In agriculture, we’re presented with research results all the time. Sometimes it’s agrichemical reps promoting new pesticides. It could be seed people showing better yields. Maybe it’s equipment dealers demonstrating return on investment or fuel savings. Even university research and government extension have been known to flash the occasional graph from time to time.
I was asked to develop a talk for Ontario’s tomato growers. It was intended to give a basic grounding in descriptive statistics. More to the point, it would give the grower the tools they needed to raise the red flag when they felt a presenter was, perhaps, not being completely honest. Without a basic understanding of how data can be abused, we are at the mercy of those presenting the data.
- Did they do their stats correctly?
- Are they hiding something?
- Are they skewing something to make us see what they want?
I often raise my hand during agrichemical company presentations where the next great pesticide is showcased. I’ll ask horrible questions about “n“, standard error bars and the scale used on the graphs… Fair warning.
I have since delivered this talk to many groups, which was surprising, because no one I know loves statistics.
If you’re unable to see this presentation, you may need to update your browser.
Mona Chalabi gives a fun and engaging TedTalk on the subject of sneaky statistics where she advises you on three look-outs before you believe what’s presented: