The BICL are approximate Bayes factors (Burnham & Anderson, 2002; Madigan & Raftery, 1994; Raftery, 1995), multiplied by a constant so that the best model (according to BIC) has BICL = 1.
See Akaike weights and Bayes factors (max = 1).
Click the BICL column heading to sort the models according to BICL.