Chapter 10. Bayes: Adding to What You Know Now
When presented new information, we have no other option than to relate it to what we already know—there is no blank space in our minds within which new information can be stored so as not to "contaminate" it with existing information.
In the first semester of business statistics, students learn a few methods based on a few "simplifying" assumptions. Often the assumptions don't end up simplifying very much of anything and some assumptions, like the assumption of a normal distribution, can turn out to be disastrously wrong. Later on in statistics, students learn about some more "advanced" methods that, to me, always seemed much more intuitive than the earlier ones.
One of the key assumptions in most introduction-to-statistics courses is that the only thing you ever knew about a population are the samples you are about to take. In fact, this is virtually never true in real-life situations.
Imagine that you are sampling several sales reps about whether an advertising campaign had anything to do with recent sales. You would like to measure the campaign's "contribution to sales." One way would be simply to poll all of your sales team. But you have more information than just what they reveal. You had some knowledge prior to the poll based on historical experience with sales and advertising. You have knowledge about current seasonal effects on sales as well as ...