This dissertation aims to provide a deeper understanding of consumer emotions from user-generated content. In the first essay of my dissertation, “Digital Therapy for Negative Consumption Experiences: The Impact of Emotional and Rational Reviews on Review Writers and Readers,” we examine whether the manner in which a consumer writes a review can help them recover from a negative consumption experience, as well as how this influences other review readers.
To test these research questions, we use a multimethod approach and collect archival data, field data, and participants’ physiological responses as well as memory recall. We employ machine learning techniques to train multilabel classifiers with review textual data and categorize online reviews into one of the three types: emotional (feelings and emotions), rational (facts, beliefs, and thought processes), and integrated (the combination of the two).
We first show that, similar to writing about traumatic life experiences, when a review writer writes an integrated review about a negative experience (compared to when they only express emotional or rational aspects), they feel better afterwards and are more likely to purchase again. We also show that integrated reviews do not any have adverse effects on review readers. Finally, in two controlled experiments, we examine the underlying mechanisms for this positive effect of writing an integrated review on review writers by collecting biophysiological response data (i.e., review writers’ blood pressure and pulse) and by analyzing thought listing data.
The results show that writing an integrated review about a negative consumption experience appears to lead to catharsis and cognitive reappraisal of the negative experience, which in turn lead to better outcomes. This research shows that writing online reviews can serve as a digital therapy tool that helps consumers recover from negative consumer experiences and has positive benefits for the involved firms. This has important implications for the design of review systems and for firms, especially in situations where customers have negative experiences with their products and services.
The first essay of my dissertation suggests that consumers can benefit by expressing emotions (together with rational thoughts) in their online reviews. In the second essay of my dissertation, “Are Emotions Gendered? Gender Stereotypes in Online Reviews,” we examine whether the domain of online reviews is inclusive enough to allow all consumers to feel free to express their emotions. In this research, we examine whether review readers’ reactions to reviews where writers express their emotion vary with the gender of the review writer.
More specifically, we examine how gender stereotypes in general, and the belief that females are more emotional than males in particular, influences review readers’ reactions to reviews as well as the manner in which review writers construct their reviews. We find that even though the domain of online reviews is a relatively private and safe place for consumers to express their evaluations of products and services, a common feature of online review system designs, that is asking review writers to provide an avatar and/or names that might reflect their gender, leads to less favorable reactions to reviews written by women (vs. men) because of gender stereotypes.
Further, when the stereotype that women are more emotional than men is made salient before review writing (versus when it is not), female review writers express less emotion, possibly because emotionality has negative associations such as being “irrational,” “overdramatic,” and “sensitive.” This finding is important because other research has shown that reviews that contain more emotion are evaluated more positively by review readers. Most importantly, while we provide evidence that this stereotype is believed and has a negative impact on review writers and readers, we show that it is not true in this context – females are no more emotional than men in review writing contexts.
Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/87bk-6840 |
Date | January 2024 |
Creators | Wu, Yinghao |
Source Sets | Columbia University |
Language | English |
Detected Language | English |
Type | Theses |
Page generated in 0.0016 seconds