I remember Raghu once telling us that back in the 70's or 80's, when imputation methods were started to being developed, a lot of researchers would refuse using it because they thought they were just fabricating data, without any scientific basis.
And actually, this might not be quite the right thing since it's more of a step-by-step guide on how to do weighting using an R package that we created, but it does go through the steps and their rationale, including imputation.
This post walks through the process of weighting and analyzing a survey dataset.
Ahh, yeah, that's the one area where we do use imputation routinely. You know, this does make me think that an explainer on survey weighting (including the imputation piece) could be worth doing!
Thank you so much! Imputation doesn’t seem to come up as much in the discourse about polling (and it’s not used very much in public opinion research) so it’s not something I’ve written about. But it wouldn’t surprise me if someone has done a short explainer on the topic already.
Imputing Missing Values with External Data https://arxiv.org/abs/2410.02982 arXiv:2410.02982v1 Announce Type: new Abstract: Missing data is a common challenge across scientific disciplines. Current imputation methods require the availability of individual data to impute missing values. Often, ho 📈🤖
We will be covering multiple imputation methods to address 1) where data are missing not at random 2) imputation for multilevel models 3) imputation for survival models, and 4) imputation for propensity score analysis
We are running our short course on Advanced Multiple Imputation Methods to Deal with Missing data this December (5th and 6th). link for further info: tinyurl.com/ybu982ru
AI/ChatGPT are fancy imputation, taking things that have already been created and repurposing them for a *task*. If your job is to write and think to push any sort of boundary, and you want to use fancy imputation for that? Nope!