![]() |
![]() |
|
| HOME | ABOUT US | CONTACT US | NEWSLETTER | ADVERTISE |
|
Articles
Statisticians
Biostatistician
Statistical Analysis
Statistical Consulting
Statistical Reports
Statistics Consultant
Statistical Quality Control
Research Design
Market Research Design Consulting
Survey Research Design
Experimental Research Design
Research Design Strategies
Time Series Research Design
Quantitative Research Design
Research Design Analysis
Clinical Trial Design
Clinical Research Trials
Mathematical Modeling
Statistical Modeling
Financial Statistical Modeling
Model Building
Time Series Analysis
Interrupted Time Series Design
Time Series Forecasting
Forecasting Methods
Quantitative Forecasting
Regression Analysis
Survival Analysis
Prediction Models
Sample Size Calculation
Power Calculation
Logistic Regression
Generalized Estimating Equations
Linear Regression Models
|
Statisticians
Power CalculationA power calculation is a necessary step for any research project and is required for almost all grant applications. The purpose of a power calculation is to determine an estimate of how many participants or observations are required in order to generate statistically significant results. There are three aspects which are used in deriving your power calculation. Power Calculation InformationThe research design is built around testing the research hypothesis. While this does not empirically determine your sample size requirements, it provides the road map to how your research is structured. Differences in research design, such as having two or three groups, will change the number of participants needed. The greater the expected variability in your outcome measure, the fewer participants you will need. You can base your estimate of the variability on past studies that are similar to the one you want to conduct. If you are breaking new ground and do not have a literature review on which to make an estimate, you can run a pilot study when appropriate. You will also need to know how big of a change you are looking for. Because variability occurs in nature, you need to be looking for a measured change that is larger than would occur by chance. Try to figure out what percent of change would produce the result that you are looking for. ![]() Get all Market Research articles via
|
![]() |
v. 5.0164 © 2002 - 2008 Article Insider. All Rights Reserved. Privacy Policy | ![]() |





