alterlab-missing-data

name	alterlab-missing-data
description	Handles missing data with principled methods — forces an explicit MCAR / MAR / MNAR mechanism statement, then applies multiple imputation by chained equations (MICE) with Rubin's-rules pooling of estimates and standard errors, or full-information maximum likelihood (FIML) where a likelihood/SEM model applies. Uses statsmodels MICE / MICEData in Python or the field-standard R mice via Rscript, and warns that single (mean/regression) imputation and scikit-learn's IterativeImputer return one completed dataset without Rubin's-rules pooling, so they understate standard errors if used as multiple imputation. Use when a dataset has missing values, when choosing an imputation strategy, or when reporting how missingness was handled. For general modeling on complete data prefer alterlab-statistical-analysis; for latent-variable models with FIML prefer alterlab-sem-psychometrics. Part of the AlterLab Academic Skills suite.
license	MIT
allowed-tools	Read Bash(python:*)
compatibility	Requires (declare in-session, no runtime install on Anthropic API): Python statsmodels>=0.14 (statsmodels.imputation.mice), pandas; scikit-learn IterativeImputer only for single imputation (NOT Rubin pooling) — OR the field-standard R mice>=3.19 via Rscript (mice/with/pool). Runs locally via `uv run python` / `Rscript`; no API key.
metadata	{"skill-author":"AlterLab","version":"1.0.0","depends_on":"alterlab-statistical-analysis (complete-data modeling), alterlab-sem-psychometrics (FIML for latent models); audited by alterlab-ssci-inference-gate"}

Missing Data — Name the Mechanism, Impute Multiply, Pool by Rubin's Rules

Skill type: ANALYSIS MODULE. Missing data is not a nuisance to delete or fill with a mean. The discipline: state the missingness mechanism, then use a method whose uncertainty is honest — multiple imputation with Rubin's-rules pooling, or FIML. The dangerous shortcut is single imputation, which treats guessed values as observed and understates standard errors.

Core Mission

STATE THE MECHANISM (MCAR / MAR / MNAR). MULTIPLY IMPUTE AND POOL BY RUBIN'S RULES —
SINGLE IMPUTATION FAKES CERTAINTY IT DOESN'T HAVE.

When to Use This Skill

• "My dataset has missing values — how should I handle them?"
• "Should I use multiple imputation? How many imputations?"
• "How do I pool results across imputed datasets?"
• "Is mean imputation / listwise deletion okay here?"

Does NOT Trigger

The request is really about…	Route to	Why not this skill
Modeling on already-complete data	alterlab-statistical-analysis / alterlab-statsmodels	No missingness to handle.
FIML inside an SEM / latent-variable model	alterlab-sem-psychometrics	That skill fits the latent model with FIML.
Survey weights / design (a different kind of "adjustment")	alterlab-survey-analysis	Design-based inference, not imputation.
Whether the study design is sound	alterlab-ssci-design-gate	Design routing.

Step 1 — state the mechanism (Rubin)

Mechanism	Meaning	Consequence
MCAR	missingness independent of everything	listwise deletion is unbiased (but wasteful)
MAR	missingness depends on observed data	MI / FIML are valid (the workhorse assumption)
MNAR	missingness depends on the unobserved value itself	needs a selection / pattern-mixture model + sensitivity analysis