How do you calculate statistical validity of data?

I want to know what sample size of data I need to make decisions that have a 90% likelihood of accuracy. Can anyone help?How do you calculate statistical validity of data?
that is kinda like accept or reject the hypothesisI think. Check out this site http://www.itl.nist.gov/div898/handbook/鈥?/a>





Cheers!How do you calculate statistical validity of data?
Probability
you could use the chi squared test
We can choose appropriate sample size which suits to our objective of the study. I think following link will help you





http://www.itl.nist.gov/div898/handbook/ppc/section3/ppc333.htm
use this formula...





n = N / [ 1 + N( e ^ 2 ) ]





n = sample size


N = total population


e = margin of error





this formula is called slovin's formula...


since you want your survey or something to be 90% accurate your margin of error should be 10 % or 0.1


the rest follows
As long as you put pretty pie charts and bar graphs there you can make any data look good.





Problem with statistics is you can biase these by asking only specific questions or specific types of people people - which can skew the results.





Guess you can take different factors and give them different weighting i.e point out of a 100 of how accurate they are to increase your chances.





There is also probabality equations you can do as well -i.e. if you roll a dice you will have 1/6 chance of landing on a 2 - if you do roll based on chances of getting an odd number your odds increase to 3/6 (divide this by 3 and you get 1/2) chance of getting an odd number.





generally speaking though statistical data can be manipulated to provide a specific point of view -and you have to take that random factor into account as well