Null hypothesis is that the difference between the two groups is 0.
Alternative hypothesis - the difference between two means is not zero.
Test statistics can be found by the formula
where x
1, x
2 - are the means of two samples, SD
1,2 - standard deviations , n
1,2 - the number of observations of two samples respectively.
Then you have to calculate the number of degrees of freedom:
![https://latex.codecogs.com/gif.latex?d.f.%20=%20%5Cfrac%7B%28SD_1%5E2/n_1%20+%20SD%5E2_2%20/n_2%29%5E2%7D%7B%28SD%5E2_1/n_1%29%5E2/%28n_1%20-1%29%20+%20%28SD%5E2_2/n_2%29%5E2/%28n_2%20-1%29%7D](https://latex.codecogs.com/gif.latex?d.f.%20=%20%5Cfrac%7B%28SD_1%5E2/n_1%20+%20SD%5E2_2%20/n_2%29%5E2%7D%7B%28SD%5E2_1/n_1%29%5E2/%28n_1%20-1%29%20+%20%28SD%5E2_2/n_2%29%5E2/%28n_2%20-1%29%7D)
and find p-value from the tables (or using t to p value calculator).
If p-value is more than 0.1 (performing two-sided test you must use doubled alpha), we do not reject the null hypothesis; otherwise - we reject it.
Comments
Leave a comment