Loading articles...

You can prove anything with statistics

VANCOUVER (NEWS1130) – Sixty per cent of the time it works every time!  A new report suggests statistics can be used to prove whatever you want them to prove, even complete falsehoods.

Seventy-one per cent of news reporters think the title is too long, but False-Positive Psychology: Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant appears in the latest issue of the journal Psychological Science.

The authors, two scientists at the University of Pennsylvania and one from Berkeley, argue that academic psychologists have too much flexibility with juggling numbers.  They say it is 100 per cent too easy to gather statistically significant evidence to prove a false hypothesis, and those false results can be persistent in scientific literature.

The researchers want new scientific guidelines making study authors disclose everything in an experiment, so we can see anything that might skew results.

By the way, nine out of 10 dentists agree.