# Is normal distributed data more robust than skewed distributed one ?

I am learning about the skewed and normal distributions so I apology for some novice questions in advance. I know the definitions of them but I do not know if they can also indicate the robustness of the data. What other meanings do they have ?

For example, assume that there are 2 drugs A and B. Every drug is formed by a few chemical components. Among these components of one drug, there are interacting forces. Then, we label each force by a numerical value. If all forces in drug A follow a normal distribution while those in B follow a skewed distribution, can I say A is more robust than B statistically ?

Thank you very much,

-
In statistics one concept of robustness is that a test statistic which is based on a sample from a known distribution from a paramtric family has its distribution under the null hypothesis be insensitive to departures from that distribution that it has for the sample having the parametric distribution. It is not the data that is robust it is the test statistic and it is relative to some assumed null distribution. You seem to be asking about some other property. I am not familiar with your version of "robustness". Can you explain what you mean? –  Michael Chernick Sep 23 '12 at 11:28
Also in statistics it is often the normal distribution that is the parametric family that is assume for the population and the robustness property of the test statistic is relative to it. –  Michael Chernick Sep 23 '12 at 11:32