Statistics of Response to novelty as a predictor for drug effects: the pitfalls of some correlational studies