A Course in Statistics with R

Appendix B

The Statistical Tables

Statistical tables have played a very prominent role in the development of the subject, especially before the advent of the computer era. Through powerful software evolution, it may appear to be a very trivial matter. However, the contribution of pioneering Statisticians, such as Sir R.A. Fisher, Karl Pearson, and many eminent others cannot be undermined simply because those tasks can be easily carried out in a software with the user not even being aware of them being in action. The importance of statistical tables may be understood with Horgan (2008), Appendix C, giving the standard normal tables as a useful reference point to her table. In this very brief appendix, we attempt to help the reader to generate some of such tables.

Let us first undertake the task of generating Table A.2 of Snedecor and Cochran (1980) on page 464. The standard density function for -values varying from 0 to 2.99 with a delta of 0.01 is required to generate this table. To set up this, first declare z <- seq(0,2.9,0.1); zpart <- seq(0,0.09,0.01). The half-line simple R code dnorm(outer(z,zpart,“+”)) accomplishes the task, and to obtain the four decimal accuracy, as required on their page 464, simply round off the result with round(dnorm(outer(z,zpart,“+”)),4).

Next, consider setting up Table A.3 from the same reference, page 465. The z needs to be slightly changed to ...