How do I find statistical significance between a set of six pairs of numbers?

I have six pairs of numbers. The pairs represent the results of two different conditions. I want to know if there is a statistically significant difference between each pair.


These are the pairs, in case you want to see what kind of numbers I am working with:


0.14 --%26gt; 0.17


0.09 --%26gt; 0.13


0.1 --%26gt; 0.196


0.03 --%26gt; 0.088


0.117 --%26gt; 0.217


0.148 --%26gt; 0.151





Since I only have six pairs, is this possible to do?How do I find statistical significance between a set of six pairs of numbers?
I find it hard to believe that the pairs are treated independently. There's no point to that, it doesn't give you any useful statistical information.





However, if you're trying to compare all the values on the left side as a group with their pairs on the right side as a group, then you could use a modified version of the Chi^2 test. Basically multiply everything by 100, and then do a normal Chi^2 test.





Your degrees of freedom will be (6-1)(2-1) = 5How do I find statistical significance between a set of six pairs of numbers?
Actually, Gregola, I think the paired t-test would be better.





t = (average of differences) / (std. error of mean changes)





Then look it up on a t-chart with n-1=6 degrees of freedom.





You shouldn't perform the Chi^2 test because the pairs are correlated. Report Abuse

Try calculating the % difference value for each statistic. If there is a general trend or if there is a significant lack of precision, this can be found by analyzing the values.
The question as posed cannot be answered as we cannot tell if these results on the left were obtained independently from the results on the right? If these are measurements, say, bloodpressure, on individuals before and after a treament (although these numbers could not represent blood pressure as they are too small), then you would need to perform a statistical tests that take into account the correlated nature of the data, such as a paired-sample t-test. Otherwise, you could analyze the data with a regular t-test assuming normality of the underlying distribution.





I would avoid performing any chi-squared test with this sample size.