The informational value of multi-attribute online consumer reviews: A text mining approach

In the digital age, customers use online reviews to minimize the risks associated with purchasing a product. Major online retailers help customers choose the right product by exposing reviews that received many “helpful” votes at the top of the review section. Given that reviews that have received the maximum helpfulness votes are considered more important in purchase decisions, understanding determinants of helpfulness votes offers clear benefits to online retailers and review platforms. This study focuses on the effect of review informativeness, which is measured by the number of attributes discussed in a review, and its interplay of review valence on customers' perception of review helpfulness. We applied a word-level bigram analysis to derive product attributes from review text and examined the influence of the number of attributes on the review's helpfulness votes. More importantly, we also suggested the moderating role of review valence. Estimation results of the Zero-inflated Poisson models on 21,125 reviews across 14 wireless earbuds indicated that as more attributes are discussed in a review, the more the review can earn helpfulness votes from customers. Furthermore, the positive association between the number of attributes and helpfulness was enhanced among negative reviews. This study contributes to customers' information processing literature and offers guidelines to online retailers in designing a better decision support system.     

Mining the text of online consumer reviews to analyze brand image and brand positioning

The growth of the Internet has led to massive availability of online consumer reviews. So far, papers studying online reviews have mainly analysed how non-textual features, such as ratings and volume, influence different types of consumer behavior, such as information adoption decisions or product choices. However, little attention has been paid to examining the textual aspects of online reviews in order to study brand image and brand positioning. The text analysis of online reviews inevitably raises the concept of “text mining”; that is, the process of extracting useful and meaningful information from unstructured text. This research proposes an unified, structured and easy-to-implement procedure for the text analysis of online reviews with the ultimate goal of studying brand image and brand positioning. The text mining analysis is based on a lexicon-based approach, the Linguistic Inquiry and Word Count (Pennebaker et al., 2007), which provides the researcher with insights into emotional and psychological brand associations.     

Creating and detecting fake reviews of online products

Customers increasingly rely on reviews for product information. However, the usefulness of online reviews is impeded by fake reviews that give an untruthful picture of product quality. Therefore, detection of fake reviews is needed. Unfortunately, so far, automatic detection has only had partial success in this challenging task. In this research, we address the creation and detection of fake reviews. First, we experiment with two language models, ULMFiT and GPT-2, to generate fake product reviews based on an Amazon e-commerce dataset. Using the better model, GPT-2, we create a dataset for a classification task of fake review detection. We show that a machine classifier can accomplish this goal near-perfectly, whereas human raters exhibit significantly lower accuracy and agreement than the tested algorithms. The model was also effective on detected human generated fake reviews. The results imply that, while fake review detection is challenging for humans, “machines can fight machines” in the task of detecting fake reviews. Our findings have implications for consumer protection, defense of firms from unfair competition, and responsibility of review platforms.     

An empirical comparison of machine learning methods for text-based sentiment analysis of online consumer reviews

The amount of digital text-based consumer review data has increased dramatically and there exist many machine learning approaches for automated text-based sentiment analysis. Marketing researchers have employed various methods for analyzing text reviews but lack a comprehensive comparison of their performance to guide method selection in future applications. We focus on the fundamental relationship between a consumer’s overall empirical evaluation, and the text-based explanation of their evaluation. We study the empirical tradeoff between predictive and diagnostic abilities, in applying various methods to estimate this fundamental relationship. We incorporate methods previously employed in the marketing literature, and methods that are so far less common in the marketing literature. For generalizability, we analyze 25,241 products in nine product categories, and 260,489 reviews across five review platforms. We find that neural network-based machine learning methods, in particular pre-trained versions, offer the most accurate predictions, while topic models such as Latent Dirichlet Allocation offer deeper diagnostics. However, neural network models are not suited for diagnostic purposes and topic models are ill equipped for making predictions. Consequently, future selection of methods to process text reviews is likely to be based on analysts’ goals of prediction versus diagnostics.     

Emotional contagion triggered by online consumer reviews: Evidence from a neuroscience study

People tend to align with the emotional state of the person that is talking to them (the observed). Similarly, while processing information consumers can also experience this emotional contagion. Emotional contagion can activate in those who process information (the observer) similar responses in the autonomic nervous system and the neural responses as in those who create such information (the observed), triggering a certain level of arousal. Neuroscience enables researchers to study emotional contagion by monitoring the activation of neural structures and physiological responses. This study draws on the theory of arousal to investigate how different combinations of online consumer review (OCR) valence can trigger different emotions and customer experiences in the observer (the one who reads the OCRs). This study conducts a consumer neuroscience experiment to monitor emotional arousal. The physiological analyses (through skin conductance response) confirm that the emotional arousal of the observer aligns with that of the observed. The neural analyses (through electroencephalography) show the valence of the arousal, which indicates that negative OCRs activate arousal and pleasure in the observer, while positive OCRs are associated with arousal deactivation and displeasure.     

Assisting consumers in detecting fake reviews: The role of identity information disclosure and consensus

Current discussions in academia and in the press increase consumers’ awareness of potentially deceptive online reviews. The increasing practice of fake reviews posted online not only jeopardizes the credibility of review sites as important information sources for individuals but also endangers a valuable source of information for service providers. Two studies shed further light on the role of consensus and identity-related information in assisting consumers detect potentially faked reviews. In one preliminary study, a sample of 4826 rejected and 4881 published online reviews was analyzed to investigate the differences in the disclosure of author-related information such as name and age as well as star ratings across those reviews. In the main study, a 3 (identity disclosure) x 2 (consensus) x 2 (priming of fake reviews) experiment was carried out with 390 respondents. The results highlight the relevance of the review's consensus in relation to the overall rating of previous reviews and corroborate the results of the preliminary study from the perspective of an internet user: the value of the amount of available information on the review's author in assisting individuals detect potential fake reviews. This study complements research in computer science by highlighting the relevance of contextual—in addition to textual—indicators that assist internet users in detecting potentially deceptive online reviews.     

Examining impacts of negative reviews and purchase goals on consumer purchase decision

Extensive literature has studied the impact of positive online reviews on consumers’ purchase decisions. Research on the role of negative reviews in consumer price perception is heavily under-investigated. This study examines effects of negative reviews on consumer price perception and subsequent purchase behavior while considering the moderating effect of purchase goals. Results from two experiments show that the proportion of negative reviews has stronger negative impacts on purchase decision for consumers with a purchase goal than those without. This research contributes to growing knowledge about negative online reviews and consumer goal literature and offers practical implications for online retailers.     

Experimental analysis of consumer channel-mix use

Today's consumers have access to multiple consumer distribution channels. To remain competitive, retailers must offer different contact points to consumers. This empirical study examines channel-mix use decisions for 352 online customers' holiday booking preparations. A scenario based experimental approach studies consumer channel-mix use by channel and decision context attributes. The study models effects of time pressure, expected expenditure, channel quality, and access costs on the use of eight channel options, including traditional and online options. Over one-half of respondents use multiple channels to decide trip booking; however, most travelers book trips using the channel they initially investigated for information. Results show expected expenditure does not influence channel use; however, overall cost affects the booking's timing. Time pressures force respondents to visit a travel agent or collect additional brochures to quickly complete the purchase. Results suggest brochures fulfill an important support role.     

The impact of text valence,star rating and rated usefulness in online reviews

The influence of text valence, star rating and rated usefulness of online reviews on review readers’ impression of the review and their positive word-of-mouth intention is tested in an experimental study (n = 431). In addition, we investigate the moderating role of review readers’ product category involvement and susceptibility to interpersonal influence on the effect of the three review components. The influence of review text valence on evaluative responses is stronger for more highly involved people and for people who are more susceptible to interpersonal influence. The influence of rated review usefulness on review impression is marginally stronger for people who are more susceptible to interpersonal influence. Star ratings do not influence evaluative responses, and their effect is not moderated by either involvement or susceptibility.     

A discourse analysis of pilgrimage reviews

ABSTRACT

This article is the first to provide an account of the discursive features of online consumer reviews of pilgrimage sites. Drawing from pilgrimage studies and narrativity theory in consumer research, the authors explore how consumers communicate the spiritual and material aspects of pilgrimage experiences by examining a corpus of 833 consumer reviews on TripAdvisor of the most sacred pilgrimage sites of the world’s major five faith groups. Pilgrims include analytical discursive features to communicate the material aspect of their consumption experience. They reserve narration for spiritual transformation and the experience of strong emotions. Moreover, review ratings are only reflective of the spiritual aspect of their consumption experience. As such, this research complements previous studies by highlighting the material, physical aspect of this extraordinary consumption experience.     

Complimentary competition: The impact of positive competitor reviews on review credibility and consumer purchase intentions

Online reviews are important decision aids for consumers, but not all reviews are perceived as equally credible, which may hinder consumer decisions. This research identifies positive competitor reviews as a novel source of credible reviews. A pre-test and four studies show that consumers perceive positive reviews written by competitors as more credible than similar consumer-generated reviews. This is because, compared to lay consumers, competitors are perceived as less likely to be compensated for their positive reviews. Positive competitor views, in turn, elicit greater intentions to purchase the reviewed product. Positive competitor reviews are expected to contribute to a “coopetitive” environment.     

A comparative analysis of four scales of consumer involvement

Four recent scales of consumer involvement are compared. These scales are first scrutinized, and, where necessary, modified, on a priori grounds. The modified scales are then empirically compared in terms of unidimensionality, convergent and discriminant validity, and nomological validity. On these criteria, the pruned and modified version of each of the four scales is found satisfactory. However, some unique features of each, which are discussed as trade-offs that marketing researchers would have to consider in their choice of a scale to measure this important consumer behavior construct. © 1995 John Wiley & Sons, Inc.     

Understanding the influence of submission devices on online consumer reviews:A comparison between smartphones and PCs

Online consumer reviews (OCRs) are of immense value to businesses and consumers. With the rapid development of mobile Internet, smartphones have become an important device for consumers to publish OCRs. Information input is more difficult on smartphones than on PCs because of the screen size, the convenience of using keyboard and the interference of use situation. Based on cognitive cost theory, this study compares the differences of the statistical characteristics of OCRs (study 1) and the narrative content of OCRs (study 2) submitted via smartphones and PCs about material purchases in B2C websites. In study 1, an analysis of 24,933 OCRs across six products shows that OCRs submitted via smartphones have fewer words, more images, higher star rating, and shorter time interval than via PCs. In study 2, by coding and analyzing 1001 OCRs, the results show that OCRs submitted via smartphones are less likely to mention price, product functionality, product quality, product aesthetics, service quality, logistics quality, and attitudinal loyalty than via PCs, whereas no significant difference is observed in the dimensions of seller trustworthiness, cognitive attitude, emotional attitude, and recommendation expression. Implications for theory and practice are discussed.     

Critical thinking in consumer studies: part 2 of a content analysis of Canadian university consumer studies courses

Abstract This paper concludes a two-part study on a content analysis of Canadian university consumer studies courses offered in human ecology, home economics or related programmes. Phase 1 of the study profiles the programmes, professors, students and the results of a content analysis of these courses. This companion research examines the inferred level of learning associated with consumer concepts in the course outlines. Instructors were choosing verbs associated with knowledge and comprehension rather than application, analysis, synthesis or evaluation. Implications regarding the appropriateness of these courses to prepare graduates for consumer-related careers are discussed. Lessons to be learned from the Canadian experience are tendered, especially the need for critical theory, a systems of actions perspective and an ethical, value-based approach in consumer studies courses.     

The impact of consumer decision‐making styles on consumer confusion in Mauritius: An empirical analysis

This article investigates the effects of consumer decision‐making styles on consumer confusion and harm. While consumer confusion has been well documented in the literature, this investigation attempts to bridge our understanding on its relationship with consumer decision‐making styles. Empirical data was collected from a field survey where 400 consumers selected at 20 supermarkets in Mauritius were asked to respond to a questionnaire measuring three variables under study: consumer confusion, consumer decision‐making styles, and situational factors. Findings from the MANOVA parametric test showed that there is a significant difference between consumer decision‐making styles and consumer confusion. Even when covariate situational factors were controlled through the parametric test MANCOVA, results still showed a significant difference between consumer decision‐making styles and consumer confusion. More specifically, ANCOVA tests showed that the findings were specifically significant for three types of confusion: product confusion, packaging confusion, and product complexity. The study tends to confirm that a consumer's style of decision making does impact on consumer confusion.     

Exploring the use of content analysis methodology in consumer research

Content analysis is a viable way to thematise consumer experiences. The purpose of this review study is to examine the use of content analysis in consumer research (1977–2017). The authors explore how content analysis has been used. The reviewed studies address consumer experiences. The results show that qualitative content analysis is not used as a method in its own right; it is more often applied as a supplement to quantitative testing. There is also a lack of rigorous reporting of methodologies in many studies. The systematic review provides four propositions, content analysis studies: 1) vary in execution and reporting; 2) have a tendency towards methodological vagueness; 3) do not apply content analysis as a sole method; 4) are versatile. The study can also serve as point of departure for novice researchers wishing to engage with content analysis research. We suggest that further research is needed to explore the use of qualitative methodologies in consumer research.     

居民消费价格指数的因子分析模型及应用

由于目前对居民消费结构的分析指标很多且指标体系中各指标之间存在着多重共线性,从而影响了分析模型的稳定性,使所得模型中出现了不符合经济学原理的现象.为此本文采用多元统计分析方法,建立了关于居民消费价格分类指数变动的因子分析模型.利用SPSS软件找出影响居民消费结构的一致性因素,消除了多重共线性,并以某一具体实例验证了该模型的实用性和有效性.     

我国个人消费信贷现状分析及应对策略

个人消费信贷业务近年来以巨大的市场潜力和较高的预期收益成为金融市场竞争的热点.大力发展个人消费信贷业务,有利于扩大国内消费需求,拉动经济增长,其前景不容忽视.目前我国消费信贷业务仍存在诸多局限性,本文根据我国个人消费信贷现状,提出各项应对策略.     

The effect of negative online customer reviews on brand equity and purchase intention of consumer electronics in South Africa

This study investigated the effect that negative online customer reviews have on brand equity and purchase intention. This research examined the influence of negative electronic word of mouth (eWOM) on brand equity, as influenced by customer review quality, product involvement and source credibility. A 3 (high quality review versus low quality review versus no review) × 2 (high product involvement versus low product involvement) factorial experiment with 236 respondents was conducted. The results revealed that the presence of negative eWOM has a significant detrimental effect on brand equity and purchase intention. Furthermore, the effect of negative online customer reviews is more detrimental to the brand equity of a high involvement product than a low involvement product. The results also revealed that high quality reviews are more influential than low quality reviews with respect to brand equity, and the difference between the levels of eWOM source credibility has no significant effect on brand equity.