Correction for Range Restriction: Lessons from 20 Scenarios
Abstract
Data are often available only for a preselected range-restricted sample in many applied settings. This creates the potential for drawing incorrect inferences and making poor decisions. This is because most inferences and decisions concern the population from which the sample was drawn. Despite these problems, researchers must try to determine statistical values as if the sample were not range-restricted. Although methods for correcting the effects of range restriction have been available for more than a century, often they are not applied or applied incorrectly.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 21, 2021
- Accession Number
- AD1135820
Entities
People
- Malcolm J. Ree
- Thomas R. Carretta
Organizations
- 711th Human Performance Wing