Contents of the supplement

1. PROGRAM CODES FOR OCTAVE

"H1CIP.m"
This code calculates confidence intervals (CI) by a new method "Correlation Imposed Permutation", CIP,
with explanations in the code. Correlation and regression for the parameters R, dif and rr

"H1Efron.m"
A code for bivariate bootstrapping to calculate CI for the parameters R and dif

"H1z"
A code to plot Fisher's z for P(H1)

"H0regr.m"
A code for the randomization test. Correlation and regression for the parameters R, dif, rr 
and hits in Fisher's tea tasting experiment (partition 4, 4 and known).

"H0booty.m"
A code for bootstrapping y (outcome) for the parameter R or for hits in the tea tasting experiment 
when the partition of the 8 cups is variable and not communicated.

"H0z.m"
A code to plot P(H0) for z


2. PROGRAM CODES FOR R

"H1BCa.r"
A code from the internet "boot Package in R" by Jeremy Albright, adapted to the present data structure


3. DATA

"4fi4.txt"
Data to apply permutation and bootstrapping on Fisher's tea tasting experiment in the appendix

"83E.txt"
Outcome y = hail energy (MJ) at ground of 83 experimantal days.

"83sc.txt"
Treatment x = seeding, variable 0 <= sc <= 1.

"83scbin.txt"
Treatment x = seeding, binary variable 0 or 1.

"253E.txt"
Outcome y = hail energy (MJ) at ground of 253 experimental cells.

"253sc.txt"
Treatment x = seeding, variable 0 <= sc <= 1.

"253scbin.txt"
Treatment x = seeding, binary variable 0 or 1.

"253Ebin.txt"
Binary outcome y. Hail =1, no hail =0


4. SREADSHEET

"DataHail-FMA.xls"
Spreadsheet with the most important data from Grossversuch IV, containing codes for cleaning up 
and a quick randomization test based on the first 4 moments.