vald   38333
invd   7
min   199.132
max   135523497.889
mean   10667.066
stdev   812291.8

Degrees of freedom for a generic variable `X` are defined as:

`df[\bar{X}_{agkt}] =\ (M-1)(1 + 1/(M+1) \frac{1/M sum_{l=1}^M V^{(l)} [\hat{X}_{agkt}]}{B[\bar{X}_{agkt}]})^2`

where `V^{(l)}[\hat{X}_{agkt}]` and `B[\bar{X}_{agkt}]` are respectively the within implicate variance and the between implicate variance of `X`.