Bayesian Analysis of Modular GCSE Science Scores

This paper shows the correct way to analyse module test scores when comparisons between classes are required.  It assumes a priori that such a comparison is valid, which of course it may not be.  The software executable (Windows only) can be obtained zipped-up with a README file by sending me an email.  (This is purely so that I am in touch with anyone who is interested in using the ideas - the software is freely available for non-commercial use.)  The paper is 'submitted' and will be linked-to from here (you may need an Athens login to access it once published).

Retaking Modular GCSE Tests

Done some stats and found out (using several assumptions) that ...
* The probability of gaining a grade on a re-sit runs from 22% (for U) to 17% (for D).
* The chance of gaining a grade within the five possible re-sits is about 71% (for U) to 60% (for D).

You can download some blurb explaining the stats behind this (so you can check me!) and the Excel file I cobbled together to get the numbers.