A simulation study using EFA and CFA programs based the impact of missing data on test dimensionality

Authors:
Shin-Feng Chen;Shuyi Wang;Chen-Yuan Chen
Affiliations:
Department of Education, National Pingtung University of Education, No. 4-18, Ming Shen Rd., Pingtung 90003, Taiwan;Department of Measurement, Statistics and Evaluation, University of Maryland, College Park, MD 20742, USA;Department and Graduate School of Computer Science, National Pingtung University of Education, No. 4-18, Ming Shen Rd., Pingtung 90003, Taiwan and Global Earth Observation and Data Analysis Center ...
Venue:
Expert Systems with Applications: An International Journal
Year:
2012

Citing 6
Cited 0

Statistical analysis with missing data

Statistical analysis with missing data
Modeling and control for nonlinear structural systems via a NN-based approach

Expert Systems with Applications: An International Journal
The stability of an oceanic structure with T-S fuzzy models

Mathematics and Computers in Simulation
Application of data mining to the spatial heterogeneity of foreclosed mortgages

Expert Systems with Applications: An International Journal
Stability analysis and robustness design of nonlinear systems: An NN-based approach

Applied Soft Computing
Modeling, control, and stability analysis for time-delay TLP systems using the fuzzy Lyapunov method

Neural Computing and Applications

Quantified Score

Hi-index	12.05

Visualization

Abstract

This study examines the impact of missing rates and data imputation methods on test dimensionality. We consider how missing rate levels (10%, 20%, 30%, and 50%) and the six missed data imputation methods (Listwise, Serial Mean, Linear Interpolation, Linear Trend, EM, and Regression) affect the structure of a test. A simulation study is conducted using the SPSS 15.0 EFA and CFA programs. The EFA results for the six methods are similar, and all results obtained two factors. The CFA results also fit the hypothesized two factor structure model for all six methods. However, we observed that the EM method fits the EFA results relatively well. When the percentage of missing data is less than 20%, the impact of the imputation methods on test dimensionality is not statistically significant. The Serial Mean and Linear Trend methods are suggested for use when the percentage of missing data is greater than 30%.