I think that the data in the link you posted does not support your hypothesis.
As seen in Chapter IV with both the regression results, as well as with the survival
analysis, once conduct-waivered soldiers pass the one year point their attrition rates begin to
worsen and quickly surpass those of their non-waivered counterparts. I believe this is happening
because the soldiers are slowly being released from the controlled environment of basic training
and specialty training. As soldiers are granted more responsibility for their own actions, and no longer are under the watchful eye of the drill instructors, they seem to have the maneuverability to allow their past morally questionable lifestyles to once again affect their decision making. This in turn allows them to make errors of judgment that causes their attrition.