t-tests
Statistical significance assessments
A t-test is a statistical method for comparing averages between groups to determine whether differences are meaningful or just random variation.
The test produces a t-value that, combined with sample size, indicates statistical significance.
Common forms include the independent t-test (comparing two separate groups) and the paired t-test (comparing the same group before and after): A t-test showed the treatment group improved significantly.
Rejecting a true null
- type II errorFailing to reject falsity
- false positiveIncorrect presence indication
- false negativeIncorrectly absent finding
- alphaFirst Greek letter
beta · sample size · error rate · correction · inflation
Probability under null hypothesis
- standard errorStatistic's expected deviation measure
- confidence intervalEstimated bounds for population parameter
- powerEnergy transfer rate
- probabilityQuantified chance of occurrence
t-distribution · degrees of freedom · significant · nonsignificant
Joined as two
- independentFree from external control
- assumptionsUnproven things taken as true
- normalityState of being typical
- homogeneityState of sameness throughout
parametric · nonparametric · degrees of freedom · sample size · between-subjects · robust
Assumed no relationship exists
- alphaFirst Greek letter
- confidence levelDegree of certainty
- directionalIndicating a path
- nondirectionalWithout preferred orientation
type I error
Rejecting a true null
- type II errorFailing to reject falsity
- false positiveIncorrect presence indication
- false negativeIncorrectly absent finding
alpha · beta · sample size · error rate · correction · inflation
p-value
Probability under null hypothesis
- standard errorStatistic's expected deviation measure
- confidence intervalEstimated bounds for population parameter
- powerEnergy transfer rate
probability · t-distribution · degrees of freedom · significant · nonsignificant
paired
Joined as two
- independentFree from external control
- assumptionsUnproven things taken as true
- normalityState of being typical
homogeneity · parametric · nonparametric · degrees of freedom · sample size · between-subjects · robust
null hypothesis
Assumed no relationship exists
- alphaFirst Greek letter
- confidence levelDegree of certainty
- directionalIndicating a path
- biostatisticsANOVA, biostatistics, confidence interval, p-value, correlation
- statisticsmean, median, mode, variance, standard deviation