Quote:
Originally Posted by Vanilla
Most statistics are pulled out of thin air on the internet. You need to do checks and find out what the significant difference is using various methods. The 95 or 99% confidence level can help but you can get false positives so the most important aspect is to have a decent sized sample size (dependent on the type of data you are measuring). Even peer reviewed journal articles can have flaws with their data because not all researchers are well versed in statistics and can take parts of their data instead of the true results. This is sometimes because they want to create certain results so they can get more government funding.
|
Well I was talking about
actual scientific studies, not just random statistics.
Threaded Mode