0

I would like to test if two means are significantly different from each other. I am using R and I do this:

t.test(SampleA,SampleB)

and I get output that looks like:

Welch Two Sample t-test

data:  SampleA and SampleB
t = -4.4195, df = 84.117, p-value = .0197
alternative hypothesis: true difference in means is not equal to 0

95 percent confidence interval:
 -16.214653  -6.160959

 sample estimates:
mean of x mean of y 
 6.44 18.25

My conclusion is that the means are significantly different because the p-value is < .025. I am using .025 as comparison because this is a two tailed test at the 95% confidence level. Is my conclusion correct?

Also - can you explain when to use a paired t-test. The test above is not a paired t test.

joesyc
  • 163
  • 1
  • 6
  • Welcome to our site. It is highly likely that an elementary question like this already has good answers. Please consider conducting searches on [paired t test](http://stats.stackexchange.com/search?q=paired+t+test) and [p-value](http://stats.stackexchange.com/questions/tagged/p-value?sort=votes&pageSize=30). – whuber Mar 27 '15 at 21:30
  • 1
    R's `t.test()` function automatically takes the two-tailed nature of your hypothesis test into account. So you don't need to compare your p-value to .025. If you wanted a one-tailed test, you would use `alternative="greater"`, eg. – gung - Reinstate Monica Mar 27 '15 at 21:41
  • Thanks GUNG - In order to accept or reject the null then I should compare the p-value to .05? If p-value <.05 null="" reject="" the="" then=""> – joesyc Mar 27 '15 at 21:43
  • only if your significance level is 0.05. (One doesn't do hypothesis tests at "95% confidence level", by the way. There's no such animal.) – Glen_b Mar 28 '15 at 01:32

0 Answers0