To quote, overinterpreting of statistical significance in studies with small sample sizes ===== This is far too common and sadly even more popular :( Common example that I often given is using simple Mathematics to fool people - you can really play with numbers/percentages and if you mix with credulous data (not credible), you can really play with people as well. For example, 90% of women drivers meet with accidents in their first 6 months of driving - sample size could be just 10 and 9 of them aptly chosen. Coffee drinking is good for heart, tea causes cancer - these are all not fictional headlines :o
Posted on: Thu, 20 Mar 2014 03:38:05 +0000