Pilot study
A pilot study is conducted when no measure of effect size is available
A pilot study is a type of observational design used when there is not a measure of effect size published in the literature. Pilot studies are also conducted when a researcher cannot hypothesize a given effect size. Without an effect size, there is no way to conduct an a priori sample size calculation to achieve adequate statistical power. In essence, the researcher is just deciding to collect some observations of an outcome along with predictor variables and risk factors to see what, if any, associations exist.
A good pilot study consists of at least 150-300 observations of an outcome. With fewer observations, the inferences yielded from statistical tests in a pilot study will possess less precision and accuracy. Many selection and observation biases exist in pilot studies.
A pilot study can also be conducted when trying to generate hypotheses or when measuring for rare outcomes.
Most often, the statistical results of a pilot study yield an effect size that is then used in an a priori sample size calculation for a future larger study. When entering into new empirical and clinical areas, pilot studies are often required to properly power studies to detect treatment effects of interest.
A good pilot study consists of at least 150-300 observations of an outcome. With fewer observations, the inferences yielded from statistical tests in a pilot study will possess less precision and accuracy. Many selection and observation biases exist in pilot studies.
A pilot study can also be conducted when trying to generate hypotheses or when measuring for rare outcomes.
Most often, the statistical results of a pilot study yield an effect size that is then used in an a priori sample size calculation for a future larger study. When entering into new empirical and clinical areas, pilot studies are often required to properly power studies to detect treatment effects of interest.
Click on the Statistical Power button to continue.
Statistician For Hire
DO YOU NEED TO HIRE A STATISTICIAN?
Eric Heidel, Ph.D. will provide statistical consulting for your research study at $100/hour. Secure checkout is available with PayPal, Stripe, Venmo, and Zelle.
- Statistical Analysis
- Sample Size Calculations
- Diagnostic Testing and Epidemiological Calculations
- Psychometrics