tTest
Syntax
tTest(X, [Y], [mu=0.0], [confLevel=0.95], [equalVar=false])
Arguments
X is a numeric vector indicating the sample for the t-test.
Y (optional) is a numeric vector indicating the second sample for a paired-sample t-test.
mu (optional) is a floating number. If Y is not specified, mu is the mean value of X in the null hypothesis; if Y is specified, mu is the difference in the mean values of X and Y in the null hypothesis.The default value is 0.
confLevel (optional) is a floating number between 0 and 1 indicating the confidence level of the test. The default value is 0.95.
equalVar (optional) is a Boolean value indicating whether the variance of X and Y are the same in the null hypothesis. The default value is false.
Details
If Y is not specified, conduct a one-sample t-test on X. If Y is specified, conduct a paired-sample t-test on X and Y.
Return a dictionary with the following keys:
-
stat: a table with p-value and confidence interval under 3 alternative hypotheses.
-
df: degree of freedom
-
confLevel: confidence level
-
method: type of t-test used
-
tValue: t-stat
Examples
One-sample t-test:
x = norm(10.0, 1.0, 20)
tTest(x, , 10.0);
// output
stat->
alternativeHypothesis pValue lowerBound upperBound
---------------------------- -------- ---------- ----------
true mean is not equal to 10 0.499649 9.68582 10.621998
true mean is less than 10 0.750176 -Infinity 10.540616
true mean is greater than 10 0.249824 9.767202 Infinity
df->19
confLevel->0.95
method->One sample t-test
tValue->0.688192
Paired-sample t-test with equal sample variance:
x = norm(10.0, 1.0, 20)
y = norm(4.0, 1.0, 10)
tTest(x, y, 6.0, , true);
// output
stat->
alternativeHypothesis pValue lowerBound upperBound
------------------------------------ -------- ---------- ----------
difference of mean is not equal to 6 0.438767 5.539812 7.03262
difference of mean is less than 6 0.780616 -Infinity 6.906078
difference of mean is greater than 6 0.219384 5.666354 Infinity
df->28
confLevel->0.95
method->Two sample t-test
tValue->0.785483
Paired-sample t-test with no restriction on sample variance:
x = norm(10.0, 1.0, 20)
y = norm(1.0, 2.0, 10)
tTest(x, y, 9.0);
// output
stat->
alternativeHypothesis pValue lowerBound upperBound
------------------------------ ----------------- ---------- ----------
true difference of mean is n...0.983376 7.752967 10.271656
true difference of mean is l...0.508312 -Infinity 10.04285
true difference of mean is g...0.491688 7.981773 Infinity
df->12.164434
confLevel->0.95
method->Welch two sample t-test
tValue->0.021269