Introduction
PASS provides several modules for power analysis and sample size calculation of regression and correlation, including:

Coefficient Alpha
Coefficient alpha, sometimes called Cronbach’s alpha, is used as a reliability measure. Two procedures are available: one for analyzing a single set of variables and the other for comparing the coefficient alphas from two sets of variables.

Correlation Coefficients
The correlation coefficient, r, is a popular statistic for describing the strength of the relationship between two variables. PASS lets you calculate the sample size for testing a correlation coefficient versus a specific value or for the comparison of two correlations. Power calculations are based on the exact distribution of the correlation coefficient.

Cox Regression
Cox regression is similar to multiple regression except that the dependent variable is a hazard rate.

Intraclass Correlation
The intraclass correlation coefficient is often used as an index of reliability in a measurement study. In these studies, there are N observations made on each of K individuals. These individuals represent a factor observed at random. This design arises when K subjects are each rated by N raters.

Logistic Regression
Logistic regression is similar to multiple regression except that the dependent variable only has two values. Let Y be equal to one if a certain event occurs and equal to zero otherwise. The logistic regression model relates the probability distribution of Y to one or more covariates (X1, X2, ..., Xk). PASS lets you study the power of a test of this relationship. The independent variable may be binary or continuous.

Linear Regression
Use this module to calculate the power and sample size in the case when you have one dependent variable and one independent variable. The results are similar to those for the correlation coefficient, but they are not exactly the same because of the difference in the underlying models (and assumptions).

Poisson Regression
Poisson regression is similar to multiple regression except that the dependent variable is a count.

Multiple Regression
PASS lets you study the power and sample size requirements of various multiple regression F-tests. The tests are put in the context of the R-squared value. When performing a regression analysis, a typical hypothesis involves testing the significance of a subgroup of the independent variables after considering a second, non-overlapping, group of variables. For example, suppose you have five independent variables. One common hypothesis asks whether a certain variable is `important' or `useful' (has a nonzero coefficient in the regression equation) after considering the other four variables. The power of this type of situation is easily examined (with often surprising results).

[Back to PASS] [Back to PASS Procedure List]

Download a 7-day trial copy of PASS 2008

Place an order using our secure online order form