Modeling Outcomes With Floor or Ceiling Effects: An Introduction to the Tobit Model

In gifted education research, it is common for outcome variables to exhibit strong floor or ceiling effects due to insufficient range of measurement of many instruments when used with gifted populations. Common statistical methods (e.g., analysis of variance, linear regression) produce biased estima...

Full description

Saved in:
Bibliographic Details
Published in:The Gifted child quarterly 2010-10, Vol.54 (4), p.314-320
Main Author: McBee, Matthew
Format: Article
Language:English
Subjects:
Educational Research
Estimation bias
Evaluation Methods
Gifted
Gifted children
Gifted education
Item Response Theory
Multiple Regression Analysis
Outcomes of Education
Regression (Statistics)
Regression analysis
Simulation
Statistical methods
Structural Elements (Construction)
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In gifted education research, it is common for outcome variables to exhibit strong floor or ceiling effects due to insufficient range of measurement of many instruments when used with gifted populations. Common statistical methods (e.g., analysis of variance, linear regression) produce biased estimates when such effects are present. In practice, it is frequent for researchers to ignore ceiling effects and proceed with traditional analysis. However, the problems caused by ceiling effects are not without possible solutions. This Methodological Brief describes a variation of multiple regression, called the Tobit model, which is capable of correct inference when floor or ceiling effects are present. A brief simulation study illustrates the performance of the Tobit model with a dataset exhibiting a ceiling effect.
ISSN:0016-9862
1934-9041
DOI:10.1177/0016986210379095