37For the data {(xi,yi)} , i = 1,...,n with error δyi which are fitted to −x∕t f (x;c,t) = ce , the 2 ∑n 2 2 χ (c,t) = i=1(yi − f (xi;c,t)) ∕δyi . The 2 χ ∕dof is normalized to the number of degrees of freedom (dof = degrees of freedom= n − 2 ) which is the number of data points n used in the fit minus the number of fitting parameters (here c , t which makes 2 parameters).