites publish fake reviews or do not adequately police their sites for such fake reviews, so a different picture may emerge. Therefore, periodic new research into Generation Y individuals' perceived credibility on online consumer reviews is important. Moreover, it would be valuable to determine empirically the degree to which perceived information and entertainment value, perceived credibility, perceived value and usage frequency of online consumer reviews influence Generation Y students' purchase intentions towards goods reviewed.
The study was also country-specific, focussing only on the South African market. Even though international travel, global media channels and social media are increasingly leading to a more global consumer physiognomy, cross-cultural differences do still exist. Considering online consumer reviews in particular, individualistic and collective cultures are likely to differ in both their contribution to and consultation of such reviews. Future research exploring the effect of such differences on the usage frequency of online consumer reviews will be of value, both academically and practically.
Consideration also needs to be given to other generations' usage frequency of online consumer reviews. The oldest members of Generation Z (individuals born after 2005) (Markert, 2004) turn 18 in 2023are such reviews relevant to them and, if so, will they continue being relevant to them in the foreseeable future? The same holds true for Generation X (individuals born between 1966 and 1985) (Markert, 2004). Applying the model tested in this study to Generations X and Z will provide important insights for marketers targeting these two generational segments.

Conclusion
Using an extended version of the advertising value model, this study investigated the antecedents of Generation Y students' usage frequency of online consumer reviews. The findings confirm that perceived information and entertainment value, perceived credibility Antecedents of frequency of reviews and perceived value have a positive influence on Generation Y students' usage frequency of online consumer reviews. The findings highlight the importance of ensuring that only authentic reviews are published and of not censoring negative reviews, and suggest the need to implement tactics to ensure that such reviews are informative and entertaining.