Missing data: our view of the state of the art.

By Joseph L. Schafer, J. Graham. Psychological methods, 2002

11,843 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/f867e56d3a8ac2bbef6bfe5c74ee80ab43239718

How researchers see this paper

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

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

The effectiveness and relative importance of choice in the classroom. (2010, methodology): “Unlike previous studies, this study experimentally manipulated the provision of choice in the context of real classrooms with real students and teachers.”
Advances in Geometric Morphometrics (2009, background): “There exists a large body of literature on methods for estimating population parameters such as means or regressions in the presence of missing values (for reviews and examples see Schafer 1997; McLachlan and Krishnan 1997; Schafer and Graham 2002).”
Social competence, social support, and academic achievement in minority, low-income, urban elementary school children. (2008, methodology): “It should be noted that the imputation procedure uses the SEM algorithm and assumes that data are missing at random (Dempster, Laird, & Rubin, 1977; Schafer & Graham, 2002).”
Servant Leadership Behaviour Scale: A hierarchical model and test of construct validity (2011, methodology): “The Expectation Maximization (EM) approach in SPSS for Windows was used to impute missing data prior to analysis (Schafer & Graham, 2002).”
Does impulsivity relate to perceived dependence on and actual use of the mobile phone (2007, methodology): “This method is reasonable when a scale has an alpha greater than 0.70 (Schafer & Graham, 2002).”
Degrees of Freedom in Planning, Running, Analyzing, and Reporting Psychological Studies: A Checklist to Avoid p-Hacking (2016, methodology): “Missing data can be dealt with by listwise deletion, pairwise deletion, multiple imputation, full information methods, and other methods (Schafer and Graham, 2002).”
Biased attention and dysphoria: Manipulating selective attention reduces subsequent depressive symptoms (2010, methodology): “Namely, our method of estimating missing data (FIML) is considered a ‘‘state-of-the-art’’ approach (Schafer & Graham, 2002).”
Changing the ideological roots of prejudice: Longitudinal effects of ethnic intergroup contact on social dominance orientation (2014, methodology): “…procedures has been shown to produce more reliable parameter estimates and standard errors as compared to conventional methods of dealing with missing data, for example, pair-wise or list-wise deletion (Schafer, 1997; Schafer & Graham, 2002; see also Al Ramiah & Hewstone, 2012; Swart et al., 2011).”
Machine Learning and Decision Support in Critical Care (2016, methodology): “[107] published an algorithm for performing expectation–maximization (EM) with missing data, and this represented a fundamental shift of thought among statisticians from removing missing data as a nuisance toward averaging over the uncertainty caused by missing data [106].”
The Big-Fish-Little-Pond Effect in Mathematics (2014, methodology): “To take into account the missing data (less than 2% for all survey items and none for achievement test scores), we relied on full information maximum likelihood (Graham, 2009; Schafer & Graham, 2002).”
Comparison of Performance of Data Imputation Methods for Numeric Dataset (2019, methodology): “MNAR is often considered as worst missing type as it may lead to biased result whereas MCAR and MAR might lead to loss of statistical power (Graham 2009; Schafer and Graham 2002).”
The design of simulation studies in medical statistics (2006, background): “The empirical SE should be close to the average of the estimated within simulation SE if the estimates are unbiased [21] and therefore, it may be sensible to consider both estimates of uncertainty.”
The Role of Spirituality Versus Religiosity in Adolescent Psychosocial Adjustment (2006, methodology): “Methodological research has demonstrated that this method of dealing with missing data is preferable to more common methods such as pair-wise deletion, list-wise deletion, or mean substitution (see Schafer and Graham, 2002 ).”
Predicting cardiovascular risk in England and Wales: prospective derivation and validation of QRISK2 (2008, methodology): “Multiple imputation is a statistical technique designed to reduce the biases that can occur in “complete case” analysis along with a substantial loss of power and precision.”