Ch15: Add item–total tables and alpha variants to reliability analysis

[nicholaskarlson]

Summary

No response

Details

Summary

Extend the Chapter 15 reliability analysis (scripts/ch15_reliability_analysis.py)
with richer psychometrics output:

1. An item–total correlation table for the survey items.
2. One or two alpha variants (e.g., standardized alpha, alpha if item dropped).

Goals

• Use pingouin or pandas to compute item–total correlations for each item.
• Save an item–total table to outputs/ch15/ch15_item_total.csv, including:
  • item name (e.g., q1, q2, …),
  • item–total correlation,
  • (optionally) “alpha if item dropped”.
• Report at least one alpha variant in the JSON summary
  (ch15_reliability_summary.json), e.g.:
  • alpha_standardized
  • alpha_if_item_dropped (as a dict keyed by item)
• Keep existing outputs (Cronbach’s alpha, ICC, Bland–Altman plot) unchanged.

Hints

• pingouin has helpers like cronbach_alpha; its documentation also
  discusses item-level diagnostics.
• An item–total correlation is typically the correlation of each item with
  the sum of the remaining items.
• Follow the existing JSON structure so downstream code and docs remain simple.

Difficulty

Low/medium: good first issue for someone comfortable with pandas and basic stats.

Files to Touch

No response

Contributor Checklist

• I have read CONTRIBUTING.md.
• I can run make lint locally.
• I can run make test locally.
• I have checked for existing issues/PRs that might overlap.