Determine the squared log-difference between alpha and beta error given a certain chi-square value from central chi-square(df) and a non-central chi-square(df, ncp) distribution.
getErrorDiff(critChiSquare, df, ncp, log.abratio)
squared difference between alpha and beta on a log scale
evaluated chi-squared value
the model degrees of freedom
the non-centrality parameter
log(alpha/beta)