The t-test divides the difference between means by its standard error and compares the ratio with the t distribution for the available degrees of freedom. The two-sample version compares two groups, such as tool A against tool B, and the Welch form avoids assuming equal variances. The paired version applies when the same units are measured under both conditions, such as the same wafers on two metrology tools, and removes unit-to-unit variation from the comparison.
Use a paired t-test, it's the same 20 wafers on both ellipsometers. Unpaired throws away most of your power.
Also heard as
- Student's t-test
- two-sample t
- paired t
- Welch test
Related terms
p-value
Probability of seeing a result at least this extreme if there were truly no effect, used to judge statistical significance.
ANOVA
Analysis of variance: splitting total variation into parts from each factor and from error, then testing which parts are real.
Power
Probability that a test will detect a real effect of a given size, set by sample size, noise and alpha.
Split lot
Lot of wafers divided between two or more process conditions so the comparison shares the same incoming material.