Find out if the sample mean is significantly different from a certain value.
Find out if there is a significant difference between the two sample means. However, the method is slightly different depending on whether there is a correspondence between the two variables. See the link below for details Functions of stats module
Test if there is a difference in the mean between three or more levels. After finding the F ratio, calculate the p-value from it. Assuming that the population distribution is a homoscedastic normal distribution, the sample distribution of the F ratio (F distribution) is already clear, so the p-value can be obtained by using the cumulative distribution of the F distribution.
If the F ratio is large, it is judged that the effect is larger than the error.
F ratio=\frac{The magnitude of the effect dispersion}{The magnitude of the error variance}
The magnitude of the effect is the distance between the violin plots. Called intergroup variation. The magnitude of the error is the magnitude of the violin plot. Called intra-gun fluctuation.
Test the independence of the data for the contingency table.
\chi^2=\sum_{i=1}^{m}\sum_{j=1}^{n}\frac{(O_{ij}-E_{ij})^2}{E_{ij}}
In the above formula, the contingency table with only the data part is m rows and n columns. O is the observed frequency and E is the expected frequency.
https://qiita.com/kanamae879123/items/ec1226fc6d0ba789ae65 https://qiita.com/kanamae879123/items/2502258737a7d8e181c6
Recommended Posts