Statistical Techniques

Abstract

In statistics, there are two schools of thought about the exact nature of probability, one being the Frequentist school which defines probability as ‘the ratio of the times the event occurs in a test series to the total number of trials in the series’ [1], or the ‘frequencies of outcomes in random experiments’ [2] the other being the Bayesian school of thought which defines probability as ‘a measure of the degree of belief that an event will occur’ [1]. In my work in cosmological parameter inference and model selection I adopt a Bayesian approach, partly as it allows for a more holistic approach to solving problems involving probability, and partly as it allows us to explicitly include in the probability, information based on our prior experience and physical understanding of the situation. Rather than a collection of statistical tests, a Bayesian approach to probability provides a more flexible framework within which to construct solutions to problems of parameter inference and model selection. Standard texts describing Bayeisan statistical methods used in this research include [3-6] for Bayesian statistics in the cosmological context see especially [7-9]