Should a Normal Imputation Model be Modified to Impute Skewed Variables
By Paul T. von Hippel. 2013
86 citations recorded by the source corpus.
Researchers often impute continuous variables under an assumption of normality–yet many incomplete variables are skewed. We find that imputing skewed continuous variables under a normal model can lead to bias. The bias is usually mild for popular estimands such as means, standard deviations, and linear regression coefficients, but the bias can be severe for more shape-dependent estimands such as percentiles or the coefficient of skewness. We test several methods for adapting a normal imputation model to accommodate skewness, including methods that transform, truncate, or censor (round) normally imputed values as well as methods that impute values from a quadratic or truncated regression. None of these modifications reliably reduces the biases of the normal model, and some modifications can make the biases much worse. We conclude that, if one has to impute a skewed variable under a normal model, it is usually safest to do so without modifications–unless you are more interested in estimating percentiles and...
Subfields: Artificial Intelligence, Statistics and Probability
View Paper: https://www.semanticscholar.org/paper/74b4a31619809c99866760109e00c34ba8830728
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
- Paul T. von Hippel
- Craig K. Enders
- T. Macías
- A. Jenkins
- Kristin Turney
- Christina J. Cross
- Todd E Bodner
- Amanda C. Gottschall
Citations in the corpus
- New Confidence Intervals and Bias Comparisons Show That Maximum Likelihood Can Beat Multiple Imputation in Small Samples (2013, methodology): “The true small-sample distribution of ML estimates with incomplete data is not exactly t (Little & Rubin, 2002; von Hippel, 2013b) ̂ ∗ 1 (31) This formula can be used whenever we estimate a linear regression.”
- How Many Imputations Do You Need? A Two-stage Calculation Using a Quadratic Rule (2016, background): “5 BMI does not follow a normal distribution, but it doesn’t need to for a normal model to produce approximately unbiased estimates of the mean (von Hippel, 2013).”