PASS Getting Started Tutorial (Old Version)

In this video, we’ll guide you through a simple sample size calculation using the Two-Sample T-test procedure to highlight some of the important features and capabilities of the software. When you start PASS, the first window that appears is called the PASS Home window. This window is primarily used to find and load power and sample size calculation procedures. You can find any procedure by entering key words in the search bar… or you can locate a procedure using the category tree. As you select each topic in the tree, a list of associated procedures is displayed. You can shorten the list by expanding the tree and selecting subcategories. To find the Two-Sample T-test procedure, we expand Means, Two Independent Means, and then click on T-Test. Now we’re presented with a short list of available procedures associated with this topic. Click on the view buttons to change how this list is displayed. For this demonstration we’ll calculate sample size for the Equal Variance T-Test. The choices indicate that we can specify the effect size using differences or means. Let’s load the procedure that uses differences. The window that now appears is called the Procedure window. This window contains all of the calculation input parameters and options organized on tabs. The important design parameters that directly affect power and sample size calculations are contained on the upper-most tabs, located above the horizontal line. You should consider every option on each of these tabs before running your sample size calculation. The actual options on these tabs vary from procedure to procedure, but the layout will be similar to what you see here. Additional report and plot customization options are provided on the tabs located below the horizontal line. These options do not affect calculations and do not always need to be considered. The options on tabs below the line are about the same in every procedure. The Help Pane provides several resources to help you use PASS effectively. As you mouse over each option, the Option Info section provides in-depth information. If you’d like more help, expand the Help Center. The Help Center contains links to online training videos and the procedure documentation. Each procedure in PASS is fully-documented, with technical details, examples, and validation! If you don’t want to see the options in the Help Center, you can collapse it. Let’s now specify the input parameters for a simple sample size calculation according to an example study design. Each procedure has different input parameters, but as we fill in the options for the T-test, you’ll become familiar with how it’s done generally. First, choose the parameter to solve for based on the other input parameters. In this example, we’ll solve for sample size. Next, leave the hypothesis test direction as two-sided, the default value. Enter 0.8 for power to calculate the sample size required to achieve 80% power. Let’s leave alpha at 0.05 and assume equal group sample size allocation; both are the defaults. For delta, the difference in means to detect, enter 5, and for sigma, the standard deviation of both groups, enter 15. That’s it! We’ve entered all the information needed to perform this sample size calculation. Click the green Calculate button to get the results. The report is displayed in the output window. The Numeric Results Section reports that the sample size required to achieve 80% power for this combination of design parameters is 143 per group, for a total of 286. The report also includes optional literature references, report definitions and a summary statement about the results in the table. There was no plot generated in this case because it would contain only one point. By default, plots are only generated when there are enough data points to produce a meaningful graph.