More efficient: it reduces the part of the experiments variance not related with the characteristics of the populations

More efficient: it allows to reduce the variance because the experiments have less restrictions

When all samples are independent from each other

When the samples are grouped in pairs, one element from the first population and the other from the second population, and this pairs are comparable.

When the response variable is numerical

When the response variable is binary

When the response variable is numerical

When the response variable is binary

Both R-square and adjusted R-square values indicate that it is a good model

Both R-square and adjusted R-square values indicate that it is not a good model

The coefficient is significantly different from 0

El coeficient  is not significantly different from 0

El coeficient  is not significantly different from 0

The coefficient  is significantly different from 0