A.M. Bruning et al. (see ibid., vol.3, no.4, p.567-76, 1996) have presented an application of the dimensional effect expression to data obtained from a life test carried out under thermal stress and humidity. The basis of this approach is the correct identification of the probability distribution function of failure times, having performed proof tests under voltage for thermal life inference. The purpose of this discussion is to make use of the extremely detailed results of these proof tests to show an application of statistical fitting tests to the above probability functions, as well as to a truncated distribution, called 'like-Weibull', which can give advantageous information, Moreover, the limits of application of the dimensional effect formula will be pointed out.