March 2015 Archives

Last time we proved that if the waiting time for an event does not depend upon how long we have already been waiting then it must be exponentially distributed. I promised that there was a lot more to say about such memoryless waiting time processes and in this post we shall begin to investigate them further. Specifically, we shall ask how long must we wait for more than one event to occur?

Whilst we're back on the subject of probability distributions I should like to cover the one that first drew my interest to probability and statistics. Back in my undergraduate days I was chiefly interested in pure mathematics, in deriving mathematical truths from first principles, and my impression of statistics was, rather unfairly in retrospect, that it involved the rote learning of formulae and how to use them rather than proving why they should be used in the first place.
Then I chanced upon a description of a waiting time problem and learned the error of my ways.