I want to use the t-test on samples grabbed from performing a procedure in three slightly different ways. These are very correlated. I have gotten several samples of results from all 3 of these. I've generated their means, standard deviations, confidence intervals, etc.
How do I now run the student's t-test on it?
Your question is an example of the wrong way to ask a statistics question! Do not ask "...How do I use the X-test", but describe yout data, your experiment, and the research question you want to solve. maybe then we can fins an answer, and maybe that answer do not inculde an t-test.
The procedure you follow to collect the data from the processes that you are observing, define a mathematical model. When you perform a hypothesis test, you are asking questions about the parameters of the model. A t-test refers to the t-distribution, a statistical distribution that pops up frequently when asking questions about model parameters.
I would suggest that you pull out your favorite elementary statistics book and read about one-way analysis of variance, keeping in mind the procedure that must be followed for this version of an ANOVA to be legitimate.
Since the Hypothesis Testing is being done on different values of the same variable, the technique that is used is paired/dependant t-test.
Here is the formula : (Observed Mean Difference - Expected Mean Difference)/Standard Error
Step 1 : You need to calculate the mean difference between the two variables of all the observations. This is the Observed Mean Difference. Step 2: If you assume that the Null Hypothesis is true, the Expected Mean Difference is to be taken as 0. Step 3: Standard Error is the variation in the standard deviation due to chance. This is calculated by dividing the standard deviation of the population by the sample size. Step 4: Use these to compute the t-value.
