Stack Exchange Network

Stack Exchange network consists of 174 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

Multiple imputation refers to a set of stochastic imputation routines aimed at preserving the multivariate features of the data

2
votes
1answer
23 views

t test of individual coefficient and wald test of euqality of two coefficients

I've run a regression $Y = a + \beta_1X_1 + \beta_2X_2 + \epsilon$, and I have an interest in testing which coefficient of $X_1$ and $X_2$ has a stronger impact on $Y$. Here are the parameter ...
2
votes
0answers
33 views

Why don't people impute missing exposure data in database studies?

Investigators doing studies in large databases (e.g., EMR) in which there is often a lot of missing data usually (in my experience) want to exclude all subjects missing the exposure or outcome of ...
0
votes
0answers
7 views

Multiple Imputation on multi-site data

Suppose I would like to assess the relationship between Y and X (i.e. Y|X) on data collected from several different sites (i.e. 5) with one covariate Z and multiple auxiliary covariates which may ...
1
vote
0answers
10 views

Choosing Among Multiply Imputed Datasets

I am using multiple imputation to estimate treatment effects in a dataset that contains missing data. In some of my imputed datasets, the algorithm used in the analysis fails to converge; it's not ...
0
votes
0answers
19 views

Flag outliers first or conduct multiple imputation first?

I am working with a data set in which the dependent variable, Y, is constructed from three variables (y1, y2, y3) that each have missing data. To address this issue with multiple imputation, I've used ...
1
vote
0answers
31 views

When to use Multiple Imputation over Maximum Likelihood for Missing Data and vice-versa?

I've seen these being called the best techniques for dealing with missing data. But I'm wondering when to use one over the other and why? Edit: Why is this getting downvotes? I'm legitimately ...
0
votes
0answers
17 views

Imputing missing data with MICE where each observation has different levels

I have a set of observations that each consists of different levels. For example, I ask a $P$ individuals $N$ questions, each question with a possible $k_n$ discrete responses. This produces a table ...
1
vote
1answer
20 views

Per protocol or Imputation when missing is small (<5%)

if ~2% of my data is missing on the outcome (continuous scale), out of a total of 200, two in control and three in intervention group, do I need to impute? Or can I make a case that with such small ...
0
votes
1answer
62 views

Predictive Mean Matching as Single Imputation?

Multiple imputation is known to be advantageous compared to single imputation. However, in practice there are often non-statistical reasons why multiple imputation can not be used (e.g. the data ...
0
votes
0answers
22 views

Multiple imputation with composite variables

In my analyses, I often use urinary concentrations as measure of exposure to various compounds. As these are generally spot urines, they are 'adjusted' for dilution using urinary creatinine ...
0
votes
0answers
15 views

multilevel multiple imputation

I have been using the mitml package in R for multilevel multiple imputation of longitudinal data, there are 2 functions available for imputation, panImpute and jomoImpute , both use the joint modeling ...
0
votes
0answers
27 views

Procedure for identifying predictors via LASSO on imputed data

First time poster here, hello Everyone! I've tried to make my question as concise as possible. I am looking to identify the best set of predictors for rehospitalisation in my data set of 60 ...
0
votes
0answers
24 views

Multiple Imputation of Multilevel data

I am using Mice package in R for multiple imputation of a multilevel data where repeated measures are nested within individuals. But there is a bug in mice for which we need to convert the group ...
0
votes
0answers
12 views

Multiple Imputation in R; for one variable only certain date ranges

So I have a problem and I'm not sure there's a way to do this. I have 13 survey questions, and there are responses from 2008 to 2016. However, 3 of the items were only asked starting in 2009. Only ...
1
vote
1answer
54 views

Multiple imputation with a Cox model [closed]

I'm doing a study on 52 patients with breast cancers and looking for predictive factors of death. I have 17 variables to test for predictive value with some with 20% of missing data (some categorical ...
0
votes
0answers
43 views

Combining Gradient boosted trees after multiple imputation

Currently I am working with a gradient boosted tree model fit onto a multiple imputed dataset. For those who don't know multiple imputation: It predicts missing values and imputes that value with ...
0
votes
0answers
17 views

Combining multiple imputation with penalization

I have used Frank Harrell's excellent rms package both for penalization (pentrace and setting ...
0
votes
1answer
20 views

Hypothesis testing with randomised algorithm

Imagine I have some data D, and with a randomised algorithm I construct a parameter $\theta$. I know the distribution of $\theta$ under the null hypothesis of the data. For example, the original data ...
0
votes
0answers
19 views

Multiple imputation: How to obtain a consolidated data set from the m imputed data set

Sometimes it is necessary to compare results in contingency tables of the original data (without missings) with a single final dataset obtained from the m imputed datasets, but taking into account the ...
0
votes
0answers
9 views

validity of tobit estimates after multiple imputation

I want to estimate tobit marginal effects using multiply imputed data, however I see that tobit is not among the estimation commands supported by Stata's MI prefix - I understand that the validity of ...
1
vote
0answers
34 views

Is there a way to estimate regression coefficients?

I'm currently working on a simulation study (based on empirical data) and for this simulation I created a model with multiple interaction terms. The interaction terms are between categorical variables,...
0
votes
0answers
73 views

Generating frequency table and survival curve after multiple imputation

I'm using the MICE package to generate 10 imputed datasets. After that, I know I should perform analysis on each dataset (propensity score matching, Chi-square, and Cox-regression in my case) and ...
1
vote
1answer
34 views

What is the risk of not including all of my model predictors in an imputation process?

I have a model that I want to run but a lot of my predictors have missing values. So I ran an imputation process using mice package in R - it took me 22.5 hours! ...
2
votes
0answers
26 views

How to deal with undetectable outcome values? (data missing not at random)

I conducted a sound propagation experiment in which recorded maned wolves calls were broadcasted at different sites(x3), hours(x6: 17h,18h,23h,05h,06h,11h), and with different speaker position (x2: ...
2
votes
1answer
62 views

Forecast (impute) missing discrete values in multiple time series

I'm looking to forecast (impute) missing discrete values in multiple time series in order to reach a target volume in a consolidated time serie. The context: I have salesmen that are selling ...
0
votes
0answers
9 views

Error Detection and missing data imputation in Wireless Sensor Networks

I'm working on wireless sensor networks and I wish to be capable of detecting if there are any outliers in the sensed data as well as imputing missing ones. I read a lot of articles which made me more ...
0
votes
0answers
16 views

Is it valid to use Random Forest imputation in blocks and combine results into a final dataset?

I have an extremely large dataset (26 variables and 105,556 observations) with missing values and I would like to use Random Forest imputation to impute some missing values. Since the dataset is so ...
0
votes
0answers
34 views

Larger standard error after multiple imputations

Missing data experts, I am working on a presentation demonstrating the benefits of using multiple imputations to handle data missingness. To demonstrate its benefits, I simulate two data sets; one ...
0
votes
0answers
260 views

MCAR test for large number variables and small sample size

I have a dataset with 101 observations and 402 columns (those columns comprise several multiple-item questionnaires). Among those 402 columns, 10 of them are categorical and the remaining are ...
3
votes
1answer
81 views

Likelihood ratio test for multiply imputed datasets?

I have a set of generalized linear models fit to 5 multiply imputed datasets. I am interested in testing the statistical significance of a set of predictors, coded from dummy variables. Rubin's ...
4
votes
0answers
120 views

When using multiple imputation, is it more informative to report descriptive statistics for the imputed dataset, the original dataset or both?

My understanding is that multiple imputation is a method for dealing with bias/lack of power that results from missing data. However, it is not a method for replacing individual values. As a result, I ...
0
votes
0answers
99 views

Rubin's formula for variance in varying domains

I am using multiple imputation on a binary variable (employment status). I have to estimate the number of employed units in several different domains, and the respective variances using Rubin's rule: ...
2
votes
1answer
70 views

Binary classification with multiple imputations

I work on a binary classification problem with proteomic data, where the goal is to select the best subset of proteins which contribute to better classification (AUC). However, the problem is that ...
1
vote
0answers
59 views

Restricting a sample on a multiply imputed variable

Apologies if this is more of a statistical question. I am currently dealing with a multiple imputation problem that I am attempting to address in Stata. After the imputation stage, I would like to ...
0
votes
0answers
51 views

Evaluating a Fractional Logit model after multiple imputation

In the model I'm estimating the dependent variable falls between 0 and 1 so I'm using a fractional logit model. However, because of issues with the data I had to multiply impute it, and use ...
1
vote
0answers
187 views

Imputation using MICE: Use the train data to impute the missing test data

I'm using mice in R to impute missing values. If I understand correctly, mice specifies a fully conditional model to draw new ...
2
votes
0answers
25 views

To impute or not - community consensus for reporting accuracy of an imputed model

I have a model generated using an imputed data set with imputation accuracy of 75%. If the model using imputed data has an accuracy of 80% What would be the community consensus to report the ...
2
votes
1answer
232 views

Difference in success of mice imputation across variables

I'm trying to use mice package to do imputations, I'm pretty new to all of this. I have 9 continuous variables and I used the following code to impute, using the 'cart' (Classification and regression ...
0
votes
0answers
19 views

In risk prediction models, should development and validation be separated prior to imputation

Suppose you want to develop and (internally) validate a risk prediction model on a data set with missing data. Should you: 1) Separate your development and validation cohorts from the beginning, and ...
0
votes
0answers
83 views

Multiple imputation when explained variance of imputation model is low

I am thinking about using multiple imputation to impute missing values on the dependent variable (continuous) in my analysis. In the imputation model, I considered variables that are associated with ...
1
vote
0answers
23 views

Multiple imputation when have more than 1 outcome variable

Is there a good paper or reference for doing multiple imputation when there is more than one outcome variable? Anything that specifically addresses building the imputation model or software to use for ...
1
vote
0answers
27 views

Choosing, evaluating, and reporting data imputation

I have read about model checking for multiple imputation MI (https://ete-online.biomedcentral.com/articles/10.1186/s12982-017-0062-6), but I am not sure how one can check their model for a general ...
0
votes
0answers
56 views

how to get pooled p value after multiple imputation using amelia or mice and cox model?

I try to get pooled p value after multiple imputation using amelia or mice packages with a cox model. I try with the script posted by crsh in october 2013 using amelia and zelig packages but coxph is ...
0
votes
0answers
66 views

Is any variable off limits for multiple imputation

I've read elsewhere that (despite common beliefs on the topic) that one should impute outcome variables (though this may have little benefit under MAR). My question is related, but I think, distinct: ...
0
votes
0answers
245 views

Calculating pooled standard deviation from multiply imputed datasets

I need to calculate group mean and standard deviation from multiply imputed datasets. I assume, carrying forward from Rubin's rules about obtaining pooled parameter estimates, that the best way to ...
0
votes
0answers
20 views

Do I need to test for OLS violations with multiple imputations?

I am trying to understand whether it would be logical to use post estimation to test for violations of OLS, when I use multiply imputed values. I am running regressions on panel data, which has 324 ...
2
votes
1answer
441 views

Calculating pooled p-values manually

For reasons I won't go into I need to calculate parameter estimates from several imputed datasets. Based on this CV post about Rubin's rules I have determined how to manually calculate both the pooled ...
1
vote
0answers
26 views

Multiple Imputation query

I am using multiple imputation to deal with around 49 missing observations for my outcome variable from my 324 observation panel dataset. I used Stata to perform 10 imputations for this, using ...
0
votes
0answers
30 views

MNAR Sensitivity Analysis Specifying Delta

I have a dataset from a cross-sectional study that has a binary variable with approximately 20% of the data missing. I highly suspect that the mechanism of missingness is MNAR. The data that is ...
0
votes
1answer
46 views

How to manually do multiple imputation to fill in missing data points

I am trying to calculate a pearson's correlation between scores on a test to determine test-retest reliability. However, there are only 4 items on the test, with a maximum score of 4 and minimum score ...