A Review of Methods for Missing Data - University of Chicago

Educational Research and Evaluation1380-3611/01/0704-353$ , Vol. 7, No. 4, pp. 353 383#Swets & ZeitlingerA Review of Methods for Missing DataTherese D. PigottLoyola University Chicago , Wilmette, IL, USAABSTRACTThis paper reviews Methods for handling Missing data in a research study. Many researchers usead hoc Methods such as complete case analysis, available case analysis (pairwise deletion), orsingle-value imputation. Though these Methods are easily implemented, they require assump-tions about the data that rarely hold in practice. Model-based Methods such as maximumlikelihood using the EM algorithm and multiple imputation hold more promise for dealing withdif culties caused by Missing data . While model-based Methods require specialized computerprograms and assumptions about the nature of the Missing data , these Methods are appropriatefor a wider range of situations than the more commonly used ad hoc Methods . The paperprovides an illustration of the Methods using data from an intervention study designed toincrease students' ability to control their asthma researchers have faced the problem of Missing quantitative data at somepoint in their work.

Educational Research and Evaluation 1380-3611/01/0704-353$16.00 2001, Vol. 7, No. 4, pp. 353–383 # Swets & Zeitlinger A Review of Methods for Missing Data Therese D ...

Tags:

  Data, Evaluation, Missing, Missing data

Information

Related search queries