Indicator and stratification methods for missing explanatory variables in multiple linear regression

By Michael P. Jones. 1996

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

Missing Data Imputation for Supervised Learning (2016, background): “One-hot encoding for missing data yields biased estimates when the features are correlated, which is often the case with survey data, even when data are MCAR (Jones, 1996).”
Intimate Partner Violence: The Influence of Job Opportunities for Men and Women (2019, background): “The study chooses to drop missing observations rather than, for example, choosing to code missing values under a new variable (i.e. an indicator for missing religion) due to concerns that inclusion of such a variable can bias estimates (Jones, 1996).”
Missing data (2019, unknown): “Unfortunately, Little and Rubin (2002) have shown that the Heckman model is exquisitely sensitive to the normality assumption on e in the regression equation.”
Working With Missing Values (2005, background): “Table 2 summarizes the results and average margin of error for b weights, and Table 3”
Regression Modeling Strategies (2003, background): “ Including missing data but adding a category ‘missing’ causes serious biases [1, 104, 189] Problem acute when values missing because subject too sick Difficult to interpret”
Missing data techniques for structural equation modeling. (2003, methodology): “It is well known that the use of inappropriate methods for handling missing data can lead to bias in parameter estimates (Jones, 1996), bias in standard errors and test statistics (Glasser, 1964), and inefficient use of the data (Afifi & Elashoff, 1966).”
Much Ado About Nothing (2007, background): “These adhoc approaches have the potential to induce bias and are not recommended (Jones 1996; Greenland and Finkle 1995).”
Missing covariate data in clinical research: when and when not to use the missing-indicator method for analysis (2012, background): “Furthermore, the first association is based on a complete case analysis, so this association is unbiased only if missingness is conditionally independent of the outcome.(9,10) But, given the nature of nonrandomized studies, in which covariates are commonly mutually related, the missing-indicator met”
A Review of Methods for Missing Data (2001, unknown): “Jones (1996), however, examines the use of missing-indicator variables, ﬁnding that this method results in the overestimation of the residual variance of the regression.”
Regression with Imputed Covariates: A Generalized Missing Indicator Approach (2011, unknown): “Although frequently used in practice, this approach is known to produce biased estimates (Jones, 1996; Horton and Kleinman, 2007).”
Combining Multiple Imputation and Inverse-Probability Weighting (2012, methodology): “The missing indicator method can cause bias when used for variables in an analysis model (Jones, 1996).”
Site Selection Bias in Program Evaluation (2015, unknown): “In earlier drafts, I instead used the missing indicator method, which is only unbiased under stronger assumptions (Jones 1996) but gives very similar results.”
The Struggle to Pass Algebra: Online vs. Face-to-Face Credit Recovery for At-Risk Urban Students (2017, background): “, the percent of students missing prior achievement data was almost identical in the treatment and control groups) and the missing data should not bias impact estimates (Jones, 1996).”
Challenges With Propensity Score Strategies in a High-Dimensional Setting and a Potential Alternative (2011, methodology): “For each categorical predictor with missing data we simply added an additional category representing “value missing” (for a helpful discussion, see Jones, 1996 ).”