Table of contents

First, the available evidence must be relatively strong in its own terms. Such evidence might include, for instance, time-series, single case studies, or multivariate observational and econometric studies with statistical controls with each of good quality by the standards for such research when it is used to estimate intervention effects. ................
................