Study 1 findings reveal that WOM recommendation sources and expert recommendation sources play a more important role than advertising recommendation sources do in determining the perception and intentions regarding the use of e-books. This result agrees with Yeung and Yee (2003), who indicated that the reliability of the level of expert opinions was lower than that of WOM recommendation sources in the decision-making process.
Agreeing with the suggestions of Gilly, et al. (1998), this study found that interpersonal sources (WOM recommendation sources or expert recommendation sources) generally appeared to be more preferable than non-personal sources (advertising recommendation sources) did.
An academic digital library can be created to meet academic resource needs. Librarians must emphasize the benefits and advantages of e-books (e.g., easy searches and indexes, a reduction in book costs, and increased convenience) to encourage users to search academic resources, use e-books and share their e-book usage patterns with other potential users, such as family, friends, and classmates. Experts (such as professors and relevant professionals) may also provide practical operational manuals and guides to help users enhance their knowledge of e-books. Experts’ ideas and suggestions should be considered, and can be useful in influencing usage intentions.
Furthermore, advertising effects were less than the effects of WOM recommendation sources and expert recommendation sources on behavioral intentions; these effects can impress users or potential users and persuade them to action (Kanso and Nelson, 2004).
Advertising links were the most frequently used methods on library Web sites (Švenčionytė, 2005). Librarians can apply different advertising strategies (e.g., presenting online advertising and posting library announcements) to promote a linking portal to e-books and attract greater
user attention and interaction.
In addition, the mediators of perceived trust and perceived risk appeared to be powerful elements in explaining the intention to use e-books using recommendation sources. Indeed, recommendation sources effects can enhance the perceived trust and reduce the perceived risk, leading customers to determine whether or not to use e-books. Consistent with Pavlou (2003), who suggests that integrating trust and risk can explain user acceptance of e-commerce, a possible explanation for this result came from the finding that users can maximize their confidence and minimize uncertainty by gaining recommendation sources as they decide whether to use e-books in an academic digital library.
As mentioned previously, users can obtain an effective recommendation sources from different sources. Thus, librarians should create a discussion forum on library Web sites to allow greater exposure to e-book information, stimulating potential users to browse all of the relevant information on e-books. This will also increase the desire to use e-books and navigate topics users wish to investigate more easily. In this way, users can quickly find recommendations from others (through social networks, user communities, and expert networks) to foster their intentions to use e-books.
Consistent with the findings of Bickart and Schindler (2001), we found that discussion forum messages have greater credibility in inducing empathy among potential e-book users than advertising does. Providing online discussions about e-books and their benefits could diminish future hesitancy to use e-books and raise self-confidence in this new medium.
Furthermore, in study 2, this investigation examined how personal variables and recommendation sources influence users’ behavioral intentions as related to academic e-books.
The “at least 5” and “at least 4” classes correspond to users having behavioral intentions to use e-books in academic libraries, and the antecedents to these classes of rules tell academic libraries what they should consider. The analytical results show that users who trust
recommendation sources that vouch for e-books are more likely to use e-books and that WOM recommendation sources influence customers’ behavioral intentions to use e-books more than advertising recommendation sources and expert recommendation sources do. This is consistent with Study 1. In addition, users who conduct more daily Internet usage rely on recommendation sources to use e-books, especially from WOM recommendation sources and advertising recommendation sources. Finally, females who trust WOM recommendation sources and males who trust expert recommendation sources are more likely to use e-books if those recommendations give positive feedback.
The “at most 3” and “at most 4” classes correspond to users who have no or little behavioral intention to use e-books. The analytical results showed that the behavioral intentions of users who have no confidence in recommendation sources to use e-books would decrease, especially with respect to WOM recommendation sources and expert recommendation sources.
The results of this study have implications for decision-makers. One implication is that academic libraries may use combined WOM recommendation sources and expert recommendation sources strategies to promote academic e-book usage. For instance, an academic library can design recommendation activities where users who recommend e-books to others are rewarded. Especially in an online environment, our suggestion is consistent with Park, Lee and Han (2007), who pointed out that marketers should consider providing user-generated information services and recommendations by previous users in the form of electronic word-of-mouth (eWOM) communication.
Another implication is that different types of recommendations can attract different types of users. There are differences in how recommendation sources impact the two genders, so the academic libraries can apply different recommendation-based strategies. The research regarding gender differences in this study is similar to the findings of prior studies for
technology-based environments; these studies have demonstrated that national, ethnic and gender differences simultaneously influence perceptions and behaviors (Gefen and Straub, 1997). Hence, academic libraries could target mass media (e.g., library news) to male users and alternative media (e.g., a discussion forum) to female users.