# When is it admissible to combine two categories?

In a data mining project, I have a categorical variable $$C$$ which takes possible values $$c_i$$ and the outcome variable is continous.

When I plot the conditional probability distribution of the outcome variable given $$c_i$$ i.e. $$P(O|c_i)$$ I get similar distributions for some of the categories e.g. $$P(O|c_k)$$ is $$N(0,1)$$ for $$k=1,2,3$$.

Am I elligible to combine these similar groups together in this case?

Note that I have many other categorical and continuous variables.