Assessing Quality of Consumer Reviews in Mobile Application Markets: A Principal Component Analysis Approach

This study presents a simple, theory-based method for calculating a metric which reflects the quality of online consumer reviews in mobile application markets. Derived from prior online consumer review studies based on psychology, information quality, and economics literature, a metric for measuring online consumer review quality is developed. The metric is a weighted sum of three variables (Squared Star Rating, Log-transformed Word Count, and Sum of Squared Negative and Positive Sentiment), and weights for calculating the metric are estimated by using Principal Component Analysis (PCA) technique. Preliminary assessment of the proposed method shows that metrics computed by using the proposed method are positively correlated with helpfulness ranks of mobile application reviews in Google Play. However, PCA results show that one of the variables (i.e., sentiment) used for developing the metric did not load consistently on the first factor component. From the findings of the preliminary evaluation on the metric, limitations and future research directions of the proposed method are discussed.