Syllabus changes for January
Beginning with the January 2010 sitting, the syllabus has been changed slightly in that Pareto, Weibull, lognormal, Beta, and Chi-square distributions are no longer listed on the list of continuous distributions that you need to know. Despite this, some people have said that they still have seen questions on these distributions. In particular, this change doesn’t mean that those distributions can no longer be tested, but instead should mean that they cannot ask questions that require you to have memorized specific facts about those distributions. Other than the Chi-square distribution, I have never thought it was important to memorize anything on these distributions, so as a result, I don’t expect this to change the exam very much. There are a few Chi-square questions in the question bank that I think should be removed, but that doesn’t sound like it has happened, and there a couple lognormal questions that need to be either rephrased.
Of these 5 distributions, the Pareto and Weibull distributions are by far the most commonly tested, but the exam questions have always either given you the density (for a Pareto) or described the distribution as a transformation of an exponential (for a Weibull). In particular, the words Pareto and Weibull have never appeared on 1/P, and those questions are probably going to remain unchanged. The lognormal distribution usually comes up as a transformation of a normal, and again that is still testable as both normal distributions and transformations are testable. Finding the mean and variance of a lognormal only requires knowing the mgf of a normal, so that is still potentially testable but I think unlikely to appear. I think it is still worth watching the lognormal lesson in the seminar to see in practice how these transformations work. Likewise, the Beta distribution has almost always been tested by looking at order statistics of uniform random variables, and knowing about it can still provide you with some useful shortcuts there.