Statistical power and ordinal outcomes
Ordinal outcomes decrease statistical power and increase the needed sample size
Ordinal outcomes tend to come in the form of Likert-type scales or instruments that have not been fully validated. If ordinal outcomes are normally distributed, then they can be considered to be an interval level measurement, and more powerful parametric statistics can be used.
The decreased power and increased sample size is due decreased precision and accuracy in ordinal measurement. Less powerful non-parametric statistics are used with ordinal outcomes. Larger sample sizes will be needed to detect significant effects with ordinal outcomes.
The decreased power and increased sample size is due decreased precision and accuracy in ordinal measurement. Less powerful non-parametric statistics are used with ordinal outcomes. Larger sample sizes will be needed to detect significant effects with ordinal outcomes.
Ordinal outcomes decrease statistical power and increase the needed sample size.
Click on the Power and Research Design 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