Hi all, I'm wondering if there is some way to compare the means of 2 diff samples if I have 2 t-values, one for each of the sample. Thanks in advance for any advice!

Printable View

- Nov 10th 2012, 08:42 AMboredaxelComparison of 2 means using 2 t-values
Hi all, I'm wondering if there is some way to compare the means of 2 diff samples if I have 2 t-values, one for each of the sample. Thanks in advance for any advice!

- Nov 10th 2012, 02:30 PMAPrimRe: Comparison of 2 means using 2 t-values
What type of t-values do you have? It depends on the T-test you ran to obtain the value, but if you have some values (such as the standard deviation of the samples, the sample size, etc.) and know what test was run to obtain the t-value (e.g. a one sample t-test), I think you could just set up an algebra equation to obtain the mean. If you don't have any of the other information, I am at a loss. If you are comparing the two groups of samples, than you should have only 1 t-value.

- Nov 10th 2012, 02:58 PMabenderRe: Comparison of 2 means using 2 t-values
Please provide more context. Perhaps you are looking for one of these:

1) A pooled t-test for two population means. The underlying assumptions are population is normally distributed, population variances are equal (or sample variances are reasonably close to one another), and samples were drawn randomly and independently.

2) Separate Variance t-test: Populations normal, indep. random samples. () This test is not as statistically powerful as the pooled test in 1, but it doesn't demand any similar variance assumption.

3) Paired t-test: Say you wish to test a particular diet's efficacy on lowering athletes' mile times. You may pick 20 subjects, and have them race before starting the diet and then 6 weeks after. You are not interested in the differences among the 20 different athletes; you are interested in the diet's effect on mile times of 20 independent sampling units. That is, you are interested in 20 independent time DIFFERENCES. The paired t-test assumes that the time differences are independent and normally distributed.