Media Effects Research Lab - Research Archive
Putting Negativity Bias to the Test
Student Researcher(s)
Xiaoye Zhou (Masters Candidate);
Na Sun (Ph.D Candidate);
Nia Marie Sherony (B.A. Candidate);
Faculty Supervisor
Introduction
With its great convenience and practicality, online shopping is more popular than ever. Online shoppers have been shown to rely heavily on online reviews when making purchase decisions. A 3 × 3 × 2 ANCOVA test (N=206) found significant main effects of the modality of positive and negative reviews on consumers’ attitude and behavioral intentions toward products. Also, attitude was found to mediate the effect of positive review modality on consumer behavioral intention, which is subsequently moderated by negative review modality. Theoretical implications and practical implication are discussed in this paper.
Hypotheses
H1a: The modality of positive reviews will have a main effect on perceived review usefulness.
H1b: The modality of negative reviews will have a main effect on perceived review usefulness.
H1c: The modality of positive reviews will have a main effect on perceived review credibility.
H1d: The modality of negative reviews will have a main effect on perceived review credibility.
H2a: The modality of positive reviews will have a main effect on attitude toward the product.
H2b: The modality of negative reviews will have a main effect on attitude toward the product.
H2c: The modality of positive reviews will have a main effect on behavioral intention toward the product.
H2d: The modality of negative reviews will have a main effect on behavioral intention toward the product.
RQ1: Will there be an interaction effect between the modalities of positive and negative reviews on consumers’ attitude toward the product?
RQ2: Will there be an interaction between the modalities of positive and negative reviews on consumers’ behavioral intention toward the product?
RQ3: Will the modalities of positive and negative reviews indirectly influence behavioral intentions toward the product, mediated by consumers’ attitudes toward the product?
Methods
Our study aims to reveal modality prominence in cross-valence product review pairs. We conducted a 3 (modality of positive reviews) × 3 (modality of negative reviews) × 2 (product types) mixed full factorial design. The basic experiment involved participants exposed to two types of products, a USB drive and a printer, each of which was presented with a pair of reviews – one positive, one negative – following two rounds of pretests. In addition to product type as the within-subject variable, we include two other between-subject variables: modality of the positive review (text, text-plus-picture, or video), and modality of the negative review (text, text-plus-picture, or video). The experiment was constructed so that independent variables are mixed and interact together in full factorial design, giving rise to a collection of 18 cases in total. Each experimental condition always has a cross-valence pair of reviews due to the combination of the modalities under opposite valence (positive and negative). The details for each of the independent variables will be introduced in the following section. It should be noted that the product difference, or the within-subject variable is not the focus of the study. Instead, its function in the experiment is (1) to rule out the possible personal relevance or bias to the product being reviewed, and (2) to enhance the external validity of our study in general in positive (or negative) review modalities.
Results
We found main effects of positive review modality, negative review modality, and product type on consumers’ attitude toward the product and behavioral intention toward the product. We also found significant interactions of positive review modality and negative review modality on both attitude and behavioral intention toward the product. When negative video is present, the modality of the positive review does not make many difference. When the negative review becomes less rich in terms of modality, from video to text-plus-picture, and to text only, the positive video failed to counteract the negative review valence, leading to a sharp drop in the graph from right to left. We also found that when cross-valence reviews were present in the same modality (i.e., positive text and negative text, positive text & picture and negative text & picture, positive video and negative video), positive valence reviews seemed to outweigh their negative counterparts. Last but not least, we found the effect of positive review modality on behavioral intention was mediated by consumers’ attitude toward the product, and moderated by negative review modality.
Discussion
For both negative and positive reviews, the text modality dominated the counterpart modality when the modalities in the cross-valence review pairs differ. In contrast, video is of the least importance in predicting product attitude and behavioral intent when balancing out the effect of modality in the opposing valence review conditions. positive video is more subject to the influence of negative modality, especially when the level of the latter variable changes from negative video to negative text, only. The rating of product attitude of the same modality was significantly higher than the rest, and increased compared with prior brand evaluation.
For more details regarding the study contact
Dr. S. Shyam Sundar by e-mail at sss12@psu.edu or by telephone at (814) 865-2173