Hey Aditya! Thanks for contacting careers 360.

Non parametric tests when your data isn't normal.
Therefore the key is to figure out wether you have a normally distributed data.

For example - if you don't have a graph then how do you figure out if your data is normal distributed?

Check the skewness and kurtosis of the distribution.
A normal distribution has no skew.

If the distribution is not normal then the skewness and kurtosis deviate a lot. You should use a non parametric test like a chi square test.

Another example - another example for the one way anova is that the data comes from a normal distribution. If your data isn't normally distributed then you cannot run the anova, but you can run the non parametric alternative test - kruskal Wallis test.

Wish you all the best!