James Ming Chen, Weighted-Average Methodologies for Evaluating Bar Examination Passage Rates, available at http://ssrn.com/abstract=2532800:
There are few truly “national” law schools in the United States. Most American law schools in the United States have a “dominant” state bar. A greater number of the graduates of nearly any law school take the bar examination administered by one state than any other bar examination. The American Bar Association and U.S. News and World Report's law school rankings rely on bar passage rates for the single largest cohort within any school’s graduating class. But the modal passage rate is misleading as a measure of any one school’s overall bar passage rates. The modal passage rate also fails to facilitate direct comparisons of bar examination performance at different schools.
To evaluate the overall bar examination performance of the graduates of any law school, I propose the use of weighted-average methodologies. Ideally, we should be able to measure, by use of weighted averages, each school’s bar passage z-score. Since the data needed to conduct proper standard scoring is extremely difficult, if not impossible, to procure, I propose a simplified weighted-average methodology. The weighted average of school-specific bar passage rates by jurisdiction, minus the weighted average of passage rates from all jurisdictions where its graduates, enables us to evaluate each school's bar exam performance, relative to the bar passage rate in its modal state, and relative to the weighted average bar passage rate across the entire United States.
In the interest of completeness, I propose two other methodologies. One of those methodologies is based on ratios, in emulation of U.S. News and World Report’s law school rankings. The other is based on what I consider a reasonable parametric estimate of standard deviations in state-wide bar passage rates weighted by school, through which we can estimate z-scores for bar exam passage rates for all states and all schools.