One of the foundation blocks of good science is replication. However if the agency holding the data will not share it with other qualified scientist, then they may have something to hide. No Replication = Bad Science
This letter to the Editor in the WSJ supports that view:
Here is what the writer is looking for a relationship between an increase in CO2 as it relates to an increase in temperature. If there was a direct relationship an r=1 this is what a graphic of CO2 and temperature would look like:
Figure 1. A perfect positive linear relationship, r = 1
Here is the real world relationship:
With out having the EPA data it is impossible to calculate the r relationship.
It could be a total negative relationship:
Figure 2. A perfect negative linear relationship, r = -1.
It is more likely some factor in between 1 and -1, and could even be a number greater than zero which would indicate some positive relationship. But, we will never know unless the EPA releases the data so that a qualified third party scientist can do the analysis. For now, I think we can conclude they have something to hide.