HOW TO IMPUTE INTERACTIONS, SQUARES, AND OTHER TRANSFORMED VARIABLES

By Paul T. von Hippel. 2009

365 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/7fade7f65335d34e394e652f5b4d425b0f7f5e10

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

The Effect of Gender and Race Intersectionality on Student Learning Outcomes In Engineering (2015, background): “Researchers suggest that creating interaction variables first and then imputing yields good regression estimates, when the regression model includes interaction (Allison, 2003; Von Hippel, 2009)squares, and other transformed variables</ title>”
Fitting multilevel multivariate models with missing data in responses and covariates that may include interactions and non‐linear terms (2014, neutral): “This approach was criticized by Von Hippel (2009), who instead proposed simply treating each interaction or polynomial term as ‘just another variable’.”
Multiple imputation of covariates by fully conditional specification: Accommodating the substantive model (2012, neutral): “Second, we imputed the missing X values using the ‘transform then impute’ or ‘just another variable’ (JAV) approach proposed by [15], that is, by treating X(2) as another variable to be imputed in the ice command in Stata.”
New Confidence Intervals and Bias Comparisons Show That Maximum Likelihood Can Beat Multiple Imputation in Small Samples (2013, background): “The potential disadvantage is that bias can be introduced by unnoticed conflicts between the assumptions of the imputation model and the assumptions of the analytic model (e.g., von Hippel, 2009, 2013a).”
The role of adolescents’ hopeful futures in predicting positive and negative developmental trajectories: Findings from the 4-H Study of Positive Youth Development (2011, background): “With this proportion of missing values, statisticians and developmental methodologists endorse multiple imputation (MI), recognizing the limitations of other methods (eg, listwise deletion) that significantly reduce sample size and thereby potentially bias results (Allison, 2002 ; Jelicic et al, 20”
Should a Normal Imputation Model be Modified to Impute Skewed Variables (2013, background): “…words: missing data; missing values; incomplete data; regression; transformation; normalization; multiple imputation; imputation Paul T. von Hippel is Assistant Professor, LBJ School of Public Affairs, University of Texas, 2315 Red River, Box Y, Austin, TX 78712, paulvonhippel.utaustin@gmail.com.”
Multiple imputation as a flexible tool for missing data handling in clinical research. (2017, methodology): “However, this so-called impute-then-transform approach (von Hippel, 2009) pushes the product term coefficient toward zero because the imputation phase fails to model the interactive effect (i.e., an imputation model based solely on the lower-order terms explicitly assumes that b5 equals zero).”
Multiple Imputation of Missing Data for Multilevel Models (2018, background): “As an alternative, product terms may be regarded as “just another variable” (von Hippel, 2009).”
Multiple Imputation Strategies for Multiple Group Structural Equation Models (2011, methodology): “Finally, von Hippel (2009) used analytic methods to study the impact of different imputation models for interaction terms (as noted previously, he uses the phrase transform, then impute to refer to PTI and uses stratify, then impute to denote SGI).”