Missing data

By John L.P. Thompson, Gilberto Levy. 2004

How researchers see this paper

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

Citations in the corpus

Big Questions for Social Media Big Data: Representativeness, Validity and Other Methodological Pitfalls (2014, background): “In contrast, Facebook’s backbone is mostly an undirected graph in which “friending” is a two-way relationship and requires mutual consent.”
Fear of Crime, Incivilities, and Collective Efficacy in Four Miami Neighborhoods (2013, background): “Allison (2001) indicates that listwise deletion of missing data performs well when the fraction of missing data is small even if these data are MAR instead of MCAR.”
What’s the best way to characterize the relationship between working memory and achievement?: An initial examination of competing theories. (2020, methodology): “Therefore, we elected to correct for missingness using Mplus to maximize sample size and reduce possible bias in estimates (Allison, 2001; Enders, 2010; Rubin, 1987).”
Longitudinal Modeling with Randomly and Systematically Missing Data: A Simulation of Ad Hoc, Maximum Likelihood, and Multiple Imputation Techniques (2003, methodology): “Although FIML gave the best results in this simulation (for both parameter estimates and SEs), MI may be superior when data are not multivariate normal (Allison, 2002).”
The Contradictory Roles of Institutional Status in Retaining Underrepresented Minorities in Biomedical and Behavioral Science Majors (2008, unknown): “The EM algorithm represents a general method for obtaining maximum likelihood estimates when a small proportion of the data is missing (Dempster, Laird, & Rubin, 1977, cited in Allison, 2002; McLachlan & Krishnan, 1997).”
A Comparison of Maximum-Likelihood and Asymptotically Distribution-Free Methods of Treating Incomplete Nonnormal Data (2003, background): “Detailed descriptions and discussions of EM can be found in papers such as those by Dempster, Laird and Rubin (1977), Little and Rubin (1987, 1989/1990), Little and Schenker (1995), and Rubin (1991), as well as a book by Schafer (1997).”
Predicting transition and adjustment to college: biomedical and behavioral science aspirants’ and minority students’ first year of college (2007, methodology): “The EM algorithm represents a general method for obtaining maximum likelihood (ML) estimates when a small proportion of the data is missing (Dempster, Laird, and Rubin, 1977, cited in Allison, 2002; McLachlan and Krishnan, 1997).”
Breaking the Segregation Cycle: Examining Students' Precollege Racial Environments and College Diversity Experiences (2010, methodology): “Unlike conventional regression imputation, in which decisions must be made on which variables to use as predictors, the EM algorithm starts with a full covariance matrix and uses all available variables as predictors for imputing missing data (Allison, 2002).”
Foundations of Organizational Trust: What Matters to Different Stakeholders? (2010, unknown): “4EM imputation is more appropriate than listwise deletion when, as in our data, more than 5% of the cases have one or more missing data points, the data are missing at random, and multivariate normality can be assumed (Allison 2001).”
Robustness of linear mixed‐effects models to violations of distributional assumptions (2020, background): “If missing data depend on covariate values they are called missing at random (MAR) while when missing data depend on themselves (i.e. the response variable) or unobserved predictors, they are called missing not at random (MNAR; Nakagawa, 2015; Nakagawa & Freckleton, 2008).”
Directly Modeling Missing Data in Sequences with RNNs: Improved Classification of Clinical Time Series (2016, background): “While a long and rich literature addresses pattern recognition with missing data (Cohen and Cohen, 1975; Allison, 2001), most of this literature addresses fixed-length feature vectors, as indicated by reviews of the literature from Garćıa-Laencina et al. (2010) and Pigott (2001).”