Quantifying the dynamics of populations of articles
Some characteristics of dynamic populations are developed and illustrated with some examples.
Most populations of articles change in due course, due to innovation or changing demands of consumers. Existing articles disappear from the market after some time, and new ones are introduced, continuously. Each article has a finite life span. In classical index number theory populations were assumed to be static, that is without change, which in fact means that articles 'live' forever. This ideal situation, however, is quite unrealistic. Especially in our times article populations tend to change constantly, some even dramatically so. The choice of a suitable index formula depends on how dynamic such a population is. Not all index methods are suited for highly dynamic populations. This paper does not go into the choice of a suitable index method for a certain dynamic population. It only deals with the problem of characterizing the dynamics of populations of articles. Several characteristics are proposed, and are applied to a few populations of articles, with different dynamic behaviour. It is shown that certain statistical problems arise when we have to deal with samples of articles, instead of the entire population.