Statistical Ethics and Challenging Substantial Errors in Statistical Methods and Results in a Prominent Peer Reviewed Economics Journal

• Activity Number: 342 - SPEED: Sports to Fire: Fascinating Applications of Statistics
• Type: Contributed
• Date/Time: Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Consulting
• Author(s): Chris Barker*
• Companies: Statistical Planning and Analysis Services, Inc.
• Keywords: ethics; economics; substantial published statistical errors; clinicaltrials.gov

Abstract

Effectively challenging substantial errors in statistical analyses in published literature is difficult. This paper challenges substantial errors and biases in a publication in a prominent economics journal - the American Economic Review (AER) and prominent websites. The topic, long-term investment in research, used an example of oncology clinical trials. Key steps, seeking advice from the ASA ethics committee who suggested publishing the results. A couple prominent statisticians, several medical oncologists and several Health economists also suggested to publish. Rather than clinicaltrials.gov ("CTG") the authors used a database of oncology clinical trials called "PDQ", SEER and others. The manuscript states that PDQ "was not intended for research". This paper includes detailed summaries of oncology clinical trial information from CTG. compared to PDQ. The article documents the biases in use of PDQ vs. CTG. Aside from the use of PDQ were other problems and in consistencies with the original AER paper. The session provides key summaries comparing PDQ with CTG and lessons learned.