XLSTAT: Student's t test on two independent samples in Excel

 

Dataset for Example

The data are from [Fisher M. (1936), The Use of Multiple Measurements in Taxonomic Problems. Annals of Eugenics, 7, 179 -188] and correspond to 100 Iris flowers, described by four variables (sepal length, sepal width, petal length, petal width) and their species. The original dataset contains 150 flowers and 3 species, but we have isolated for this tutorial the observations belonging to the versicolor and virginica species.

Our goal is to test if for the four variables, there is a clear difference between the two species: versicolor and virginica.

 

Setting Up

• Open XLSTAT.

• Click on the Parametric tests menu and select the Two-sample t and z-test feature:

• Once you've clicked on the button, the dialog box appears.

In the General tab:

• In the Data format field, select One column per variable.

• Choose to run only the Student's t-test.

Note: If we were running the test for one variable only (out of the four), we could also have one column for each specie. The third format is for paired samples (in that case, there must be one column per sample).

• In the Options tab, leave the default options.

• In the Charts tab, activate the dominance diagram.

  

• Click on the OK button to launch the computations.

• The results are displayed on a new Excel sheet.

 

Interpreting the Results

The first results displayed in the XLSTAT report sheet are the summary statistics for each sample followed by the t-test outcome, the distribution graph as well as the comparison plots.

The results for the Sepal Length variables are displayed below.

The comparison plots below summarize in a single graph (e.g. boxplot, scattergram) the t-test results. Hence, we can visualize at the same time the distribution of each sample but also verify the significance of their difference (p-value).