Key points are not available for this paper at this time.
This note compares the power functions of the sign test for slippage with the power functions of the most powerful test for the case of normal populations. The sign test is found to be approximately 95% efficient for small samples.
John E. Walsh (Sun,) studied this question.