I'm doing the CA2 worked example and I notice they do the chi square test the long way (calculating A-E, chi square value, degrees of freedom etc) when you can get the same result using the CHITEST formula in excel. Is this formula frowned upon or is it okay to use it? Thanks
I can't comment on this from the point of view of the examiners. However, I would say that Excel is not a sound tool for statistical analysis. For example, older versions of Excel calculated standard deviations using the single pass method, which is inaccurate for sets of large data points. More generally, numerical algorithms are almost never specified and generally, seem to be derived directly from theoretical results, with no consideration given to practical numerical effects. There is a good summary and set of references here: http://www.practicalstats.com/xlsstats/excelstats.html It is also worth having a read of this: http://www.burns-stat.com/pages/Tutor/spreadsheet_addiction.html
