University of Twente Student Theses


Exploring credibility in electronic word-of-mouth

Menkveld, B.G.T. (2013) Exploring credibility in electronic word-of-mouth.

[img] PDF
Abstract:More than half of the Dutch population makes use of electronic user-reviews online by reading and evaluating these reviews as part of their decision-making process. However, an important practical issue is the rise of fake reviews on the Internet. Studies of Chevalier and Mayzlin (2003) and Zhu and Zhang (2010) showed evidence that reviews can have a direct influence on product sales, and marketers are well aware of this. As a result, the popularity of online reviews brings about growing concerns about the credibility of online reviews since there is substantial evidence of marketers cheating the public opinion with fake online reviews. Therefore, this study aims to answer the following research question: which factors influence credibility assessments of eWOM of strangers? Our sub questions were based on three pillars: - What is the influence of receiver characteristics in credibility assessments of anonymous eWOM? (characteristics of the receiver) - What is the influence of attributes of the review on credibility assessments? (characteristics of the message) - What is the influence of characteristics of the reviewer on credibility assessments? (characteristics of the source) Given the exploratory nature of the research objectives, a preliminary study was conducted. The goal of this preliminary study was to collect rich and detailed information concerning the attitudes, beliefs and trust of consumers towards anonymous eWOM via 15 interviews. This study made it possible to cover the predetermined areas of research, but it also allowed for unexpected new insights. The interviews not only helped answering some of the sub questions, but also served as a basis for the main study: an online experiment. The preliminary study showed that there is not one particular cue that influences credibility assessments, as it appears to be a combination of characteristics of the source, receiver and the message. In the second part of the study we took an experimental approach. Our experimental material consisted of a (for the purpose constructed) website that contained product information about a camera and twelve (manipulated) user reviews. Reviews were manipulated on name, expertise, message quality, negative product information, social presence, gender and consistency. There were four versions of the website that all looked the same, except for the reviews on it because these were randomly divided over the four websites. This was done to account for other (unintended) effects, such as the influence of the order of certain reviews or the combination of certain reviews. The four versions of the websites were equally distributed to the participants. Participants were asked to comment on the credibility and trustworthiness of each review by giving plusses and minuses accompanied by explanation for elements that helped or hurt perceived credibility. This could be any element, from words to sentences to profile pictures or punctuation marks. Afterwards, all 126 participants were asked to fill out a survey that measured their levels of trust on a 7-point Likert scale. When the main study was finalized, around 2000 comments on perceived credibility had to be analysed. This was done by coding each comment, based on a codebook that was both data-driven and theory-driven. The results of the main study confirmed what was said in the preliminary study: it appears that there is not one particular cue that influences credibility assessments, because it is a combination of cues. So, to answer the main research question: as expected, it appears that in this study perceived credibility assessments are influenced by a mixture of characteristics of the receiver, characteristics of the message and characteristics of the source. From the data of this study, it appears that there is not one factor that will instantly make a review credible and trustworthy. It is the interaction between characteristics of the receiver (e.g. experience with eWOM, attitude towards eWOM and trusting stance), characteristics of the source (e.g. expertise, experience and social presence) and characteristics of the message itself (e.g. message quality and balanced arguments) that influences credibility assessments. This study builds on the existing knowledge about eWOM and made a first step in adding more insight into what cues can increase or hurt perceived credibility in an online environment. Because of the nature of the study, the results cannot be seen as hard evidence, but must be seen as the first step in a journey towards full understanding of credibility in electronic word-of-mouth. Findings in this study should be used to gain a deeper understanding of what causes credibility in online reviews and even more important: how to increase the credibility and trustworthiness of these reviews. Future studies need to find out if results of our study can be generalized, for instance through a quantitative study that builds on our more explorative, qualitative study, in order to provide more definitive insights of what causes credibility in eWOM.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:05 communication studies
Programme:Communication Studies MSc (60713)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page