Why should anyone assume credibility for this type of observational study? It’s just based on data with no experimental design.
Too many people doing "statistical analysis" without the slightest understanding of what statistical science is or does.
"..."However, our findings are quite tentative for mortality because the data are very noisy," he added. "We feel that our study raises important questions about the benefits of mammography screening, but it certainly does not answer them."
More research is needed, he said.
The researchers also warn in their paper that their findings may be limited by so-called ecological bias, which can occur when assumptions are made about individuals from data of a large group.
Uh-huh. Noisy data. They don't know key things about the population. What a stinking piece of garbage. I wonder who paid for that study.