Predictor Sort Sample Size Calculations, Simple Experiment (1 Factor, 2 Levels)

In what follows, you will be asked to provide a value for the correlation between the predictor and the response. If you are not performing a predictor sort experiment, but are just performing a standard pooled T type of experiment, simply set the correlation value to 0.

This program will not explicitly tell you what sample size to use. Instead, you must give it information on the sizes of the differences that you want to be able to detect, on the variability of the response (e.g., modulus of rupture (MOR)), on the correlation between the predictor (e.g., modulus of elasticity (MOE)) and the response, on the significance level that you want to achieve, and on the sample sizes that you are considering.

Given this information, the program will calculate the probabilities that you will be able to detect the differences in which you are interested (power). If these probabilities are too low (say below .90), then you will have to find a predictor that is more correlated with the response, accept a larger significance level (say .10 rather than .05), accept a higher risk of not statistically detecting the differences in which you are interested (lower power), or be willing to consider larger sample sizes.

How many mean differences (diff) do you want to consider?

What are they?
(for example, .10 for a 10% difference in means)

How many coefficients of variation (cv) do you want to consider?
(The coefficent of variation of a property is 100 x (standard deviation)/mean.)

What are they?
(for example, .20 for a 20% coefficient of variation)

What will be the significance level of your tests?

How many sample sizes do you want to consider?

What are they?
(for example, 10 if you want 10 replicates for EACH treatment)

What is the correlation value?
(between -.99 and .99)

What power calculation approach do you want to take?

To use the power tables, the significance level must be .01 or .05. Also, the diff/cv ratio must lie between 0.0 and 3.0, and the number of replicates per treatment must lie between 2 and 7, or the diff/cv ratio must lie between 0.0 and 1.5, and the number of replicates per treatment must lie between 6 and 48. Otherwise, we cannot interpolate within the tables.

If the tables cannot be used, a non-central T approach is automatically taken.

What name do you want for your results file?

The name should be unique to you to prevent the file from being accidentally overwritten by another user.

Since you will see an html document that presents the results as soon as you execute the program, you may not want an ascii copy. However, we do provide this option.

The ascii file will be written in the pub/tt anonymous ftp directory. If a file by the same name already exists in the directory, the new results will be appended to the existing file. Check the anonymous ftp link for directions about retrieving the file.

If ftp does not run on your machine, e-mail me at sverrill@fs.fed.us and ask me to e-mail the results file to you. Do this quickly as the pub/tt directory is cleaned weekly.





For further information, please contact Steve Verrill at sverrill@fs.fed.us
Last modified on 11/24/95.