Description Usage Arguments Value Author(s) References See Also Examples
In the t test, it is usually the case that the assumption of equal variances on the two groups is violated. The test problem is known as the Behrens-Fisher (BF) problem when no assumption of equal population variances can be made. For the BF problem, the T statistic provides value for a given dataset and its statistical distribution is not easy to characterise.
To our knowledge, the best approximation thus far is due to Welch (1938). The Welch<e2><80><99>s test involves two layers of approximations: approximating the distribution of the statistic by a t-distribution, which in turn depends on an approximate degrees of freedom.
The Monte Carlo based tests improve upon Welch<e2><80><99>s approximate test by avoiding one layer of approximation, resulting in enhancement in statistical power than Welch's t-approximation.
1 | mcBFtest(x, y, method, MC)
|
x |
a (non-empty) numeric vector of data values |
y |
a (non-empty) numeric vector of data values |
method |
if "t" is used, we will use t-test assuming equal variance and df=n+m-2. If "W" is used, we assume unequal variance and the Welch approximation is used. If "Monte Carlo", the Monte Carlo procedure is applied. |
MC |
a number for Monte Carlo procedure |
The function returns a list including
p.value |
the p-value for the test. |
You-Gan Wang, Na Wang
Welch, B.L. (1938). The significance of the difference between two means when the population variances are unequal. Biometrika, 29 (3/4), 350<e2><80><93>362.
Ullah, I., Paul, S., Hong, Z., & Wang, Y-G. (2019). Significance tests for analyzing gene expression data with small sample sizes. Bioinformatics, in press.
t.test
function from package stats
1 2 3 4 5 6 7 8 |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.