quant-multi-group-indlisted

# quant-multi-group-ind — Three or More Independent Groups ## Objective Test whether an outcome variable differs across 3 or more independent groups, identify *which* specific groups differ and *by how much*. The omnibus test tells you whether any difference exists; the mandatory post-hoc analysis tells you where. --- ## Pre-flight **Required inputs:** - `output/data/cleaned_data.csv` - `output/data/column_metadata.json` **Requirements:** - Outcome variable: ordinal or continuous - Grouping variable: 3–15 unique non-NA values (warn above 10) - Minimum n per group: 10 (warn), 20 (preferred) - If grouping variable has exactly 2 values: redirect to `/quant-two-group-ind` --- ## Pipeline ### Step 1 — Load data and metadata ```python # See references/pipeline.py → load_prepared_data() ``` --- ### Step 2 — AskUserQuestion: select variables ``` AskUserQuestion: question: "Which variable would you like to compare across groups?" header: "Outcome variable" options: [list ordinal/continuous columns] ``` ``` AskUserQuestion: question: "Which variable defines the groups?" header: "Grouping variable" options: [list nominal/binary columns — show unique value count for each] ``` If grouping variable has >15 unique values, AskUserQuestion: - "This variable has {n} unique values. Please confirm you want to compare all of them, or select a subset." (with option to continue or re-select) --- ### Step 3 — AskUserQuestion: select method ``` AskUserQuestion: questi