Nutritional Sciences Professor Discusses Need for Better Data in Obesity Research
John Dawson, an assistant professor in the Department of Nutritional Sciences at Texas Tech University, is an author on a recently published paper examining the common statistical errors and biases in obesity research.
The paper, “Common scientific and statistical errors in obesity research,” is one of a three-part series published by The Obesity Society to identify common mistakes, challenge assumptions about weight loss both among researchers and practitioners and call for greater accountability in how obesity research is done.
Dawson, who is a biostatistician with expertise in statistical analysis, designing experiments and analyzing and interpreting data related to obesity and cancer research, said he and his colleagues on the paper, of which there are 17, wanted to raise these issues so all obesity researchers would produce more informative research. They also wanted to alert research journals to these issues so the journals and organizations would give more consideration to the statistical analysis and have statisticians look over studies for errors.
This is critical in all scientific research, he said, but especially so in obesity research, which can affect huge numbers of the population.
“While some of the topics might be considered esoteric, many of them touch on points that anyone might ask themselves when reading about a new study: ‘Have the researchers confused a mere correlation with a direct cause-and-effect relationship?' ‘Did they tweak the process to get a publishable result?' ‘How were incomplete questionnaires handled?'” he said. “We also discuss less conscious problems, like not providing enough precision in reported results or thorough methodological detail.
“It is our hope that this paper will be a useful reference for scientists preparing experiments, as well as for anyone in the general public who is interested in learning how to critically read new research in order to separate the wheat from the chaff.”
Dawson is available to comment on the need for appropriate methods and statistical analysis in obesity research. For more on the series, click here.
CONTACT: John Dawson, assistant professor, Department of Nutritional Sciences, College of Human Sciences, Texas Tech University, (806) 834-7615 or email@example.com