t test
2024. 10. 25. 13:15ㆍStatistics
The t test is a type of estimation test that determines whether the difference between the means of two data sets is significantly different or not.
Limitations and Assumptions: Data for performing a t test must meet the following conditions:
● ramdom sampling
● independent measurements or observations
● normal distributions
● equal variance
t-test methodology and classification according to hypothesis rejection
The t test basically assumes equal variances, but in the case of an unpaired sample t test that compares samples with different sample sizes or population means, the results of the F test (variance comparison) must be checked.
'Statistics' 카테고리의 다른 글
Confidence interval (0) | 2024.10.25 |
---|---|
Probability distribution (0) | 2024.10.25 |
Test statistics (0) | 2024.10.25 |
Hypothesis testing procedure (0) | 2024.10.25 |
Test method according to null hypothesis (0) | 2024.10.25 |