Details in Study Methodology Can Give Misleading Results

Details in measurement methodology make the results of a recent market research study irrelevant for predicting real user behavior on the Web. The finding was indeed so striking as to conflict with most other evidence: When new data contradicts ...