It's not misleading at all.
Let's say that 50% of americans are christian, 45% follow miscellaneous other religions, and 5% are atheists.
If we were to posit that criminals come from an even distribution of all religious backgrounds, then when we look at the prison data we would see 50% of all prisoners were christian, 45% other, and 5% atheists, if the hypothesis were correct.
If we were to suggest that christians were more moral and athiests less moral, then we would expect to see something like 5% christian inmates, 45% other, and 50% atheists.
But this isn't what we see. What we see is that almost all prison inmates are christian and a very, very small percentage are atheists. We should be careful not confuse correlation with causation, however we can very easily see that the suggestion that atheist are either less likely to commit major crimes than christians, or are far better at not getting caught.