memex/daily/2016-11-17.org

2016-11-17

• Quants support

Quants support   academia UoN

Describe your new note here.

Use ANOVA when you have more than two groups

Trimming is giving the outlier the highest normal value

Signifiance in Shapiro/Smirnov normality tests is not a good thing. It means the data is not normal (parametric)

For an indepdendent sample-t test you need homogeneity of variance (calculated with Laveene (looks at homogeniety of variance) test and Bartlets (looks at equality of variance) test)

For a Paired Sample's t-test you need

Sphereicity is data making sense, test that's the variants between participats is not too far apart

*en_GB