Modeling longitudinal and multilevel data: Practical issues, applied approaches, and specific examples.

By T. Little, Kai U. Schnabel, J. Baumert. 2000

798 citations recorded by the source corpus.

Subfields: Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/ae8b0de1040ccf396bf6c69dcc3a5cef0e79377b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

Citations in the corpus

A Primer on Maximum Likelihood Algorithms Available for Use With Missing Data (2001, methodology): “However, Wothke (2000) suggested that indefiniteness problems are less pervasive with FIML than with PD.”
The impact of kindergarten learning-related skills on academic trajectories at the end of elementary school (2006, background): “This approach produces parameter estimates that are less biased than listwise deletion, pairwise deletion or mean imputation (Enders, 2001; Wothke, 2000).”
Multivariate longitudinal methods for studying developmental relationships between depression and academic achievement (2007, methodology): “FIML uses all available data to generate maximum likelihood-based sufficient statistics for the covariance matrix and mean vector, is an effective way to handle incomplete data, and is available in most SEM packages (Wothke, 2000).”
Attrition in developmental psychology (2017, background): “…on a probability density function to iteratively maximize the likelihood of the estimates employing, ‘‘ . . . all of the information of the observed data, including mean and variance for the missing portions of a variable, given the observed portion(s) of other variables,’’ (Wothke, 2000, p. 3).”
A Monte Carlo Examination of an MTMM Model With Planned Incomplete Data Structures (2002, methodology): “For a further examination of these methods, see Wothke (2000).”
Early Literacy From a Longitudinal Perspective (2005, background): “An alternative is to estimate the covariance matrix for incomplete data using ‘‘full information maximum likelihood’’ (Anderson, 1957), which yields consistent and efficient solutions in the case that missingness is at random (MAR) (Wothke, 1999).”
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, positive): “…deletion, ML and MI parameter estimates exhibit less bias and greater efficiency in many circumstances (e.g., Arbuckle, 1996; Collins, Schafer, & Kam, 2001; Enders, 2001a, 2001b; Enders & Bandalos, 2001; Graham, Hofer, & MacKinnon, 1996; Graham & Schafer, 1999; Schafer, 1997; Wothke, 2000).”
A Comparison of Maximum-Likelihood and Asymptotically Distribution-Free Methods of Treating Incomplete Nonnormal Data (2003, methodology): “Wothke (2000) compared the abilities of four methods (full information ML, mean imputation, pairwise deletion, and listwise deletion) to reproduce parameter values in a growth curve model involving four measurements per subject.”
Adaptability: How students’ responses to uncertainty and novelty predict their academic and non-academic outcomes. (2013, methodology): “Research has identified potential problems with listwise, pairwise, and similar substitution methods (Graham & Hoffer, 2000), leading to recommendation of the Expectation Maximization (EM) Algorithm, as operationalized in our study using LISREL 8.80 (Jöreskog & Sörbom, 2006).”