In 1975, the federal government set out to standardize and give legitimacy to roadside impairment testing by law enforcement.  In this enterprise, they contracted with a private research company to study available methods and formulate an approach.  That led to the current three test battery which consists of the horizontal gaze nystagmus test (HGN), the walk-and-turn test (WAT), and the one-leg stand test (OLS).  These three were allegedly the best of those available and were considered more reliable.  Notably, that early version excluded the test from being used on people above a certain age and people with weight problems or other medical conditions. These tests were never submitted for peer review and were "validated" with very limited studies.  They were validated as showing that persons with sufficient "clues" would have a BAC of 0.10% or greater.  Later, that validation was claimed for BACs as low as 0.08.  

The National Highway Transportation Safety Administration (NHTSA) claimed that the walk-and-turn test was 79% accurate, the one-leg=stand was 83% accurate.