Explained variation

In the statistics of the variance is a measure of how a mathematical model can explain the dispersion (variance) of an empirical data set.

In the linear regression of the Pearson correlation coefficient is clearly referred to as the explained variance.

In the principal components and factor analysis can be any component / assign each factor its contribution to the elucidation of the total variance.

Kent (1983 ) has given a general definition of the variance explained, based on the information measure of Fraser (1965).

Criticism

The term " variance explained " is easily misunderstood because he lets the unspecified subject: who or what is to clarify? Answer: not only the mathematical model, but the mathematical model in conjunction with a specific set of input variables. The explanation of variance is not appropriate to compare the validity of models with different input data and it is not an appropriate measure of the strength of a linear relationship (King, 1990, Achen 1990).