| name | vera-data-nominal-reviewing |
| description | Runs class distribution diagnostics and primary association tests for nominal (unordered multi-class) outcome variables. Produces class frequency table, proportions, rare-class warnings, bar chart, and one interpreted association test (Chi-square/Cramer's V for categorical predictors, or ANOVA/Kruskal-Wallis for continuous predictors) with effect sizes. Outputs .R and .py scripts with 2 publication-quality plots. Triggered when user has a nominal outcome with 3+ unordered levels and says "nominal outcome," "multi-class," "unordered categories," "multinomial," or names a variable like species, diagnosis type, region. Does NOT handle binary (2-level), ordinal, continuous, count, survival, or SEM outcomes. |
| allowed-tools | Read, Bash, Write, Edit |
Nominal Outcome — Class Distribution Diagnostics & Primary Association Test
Table of Contents
Open-source skill.
Scope Boundary
Use this skill when:
- The outcome has 3 or more unordered classes and the first task is a baseline association test plus class-distribution check.
- A descriptive / inferential starting point is needed before multinomial or classifier-style modeling.
Do not use this skill when:
- The outcome has only 2 classes or has a meaningful order.
- The main need is repeated-measures, survival, or multilevel nominal modeling.
Workflow
Read each step file in workflow/ before executing that step.
| Step | Responsibility | Executor | Document | Input | Output |
|---|
| Collect | Collect Inputs | Main Agent | workflow/step01-collect-inputs.md | User input | Structured input summary |
| Diagnose | Check Distribution | Main Agent | workflow/step02-check-distribution.md | Prior step output | PART 1 code block |
| Test | Run Primary Test | Main Agent | workflow/step03-run-primary-test.md | Prior step output | PART 2-3 code blocks |
Decision Tree
1. VALIDATE OUTCOME TYPE
├── Only 2 levels → redirect to vera-data-binary-reviewing
├── Ordered levels → redirect to vera-data-ordinal-reviewing
└── 3+ unordered levels → proceed
2. PRIMARY ASSOCIATION TEST
├── Primary predictor is categorical → Chi-square + Cramér's V
└── Primary predictor is continuous → One-way ANOVA (predictor ~ outcome) or Kruskal-Wallis
Required Inputs
| Role | What to collect |
|---|
| Outcome (Y) | Variable name, what categories represent, confirm NO ordering |
| Number of levels | Confirm 3+ unordered categories |
| Primary predictor | Variable name, type (categorical or continuous) |
| Predictors | For recommendation block (not executed) |
| Covariates | For recommendation block (not executed) |
Code Structure
PART 0: Setup & Data Loading
PART 1: Class Distribution Diagnostics → plot_01_class_distribution.png
PART 2: Primary Association Test → plot_02_association_[var].png
PART 3: Recommendation Block → text listing additional analyses available
Reporting Standards
- p-values: "< .001" not "0.000"; exact to 3 decimals otherwise
- Effect sizes: Cramer's V (Chi-square), eta-squared (ANOVA) — always alongside p
- 95% CIs: always for effect sizes when available
- Degrees of freedom: always with Chi-square and F statistics
- Sample size: final analytic N
- Decimal places: 2 for frequencies/proportions, 3 for p and effect sizes
- Non-significance: "not statistically significant at alpha = .05" — never "no association"
- Chi-square: report observed vs expected when informative
- Always state which category is the reference (for later multinomial modeling)
Hypothesis Tests
| Predictor Type | Primary Test | Confirmation |
|---|
| Categorical predictor | Chi-square + Cramer's V | Fisher's exact (if any cell < 5) |
| Continuous predictor | One-way ANOVA (predictor ~ outcome as factor) | Kruskal-Wallis |
Example Dataset
R built-in iris: outcome = Species (setosa/versicolor/virginica, 3 unordered classes).
Predictors: Sepal.Length, Sepal.Width, Petal.Length, Petal.Width.
Python: from sklearn.datasets import load_iris (preferred — fully offline) or sm.datasets.get_rdataset("iris").
Method Status
| Status | Methods |
|---|
| Implemented in this skill | Class-distribution diagnostics, chi-square / ANOVA-style association testing, rare-class warnings |
Implemented downstream in vera-data-nominal-generating | Multinomial logistic regression, LDA, CART/RF/LightGBM, subgroup analysis, and IIA diagnostics |
| Out of scope in this open-source baseline | Binary, ordinal, nested-choice, or multilevel nominal workflows not named above |
Minimal Smoke Test
- Smoke-test prompt: "Run
vera-data-nominal-reviewing on iris, treating Species as the outcome and Petal.Length as the primary predictor. Produce the standard baseline artifacts."
Cross-Skill Interface
Output:
├── code_r → .R script
├── code_python → .py script
├── figures/ → 2 PNGs (class distribution + association plot)
└── recommendations → text block (additional analyses available)
Next step: Invoke vera-data-nominal-generating from this skillset to run the full pipeline (additional tests, subgroup analysis, modeling, manuscript generation). See ../../CROSS-SKILL-INTERFACE.md for the shared handoff contract.
