Product reviews that are posted on shopping websites, including Amazon.com, don’t provide an accurate reflection of the product’s actual benefit. That’s the take-away of Mícheál de Barra of University of Aberdeen from the results of research that compared clinical-trial data with user-generated online reviews of weight-loss pills and high-cholesterol treatments that were sold on Amazon. He found that the reviews spoke far more positively about the products’ effectiveness than the clinical-trial data warranted.
De Barra believes that the inconsistent nature of the reviews as compared with the clinical-trial data is a result of a bias that people have toward sharing positive rather than negative outcomes.
Does the research in conjunction with medical information apply also to nonmedical products and services? De Barra concedes uncertainty about that conclusion and says more work is required in this vein, but he suspects that the evidence translates to a larger scale.