675

**Teddy Seidenfeld**

Fisher's analysis is a tour de force of so-called ``Goodness of Fit'' statistical tests using to calculate significance levels, i.e., P-values. In this paper I attempt a defense of Mendel's report based on several themes.

- Mendel's experiments include some important sequential design
features that Fisher and (to my knowledge) others ignore. This is
relevant to Fisher's charge that, for one crucial experiment,
Mendel's protocol had approximately a 5% error of misclassifying
hybrids as pure dominants.
- Fisher uses particular statistical techniques of Meta-analysis
for pooling outcomes from different experiments. These methods are
subject to critical debate.
- I speculate on a small modification to Mendelian theory for the model of self-fertilization of garden pea that offers some relief from Fisher's harsh conclusion that, overall, Mendel's data are too good to be true. This last idea is yet another try to find an empirically plausible reason for thinking that Mendel's pea data should reflect sub-binomial variability.

Here is the full postscript text for this technical report. It is 355 bytes long.