Status Updates From How to Lie with Statistics

3. Don’t include your sample size/keep running your experiment until you get the result you want

4. Standard deviations might mean differences are not statistically significant

Quick and easy read so far.

Nothing earth shattering for today’s readers I don’t think... these tactics are well known and understood by most these days I believe.

Quick recap:

1. How you sample your demographic will give different answers - who is included/excluded

2. Average - mean/median/mode? Use whatever is convenient.

Public pressure and hasty journalism often launch a treatment that is unproved, particularly when the demand is great and the statistical background hazy. So it was with the cold vaccines that were popular some years back and the antihistamines more recently. A good deal of these unsuccessful "cures" sprang from the unreliable nature of the ailment and from a defect of logic.