quant-model-continuouslisted

# quant-model-continuous ## 1. Objective Fit an OLS regression model for a **continuous or near-continuous dependent variable** with predictors of any type. HC3 heteroscedasticity-robust standard errors are used by default, making inference valid even when residual variance is unequal across fitted values. **When to use:** - DV is continuous (reaction time, spending, composite scale) - DV is Likert 1–10 treated as interval - Goal: understand which predictors are associated with the DV and by how much **When NOT to use:** - DV is Likert 1–5 with few categories → `quant-model-ordinal` - DV is binary (0/1) → `quant-model-nominal` - DV has 3+ unordered categories → `quant-model-multinomial` - Data is nested/hierarchical → `quant-model-multilevel` --- ## 2. Pre-flight checks | Check | Threshold | Action if violated | |---|---|---| | Minimum n | ≥ 10 obs per predictor (rule of thumb) | Warn, proceed with caution | | DV type | Near-continuous, no hard floor/ceiling | Suggest ordinal model for Likert 1–5 | | Complete cases | ≥ 30 after listwise deletion | Error if < 10 | | Multicollinearity | VIF < 10 | AskUserQuestion (Q5) | | Influential cases | Cook's D > 4/n | AskUserQuestion (Q4) | --- ## 3. AskUserQuestion protocol Ask in this order. Never auto-decide — always pause and ask. ### Q1 — Input file ``` What is the path to your input CSV file? ``` ### Q2 — Outcome variable ``` Which column is the outcome (dependent) variable? This skill expects a continuous or near-conti