For top ten:
- appropriate interpretation of "negative trials"
- the fallacies of multiple post hoc subgroup analyses
Ph*
On 3/10/12 11:25 AM, "Bastian, Hilda (NIH/NLM/NCBI) [C]"
<[log in to unmask]> wrote:
>G'day!
>
>I'm working on a list of top 3 (with top 5 and top 10 too) basic concepts
>patients, journalists & clinicians need not to understand to protect
>themselves from the data/health claims that commonly mislead. Would
>really appreciate thoughts & all personal favourite lists like this.
>
>At the moment my personal list goes like this:
>
>1. Correlation vs causation
>2. Relative risks without context
>3. Increasing survival rates vs increasing length of life
>
>Thoughts/resources gratefully received. (Have a great weekend!)
>
>Many thanks!
>
>Hilda
|