In a statistical study for calculating the effect of drug on patient's sugar level the test showed significant difference when in reality there was no difference