During periods when the size of a group of participants in the TM Sidhi Programme exceeded the square root of one per cent of the national population, there was a significant improvement in a comprehensive quality of life index comprising behavioural variables (reduction in the total of homicides, suicides, motor vehicle fatalities, deaths due to other accidents, notifiable diseases, alcohol consumption, and cigarette consumption), controlling for changes in national economic trends, such as unemployment.Ref.Proceedings of the Social Statistics Section of the American Statistical Association (Alexandria, Virginia: American Statistical Association): 38–43, 1996