Impact factor distribution revisited

We explore the consistency of a new type of frequency distribution, where the corresponding rank distribution is Lavalette distribution. Empirical data of journal impact factors can be well described. This distribution is distinct from Poisson distribution and negative binomial distribution, which were suggested by previous study. By a log transformation, we obtain a bell-shaped distribution, which is then compared to Gaussian and catenary curves. Possible mechanisms behind the shape of impact factor distribution are suggested.