Choosing the Variables Needed for Sample-Size Estimation

Ideally, the likelihood of either type of error would be zero, or at least as low as possible. In reality, the possibility of making these errors cannot totally be eliminated, but their likely occurrence can be balanced one against the other. This is done by choosing various combinations of a and p, and therefore various combinations of a and power, since power is defined as 1 minus p. 

5 Choosing the Variables Needed for Sample-Size Estimation 

As noted in Section 9.1, several variables are needed for sample-size estimation, and the researcher can choose the values to be used in the formula that will yield the sample size, N. These are a, p, the estimated treatment effect, and its variance. 

The estimation of a sample size requires several variables to be chosen. In each case, the following need to be selected  

2 The Treatment Effect, Its Variance, and the Standardized Treatment Effect 

---