Talk:Poll

From iA wiki

I did statistics in college, and honestly i have never come across a more ridiculous "science". The whole field is dominated by arbitrarily designated numbers and graphs... If you have 20 poll subjects then the precision of your results is +/- 5% ... What? Who pulled that number out of his ass? It was an interesting (and not too difficult) subject, but you really have to take any kind of poll or statistic with a massive grain of salt. In fact, i think that was probably the first line of the book - even statisticians know their "science" is pretty flakey. Unfortunately there's no real, 100% scientific way to measure a lot of things, and in that case stats is about as good as you get. -- Amw

I know where you are coming from. I tried to study stats but the lecturer was unqualified and I never studied it at high school. I couldn't understand the notation either. I learnt a few tricks for spreadsheets but couldn't use statistics practically afterwards. I should dig up my notes and flesh Statistics out. Another problem is that people heavily pushing an agenda will use a flakey poll, survey and statistics to support their argument. Trying to push a point with a display of causation when none can be proven indicates a weak hypothesis. - ABliss