Imputation methods are popular for the handling of missing data in psychology. The methods generally consist of predicting missing data based on observed data, yielding a complete data set that is amiable to standard statistical analyses. In the context of Bayesian factor analysis, this article compares imputation under an unrestricted multivariate normal model (Multiple Imputation [MI]) to imputation under the statistical model of interest (Data Augmentation [DA]). The former method is popular in applied research, but the latter method is more straightforward from a Bayesian perspective. Simulations demonstrate that DA yields less-biased parameter estimates for moderate sample sizes and high missingness proportions. MI, however, yields less-biased parameter estimates for large sample sizes with misspecified models. The incorporation of auxiliary variables in DA is also addressed, and BUGS code is provided.
AitkinM.AitkinI.
(2005).
Bayesian inference for factor scores. In
Maydeu-OlivaresA.McArdleJ. J. (Eds.), Contemporary psychometrics: A festschrift for Roderick P. McDonald.Mahwah, NJ: Lawrence Erlbaum Associates, 207–222.
2.
CasellaG.GeorgeE.(1992). Explaining the Gibbs sampler. The American Statistician. 46,167–174.
3.
CollinsL. M.SchaferJ. L.KamC.-M.(2001). A comparison of inclusive and restrictive strategies in modern missing data procedures. Psychological Methods. 6,330–351.
4.
GelfandA. E.SmithA. F. M.(1990). Sampling-based approaches to calculating marginal densities. Journal of the American Statistical Association. 85,398–409.
5.
GelmanA.RubinD. B.(1992). Inference from iterative simulation using multiple sequences (with discussion). Statistical Science. 7,457–511.
6.
GenzA.BretzF.HothornT.(2006).
mvtnorm: Multivariate normal and T distribution. (R package version 0.7–5).
HolzingerK. J.SwinefordF. A.(1939).
A study of factor analysis: The stability of a bi-factor solution (No. 48).Chicago: University of Chicago Press.
9.
JennrichR. I. (2006). What do factor analysis loadings and their standard errors estimate: The signed permutation problem. Presented at the 71st annual meeting of the Psychometric Society, Montreal, Quebec.
10.
JohnsonM. S.JunkerB. W.(2003). Using data augmentation and Markov chain Monte Carlo for the estimation of unfolding response models. Journal of Educational and Behavioral Statistics. 28,195–230.
11.
KimJ. K.FullerW.(2004). Fractional hot deck imputation. Biometrika. 91,559–578.
12.
LanzaS. T.CollinsL. M.SchaferJ. L.FlahertyB. P.(2005). Using data augmentation to obtain standard errors and conduct hypothesis tests in latent class and latent transition analysis. Psychological Methods. 10,84–100.
13.
LeeS.-Y.(2007).
Structural equation modelling: A Bayesian approach.Chichester: Wiley.
14.
LittleR. J. A.RubinD. B.(1987).
Statistical analysis with missing data.New York: Wiley.
15.
LittleR. J. A.RubinD. B.(2002).
Statistical analysis with missing data (2nd ed.).
New York: Wiley.
16.
LunnD. J.ThomasA.BestN.SpiegelhalterD.(2000). WinBUGS—A Bayesian modelling framework: Concepts, structure, and extensibility. Statistics and Computing. 10,325–337.
MartinA. D.QuinnK. M. (2006). MCMCpack: Markov chain Monte Carlo (MCMC) package(R package version 0.7–2).
19.
MengX.-L.(1994). Multiple-imputation inferences with uncongenial sources of input. Statistical Science. 9,538–558.
20.
NovoA. A.SchaferJ. L. (2002). norm: Analysis of multivariate normal datasets with missing values(R package version 1.0–9).
21.
PlummerM.BestN.CowlesK.VinesK.(2006).
coda: Output analysis and diagnostics for MCMC(R package version 0.10–7).
22.
RubinD. B.(1976). Inference and missing data. Biometrika. 63,581–592.
23.
RubinD. B.(1987).
Multiple imputation for nonresponse in surveys.New York: Wiley.
24.
SchaferJ. L.(1997).
Analysis of incomplete multivariate data.London: Chapman & Hall.
25.
SchaferJ. L.GrahamJ. W.(2002). Missing data: Our view of the state of the art. Psychological Methods. 7,147–177.
26.
ScheinesR.HoijtinkH.BoomsmaA.(1999). Bayesian estimation and testing of structural equation models. Psychometrika. 64,37–52.
27.
SinharayS.SternH. S.RussellD.(2001). The use of multiple imputation for the analysis of missing data. Psychological Methods. 6,317–329.
28.
SturtzS.LiggesU.GelmanA.(2005). R2WinBUGS: A package for running WinBUGS from R. Journal of Statistical Software. 12,1–16.
29.
TannerM. A.WongW. H.(1987). The calculation of posterior distributions by data augmentation. Journal of the American Statistical Association. 82,528–540.
30.
ThomasA.O’HaraB.LiggesU.SturtzS.(2006). Making BUGS open. R News. 6,12–17.
31.
van BuurenS.Groothuis-OudshoornK.(in press). MICE: Multivariate imputation by chained equations in R. Journal of Statistical Software.
32.
WothkeW.(2000).
Longitudinal and multi-group modeling with missing data.
LittleT. D.SchnabelK. U.BaumertJ. (Eds.), Modeling longitudinal and multilevel data: Practical issues, applied approaches, and specific examples.Mahwah, NJ: Lawrence Erlbaum Associates.
33.
ZhangZ.McArdleJ. J.WangL.HamagamiF.(2008). A SAS interface for Bayesian analysis with WinBUGS. Structural Equation Modeling. 15,705–728.
