Roosevelt economists challenge value of using randomized tests
A Roosevelt University economics professor and his 24-year-old student are breaking new ground with a comprehensive analysis of the use of randomized testing in medicine and economics, concluding the method of experimentation is neither effective nor ethical.
Roosevelt Economics Professor Stephen T. Ziliak (above right) and Edward R. Teather-Posadas, a 2013 Roosevelt master’s in economics graduate who currently teaches Principles of Micro and Macro Economics to MBA students at the University, are critical of the use of randomized testing in an article to be published in the forthcoming book, The Oxford University Press Handbook on Professional Economic Ethics.
Featuring contributions by Nobel laureate economists as well as best-selling author N. N. Taleb and luminary Deirdre McCloskey, the book due out later this year is expected to be of interest globally to economists, philosophers, social scientists and media figures who write about economics.
The article by Ziliak and Teather-Posadas, entitled “The Unprincipled Randomization Principle in Economics and Medicine” will be the first ever to analyze and challenge use of randomization in testing, long considered to be the gold standard for measuring causal effects.
“In many cases we have found that the economics, statistics and ethics involved in designing random clinical trials and tests leave a lot to be desired,” said Ziliak, lead author of the critically acclaimed 2008 book, The Cult of Statistical Significance: How the Standard Error Costs Us Jobs, Justice and Lives. “We expect our research to be a game changer in the way we think about random testing and we believe our findings have the potential to positively impact real people who have been victims of drug trials and other random-test experiments,” he said.
The Roosevelt economists first met in 2012 in a Rhetoric of the Human Sciences class at the University where Teather-Posadas became interested in Ziliak’s theories about statistical significance and the opportunity to expand the work into a larger arena examining history and outcomes of randomization. Since then, the two have been working together to evaluate its impact as outlined in dozens of articles published by the American Economics Review from 2000-2009 and in the New England Journal of Medicine from 2000-2002.
Highlighting examples of ways in which randomized testing has done more harm than good, the two Roosevelt instructors are critical, for instance, of a recent major development-economics study (sponsored by the World Bank, National Institutes of Health and Spencer Foundation) in which thousands of poor Chinese youths with poor eyesight were given eyeglasses while a control group of thousands with poor eyesight went without help for years in order to test the theory that children with access to eyeglasses do better in school.
“Why would anyone want to deny thousands of elementary school students who were known to have eyesight problems access to eyeglasses?” asked Ziliak. “These types of experiments shouldn’t be about simply fitting subjects willy-nilly into a formula so that results can be obtained. They should be about making good decisions in order to arrive at sound conclusions,” he said.
Among other examples, the two were also critical of randomized testing methods used by the federal government in its infamous Tuskegee Syphilis Experiment, conducted between 1931 and 1972, in which hundreds of African American men in Macon County, Ala., as part of a control group, were deliberately not treated to prove that syphilis was deadly and use of penicillin was useful.
“The damage done daily by randomized trials in the United States is (equally) large – and growing,” the article by the two to be published in coming months by Oxford University Press states.
During their research, Ziliak and Teather-Posadas reviewed the history of experiment design going as far back as the 1890s and they also surveyed and analyzed the way that contemporary economists and medical researchers are designing their experiments in order to arrive at their own conclusions.
“We found that using stratification and a balanced approach when developing test samples is far more ethical, efficient and precise than simply rolling the dice to populate clinical trials and sample groups,” remarked Teather-Posadas. “We believe that it is better to choose sample groups for studies very deliberately (by age, gender, height, weight, class, ethnicity, etc.) rather than randomly,” he said.
The budding economist, who is younger than all of the students he currently is teaching at Roosevelt, hopes to pursue a PhD in economics beginning in the fall.
“I chose Roosevelt’s economics program because it is one of the few in the country that is based on the heterodox principle in which students are encouraged to take a pluralistic view of economics when looking at the world and analyzing situations,” he said. “I wanted that kind of expansive viewpoint before I begin to narrow my approach in theorizing about and looking at problems and solutions as a doctoral economics student.”
Last updated 01/09/2015