- Clinical Trial Solutions
- Learning Center
Explore additional benefits of nQuery
Today study designs and statistical methods have never been more varied or more complex. Yet, one of the central questions when designing your study remains the same. What is the appropriate sample size for my study?
What Is The Appropriate Sample Size For My Study?
At nQuery we have spent over 20 years adding hundreds of individual sample size tables to make sure that no matter what choice you make when designing your study, you can answer that question. So whether you're planning to do a simple t-test analysis, you want to undertake a complex simulation of a survival study when planning it, or you wish to take advantage of interim monitoring in a group sequential design. You can be confident that our sample size software will provide you with all the statistical tools you need to find the sample size for your study.
Find The Appropriate Sample Size Method For My Study
In nQuery finding the appropriate sample size method for your study is easy. Simply go to select test design and goal window and select the relevant options for your study. All your options will now appear below. So whether selecting the type of data which would be analyzed, the number of groups being compared or the type of analysis being conducted, finding the sample size method for your study is an easy task. You can also use the search bar to search for the table name that's appropriate for your study.
Powerful Plotting Features - Power Vs Sample Size Plots
After you've completed your sample size determination, nQuery provides a number of powerful options for you to further explore your results. One of these options is the plotting features. The two most commonly used plots are the power versus sample size plot and the plot user selected rows plot. The power versus sample size plot allows you to quickly and automatically see what the effect of changing sample size will be on the power in one or more columns. This then creates an interactive plot where you can get a deeper understanding of the effect of varying the sample size would have.
Powerful Plotting Features - Plot User Selected Rows
With the plot user selected rows feature, you can visually explore the effect of changing any of your assumptions over a range of values and seeing what effect it would have on any of the solvers in the table.
For example, in the two sample t-test we can see the effect of changing anything from the significance level to the standard deviation on any of the solvers in the table which in this table would be for; the effect size, sample size or power. We can then specify a range of values that we wish to look at. And also the number of values we wish to see. This then creates an interactive graph that we can use to explore our results. In this case seeing the effect of changing the significance level on the sample size per group.
nQuery also provides a number of specialized plotting options for specific tables. For example, in the two sample survival simulation table, you will be provided with a survival vs. time plot that describes the scenario being used in the simulation. Another example is in a group sequential table you'll be provided with a boundary plot which gives the boundaries of rejection for futility or efficacy at each time point.
Specify Multiple Factors
The specify multiple factors tool allows you to quickly and efficiently explore a number of scenarios in your sample size table. In the specify multiple factors tool you can enter one or more values for each of the assumptions required in that table. If you enter multiple values for a particular assumption then each of these will be used in the table above and if this is done for more than one assumption then all combinations of all the values will be included in the main table.
When determining the sample size that you need for your trial to produce a statistical significant result, you may sometimes not have a direct estimate of a required value for the sample size calculation. However, you may have access to additional information which could be used to derive this value. In nQuery in these scenarios we provide side tables.
For example, in a one way analysis of variance using nQuery if we had three groups and we want to derive a value for the variance of means if we select the variance of means row the effects size side table will appear below the main table. In this table we could enter the expected means in each of the three groups and derive a value for the variance of means by entering the values, selecting compute and then clicking transfer to place it in the main table.
In our sample size software a large number of tables provide these types of side tables, to allow you to quickly get from results that you are likely to have, to those that are required for your sample size determination. These range from effects size side tables like shown here to those that specify a covariance matrix in a repeated measure design to a complex survival simulation scenario.
Given the high failure rates and the increased costs of clinical trials, researchers need innovative design strategies to best optimize financial resources and reduce the risk to patients