Can't we calculate the P value for non parametric tests ? We have done it for One sample sign test but we are not doing it for Wilcoxon and Mann Whitney in these videos
Good observation👍 p-value repesents area under the curve, and in order to calculate it we need a proper distribution. Calculating it manually in case of non-parametric tests is a bit tricky because unlike the sign test where luckily we could bring binomial distribution, we don't have proper distributions for non-parametric tests. The only way out is to approximate to some known distribution (say normal distribution using CLT) and then calculate the p-value. This is the reason we have shown the p-value only through python code. Having said that, maybe sometime in future we'll add this portion as well.😊
Can't we calculate the P value for non parametric tests ? We have done it for One sample sign test but we are not doing it for Wilcoxon and Mann Whitney in these videos
Good observation👍
p-value repesents area under the curve, and in order to calculate it we need a proper distribution. Calculating it manually in case of non-parametric tests is a bit tricky because unlike the sign test where luckily we could bring binomial distribution, we don't have proper distributions for non-parametric tests. The only way out is to approximate to some known distribution (say normal distribution using CLT) and then calculate the p-value. This is the reason we have shown the p-value only through python code. Having said that, maybe sometime in future we'll add this portion as well.😊
@@prosmartanalytics thanks