The article presents a new way to compare two lognormal variances using Bayesian hypothesis testing. By using the intrinsic Bayes factor and the fractional Bayes factor, the researchers have found a solution to the problem of arbitrary constants in Bayes factors. Their approach helps in making more accurate model selections and hypothesis testing. The results of the study show that these two Bayes factors can be effectively used for comparing lognormal variances in various scenarios.