Comparison of Methods of Handling Missing Time-Invariant Covariates in Latent Growth Models Under the Assumption of Missing Completely at Random

Latent growth models implemented in multilevel models (MLM) or structural equation models (SEM) may be used to analyze longitudinal data with an emphasis on interindividual and intraindividual differences. The main objective of this study is to compare methods of handling missing time-invariant data under the assumption of missing completely at random. Listwise deletion (LD), mean substitution (MS), the expectation-maximization (EM) algorithm, multiple imputation (MI), and full information maximum likelihood (FIML) are compared via a computer simulation study. The findings show that FIML and LD generally perform best, whereas the standard errors of EM and MI are usually underestimated. The results on MS are generally acceptable except when the percentage of missingness is large. Practical implications and directions for future research are discussed.

[1]  David A. Hofmann,et al.  Centering Decisions in Hierarchical Linear Models: Implications for Research in Organizations , 1998 .

[2]  Bengt Muthén,et al.  General Longitudinal Modeling of Individual Differences in Experimental Designs: A Latent Variable Framework for Analysis and Power Estimation , 1997 .

[3]  H. Goldstein.  Multilevel Statistical Models , 2006 .

[4]  Mike W.-L. Cheung,et al.  Evaluating Multilevel Models in Cross-Cultural Research , 2006 .

[5]  Daniel A. Newman.  Longitudinal Modeling with Randomly and Systematically Missing Data: A Simulation of Ad Hoc, Maximum Likelihood, and Multiple Imputation Techniques , 2003 .

[6]  David E. Booth,et al.  Analysis of Incomplete Multivariate Data , 2000, Technometrics.

[7]  Robert E. Ployhart,et al.  THE SUBSTANTIVE NATURE OF PERFORMANCE VARIABILITY: PREDICTING INTERINDIVIDUAL DIFFERENCES IN INTRAINDIVIDUAL PERFORMANCE , 1998 .

[8]  G. A. Marcoulides,et al.  Multilevel Analysis Techniques and Applications , 2002 .

[9]  S. Kozlowski,et al.  Multilevel Theory, Research, a n d M e t h o d s i n Organizations Foundations, Extensions, and New Directions , 2022 .

[10]  J L Schafer,et al.  Multiple Imputation for Multivariate Missing-Data Problems: A Data Analyst's Perspective. , 1998, Multivariate behavioral research.

[11]  Patrick M. O'Malley,et al.  How Academic Achievement, Attitudes, and Behaviors Relate to the Course of Substance Use During Adolescence: A 6-Year, Multiwave National Longitudinal Study , 2003 .

[12]  D. Rubin,et al.  Multiple Imputation for Nonresponse in Surveys , 1989 .

[13]  R. MacCallum,et al.  Studying Multivariate Change Using Multilevel Models and Latent Curve Models. , 1997, Multivariate behavioral research.

[14]  John B. Willett,et al.  Using covariance structure analysis to detect correlates and predictors of individual change over time , 1994 .

[15]  Craig K. Enders,et al.  Using an EM Covariance Matrix to Estimate Structural Equation Models With Missing Data: Choosing an Adjusted Sample Size to Improve the Accuracy of Inferences , 2004 .

[16]  Kenneth A. Bollen,et al.  Latent curve models: A structural equation perspective , 2005 .

[17]  Peter C. M. Molenaar,et al.  A Nonstandard Method for Estimating a Linear G rowth Model in LISREL , 1998 .

[18]  J. Schafer,et al.  Missing data: our view of the state of the art. , 2002, Psychological methods.

[19]  David Chan,et al.  The Conceptualization and Analysis of Change Over Time: An Integrative Approach Incorporating Longitudinal Mean and Covariance Structures Analysis (LMACS) and Multiple Indicator Latent Growth Modeling (MLGM) , 1998 .

[20]  Mike W.-L. Cheung,et al.  Applications of Multilevel Structural Equation Modeling to Cross-Cultural Research , 2005 .

[21]  Kenneth A Bollen,et al.  The role of coding time in estimating and interpreting growth curve models. , 2004, Psychological methods.

[22]  Henry L. Tosi,et al.  Multilevel Theory Building: Benefits, Barriers, and New Developments , 1999 .

[23]  Robert E. Ployhart,et al.  Growth Modeling Using Random Coefficient Models: Model Building, Testing, and Illustrations , 2002 .

[24]  Victoria Savalei,et al.  A Statistically Justified Pairwise ML Method for Incomplete Nonnormal Data: A Comparison With Direct ML and Pairwise ADF , 2005 .

[25]  B. Muthén.  BEYOND SEM: GENERAL LATENT VARIABLE MODELING , 2002 .

[26]  Craig A. Wendorf.  Comparisons of Structural Equation Modeling and Hierarchical Linear Modeling Approaches to Couples' Data , 2002 .

[27]  Jürgen Baumert,et al.  Modeling longitudinal and multilevel data: Practical issues, applied approaches, and specific examples. , 2000 .

[28]  J. Rhodes,et al.  Sleepless in Chicago: tracking the effects of adolescent sleep loss during the middle school years. , 2004, Child development.

[29]  D. Chan.  Functional Relations among Constructs in the Same Content Domain at Different Levels of Analysis: A Typology of Composition Models , 1998 .

[30]  John L.P. Thompson,et al.  Missing data , 2004, Amyotrophic lateral sclerosis and other motor neuron disorders : official publication of the World Federation of Neurology, Research Group on Motor Neuron Diseases.

[31]  Stephen W. Raudenbush,et al.  The Estimation of School Effects , 1995 .

[32]  B. Muthén,et al.  How to Use a Monte Carlo Study to Decide on Sample Size and Determine Power , 2002 .

[33]  Jacob Cohen.  Statistical Power Analysis for the Behavioral Sciences , 1969, The SAGE Encyclopedia of Research Design.

[34]  Daniel J. Bauer.  Estimating Multilevel Linear Models as Structural Equation Models , 2003 .

[35]  William Meredith,et al.  Latent curve analysis , 1990 .

[36]  S. Raudenbush,et al.  Application of Hierarchical Linear Models to Assessing Change , 1987 .

[37]  M J Rovine,et al.  A Structural Modeling Approach to a Multilevel Random Coefficients Model , 2000, Multivariate behavioral research.

[38]  S. Kozlowski,et al.  From Micro to Meso: Critical Steps in Conceptualizing and Conducting Multilevel Research , 2000 .

[39]  Timothy J. Robinson,et al.  Multilevel Analysis: Techniques and Applications , 2002 .

[40]  Thomas H. Ollendick,et al.  COMPARING PERSONAL TRAJECTORIES AND DRAWING CAUSAL INFERENCES FROM LONGITUDINAL DATA , 2008 .

[41]  S G West,et al.  Putting the individual back into individual growth curves. , 2000, Psychological methods.

[42]  Tenko Raykov,et al.  Analysis of Longitudinal Studies With Missing Data Using Covariance Structure Modeling With Full-Information Maximum Likelihood , 2005 .

[43]  Anthony S. Bryk,et al.  Hierarchical Linear Models: Applications and Data Analysis Methods , 1992 .

[44]  Ita G. G. Kreft,et al.  Multilevel Analysis Methods , 1994 .

[45]  L. Collins.  Analysis of longitudinal data: the integration of theoretical model, temporal design, and statistical model. , 2006, Annual review of psychology.

[46]  R. Bosker.  Boekbespreking van "A.S. Bryk & S.W. Raudenbusch - Hierarchical linear models: Applications and data analysis methods" : Sage Publications, Newbury Parki, London/New Delhi 1992 , 1995 .

[47]  Risto Lethonen.  Multilevel Statistical Models (3rd ed.) , 2005 .

[48]  P. Lachenbruch.  Statistical Power Analysis for the Behavioral Sciences (2nd ed.) , 1989 .

[49]  C. R. Rao,et al.  Some statistical methods for comparison of growth curves. , 1958 .

[50]  Jonathon N. Cummings,et al.  Multiple Imputation for Missing Data: Making the most of What you Know , 2003 .

[51]  Craig K. Enders,et al.  A Primer on Maximum Likelihood Algorithms Available for Use With Missing Data , 2001 .

[52]  Patrick J Curran,et al.  Have Multilevel Models Been Structural Equation Models All Along? , 2003, Multivariate behavioral research.

[53]  Karl G. Jöreskog,et al.  Lisrel 8: User's Reference Guide , 1997 .

[54]  Stephen Olejnik,et al.  Treatment Of Missing Data At The Second Level Of Hierarchical Linear Models , 2003 .

[55]  Ledyard R Tucker,et al.  Determination of parameters of a functional relation by factor analysis , 1958 .

[56]  Michael C Neale,et al.  People are variables too: multilevel structural equations modeling. , 2005, Psychological methods.

[57]  Stan Lipovetsky,et al.  Generalized Latent Variable Modeling: Multilevel,Longitudinal, and Structural Equation Models , 2005, Technometrics.

[58]  D. Hofmann.  An Overview of the Logic and Rationale of Hierarchical Linear Models , 1997 .

[59]  P. Roth.  MISSING DATA: A CONCEPTUAL REVIEW FOR APPLIED PSYCHOLOGISTS , 1994 .

[60]  D P MacKinnon,et al.  Maximizing the Usefulness of Data Obtained with Planned Missing Value Patterns: An Application of Maximum Likelihood Procedures. , 1996, Multivariate behavioral research.

[61]  V. Carey,et al.  Mixed-Effects Models in S and S-Plus , 2001 .

[62]  H. S. Booker,et al.  Research Methods for the Behavioral Sciences , 2020 .

[63]  A. Boomsma,et al.  Robustness Studies in Covariance Structure Modeling , 1998 .

[64]  David P Mackinnon,et al.  Investigation of Mediational Processes Using Parallel Process Latent Growth Curve Modeling , 2003, Structural equation modeling : a multidisciplinary journal.

[65]  C. Ramey,et al.  Modeling Intraindividual Changes in Children's Social Skills at Home and at School: A Multivariate Latent Growth Approach to Understanding Between-Settings Differences in Children's Social Skill Development , 2000, Multivariate behavioral research.

[66]  Peter M. Bentler,et al.  Comparisons of Two Statistical Approaches to Study Growth Curves: The Multilevel Model and the Latent Curve Analysis. , 1998 .

[67]  Brian D. Ripley,et al.  Modern applied statistics with S, 4th Edition , 2002, Statistics and computing.

[68]  J. Schafer,et al.  A comparison of inclusive and restrictive strategies in modern missing data procedures. , 2001, Psychological methods.

[69]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .