Multiple imputation of missing covariates with non-linear effects and interactions: an evaluation of statistical methods

نویسندگان

  • Shaun R Seaman
  • Jonathan W Bartlett
  • Ian R White
چکیده

BACKGROUND Multiple imputation is often used for missing data. When a model contains as covariates more than one function of a variable, it is not obvious how best to impute missing values in these covariates. Consider a regression with outcome Y and covariates X and X2. In 'passive imputation' a value X* is imputed for X and then X2 is imputed as (X*)2. A recent proposal is to treat X2 as 'just another variable' (JAV) and impute X and X2 under multivariate normality. METHODS We use simulation to investigate the performance of three methods that can easily be implemented in standard software: 1) linear regression of X on Y to impute X then passive imputation of X2; 2) the same regression but with predictive mean matching (PMM); and 3) JAV. We also investigate the performance of analogous methods when the analysis involves an interaction, and study the theoretical properties of JAV. The application of the methods when complete or incomplete confounders are also present is illustrated using data from the EPIC Study. RESULTS JAV gives consistent estimation when the analysis is linear regression with a quadratic or interaction term and X is missing completely at random. When X is missing at random, JAV may be biased, but this bias is generally less than for passive imputation and PMM. Coverage for JAV was usually good when bias was small. However, in some scenarios with a more pronounced quadratic effect, bias was large and coverage poor. When the analysis was logistic regression, JAV's performance was sometimes very poor. PMM generally improved on passive imputation, in terms of bias and coverage, but did not eliminate the bias. CONCLUSIONS Given the current state of available software, JAV is the best of a set of imperfect imputation methods for linear regression with a quadratic or interaction effect, but should not be used for logistic regression.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Multiple imputation of covariates by fully conditional specification: Accommodating the substantive model

Missing covariate data commonly occur in epidemiological and clinical research, and are often dealt with using multiple imputation. Imputation of partially observed covariates is complicated if the substantive model is non-linear (e.g. Cox proportional hazards model), or contains non-linear (e.g. squared) or interaction terms, and standard software implementations of multiple imputation may imp...

متن کامل

An Empirical Comparison of Performance of the Unified Approach to Linearization of Variance Estimation after Imputation with Some Other Methods

Imputation is one of the most common methods to reduce item non_response effects. Imputation results in a complete data set, and then it is possible to use naϊve estimators. After using most of common imputation methods, mean and total (imputation estimators) are still unbiased. However their variances (imputation variances) are underestimated by naϊve variance estimators. Sampling mechanism an...

متن کامل

Performance evaluation of different estimation methods for missing rainfall data

There are numerous methods to estimate missing values of which some are used depending on the data type and regional climatic characteristics. In this research, part of the monthly precipitation data in Sarab synoptic station, east Azerbaijan province, Iran was randomly considered missing values. In order to study the effectiveness of various methods to estimate missing data, by seven classic s...

متن کامل

Accuracy evaluation of different statistical and geostatistical censored data imputation approaches (Case study: Sari Gunay gold deposit)

Most of the geochemical datasets include missing data with different portions and this may cause a significant problem in geostatistical modeling or multivariate analysis of the data. Therefore, it is common to impute the missing data in most of geochemical studies. In this study, three approaches called half detection (HD), multiple imputation (MI), and the cosimulation based on Markov model 2...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره 12  شماره 

صفحات  -

تاریخ انتشار 2012