Generating Random Numbers for Experiments

If you have to run field experiments where you need to put trees out or experiments in a Complete Randomized Design or in a Randomized Block Design, you will be challenged with generating numbers to represent treatments in random order. If you only have a few treatments and replications, it can be done simply with pieces of paper that you randomly pick out of a hat. However, if you have several treatments and several blocks or replicates, you may want to refer to a random numbers table. These can be found in the back of a statistics book or if you are using Excel Version 7.0 (in Office 97) or later, try using the statistics package. I will try over the next several months to give you pointers in using this useful package for simple field experiments. Many of you may be setting up experiments now, so I will give pointers on the Random Numbers Table

RANDOM NUMBER GENERATION

This is found in your Tools menu of Excel. If you don't have it you will need to install it.

Hit the help button and search for the word statistics. Hit Install the Analysis ToolPak (it will help you through the procedure). When finished, you should be able to go to the Tools tab in your menu and select Data Analysis. You will see a list of simple statistical functions that Excel can perform, including ANOVA (Analysis Of Variance) as well as the Random Numbers Generation. Hit the random numbers generation.

I will keep this simple as the many options in this tool aren't as useful as producing a uniform table. The options offered are that Distribution should be UNIFORM. Then for Number of Variables, consider this to be the number of columns on a grid you want. I use this in a Randomized Block Design as my Reps or Blocks number. It will ask you to select the Number of Random Numbers. On the same grid of your worksheet, this represents the number of rows of numbers, ie., the number of treatments of your experiment. The range should be equivalent to the number of treatments. For example if you had 6 treatments then the range is 1 to 6. In the example I have provided, the thinning experiment has 7 reps (variables) and 6 thinning treatments (# random numbers). It will ask you to select where you want the output. You can select a worksheet dedicated to the grid or just give a cell range on the current worksheet you are working on. The output will generate numbers in multiple decimal places. Simplify the output by Formating cells (Format on menu) to 0 numbers. The output is now in whole numbers. The output will produce double entries and missing entries. Adjust these by hand and you are finished.

In the example below, I have adjusted the first 3 reps and left the remaining ones as raw output.

Rep 1 Rep 2 Rep 3 Rep 4 Rep 5 Rep 6 Rep 7
1 2 6 3 2 1 5
6 3 3 2 5 1 1
4 4 2 2 3 5 5
5 6 1 2 1 5 1
2 5 4 5 4 2 5
3 1 5 2 4 5 3

1 = Control
2 = Thinner 1
3 = Thinner 2
4 = Thinner 1 with water only
5 = Thinner 2 with water only
6 = Hand Thinning

I hope the above helps. It certainly has helped me in setting up rootstock trials.

Ron Perry, MSU Hort