A Process for Identifying Products; True Qualities and Raters; Biases and Accuracies

Stanford researchers have developed a process to simultaneously determine the true qualities of products based on ratings as well as the biases of individuals who provide ratings. This process is rooted on a new statistical technique incorporated into software that draws upon a history of ratings to iteratively estimate raters' biases and accuracies as well as items' true qualities to converge to a fixed point.