An Ethical Argument Against Leaving Psychologists to Their Statistical Devices

Abstract

The frequent critiques of the statistical tools used by experimental psychologists, especially the null-hypothesis test, cast substantial doubt on the scientific integrity of experiments associated with them. This doubt leaves psychologists with weak grounds for promising results or benefits from their experiments. In addition to the methodological problems raised, the shortcomings of null-hypothesis testing and associated experiments suggest an ethical problem in which deception of the participants is involved. Psychologists who use deception claim that the scientific and educational gains from their experiments balance the ethical costs. But given a dubious prospect of scientifically respectable results, there are good reasons to wonder whether there can be anything to offset the deception. Null-hypothesis testing therefore undermines not only experimental rigor but a significant part of the ethical justification for deceptive methods as well.