Note on Relative Efficiency of Tests