feat: multi-label moral foundation ratings with z-scored frame averaging and human calibration

- Add scripts/07_multilabel.py: LLM judge rates all 7 foundations per vignette
  using violation (forward) and acceptability (reverse) frames
- Foundation definitions drawn from Clifford et al. (2015) survey rubric
- Z-score each frame per foundation before averaging to cancel range bias
- Calibrate LLM Likert → human % via per-foundation OLS (classic set only)
- Add scripts/07a_merge_labels.py: merges llm_* and calibrated_* into vignette files
- Update README and HF dataset card with methodology and calibration quality table
- Classic set: 80.3% dominant-foundation accuracy, Pearson r 0.69-0.89 per foundation

data/calibration.json

@@ -0,0 +1,87 @@
+{
+  "model": "x-ai/grok-4-fast",
+  "foundations": {
+    "Care": {
+      "n": 132,
+      "spearman_r": 0.7345042029622307,
+      "spearman_p": 1.2176410154221345e-23,
+      "pearson_r": 0.8102300980028216,
+      "pearson_p": 5.905981200185428e-32,
+      "slope": 17.91275224831169,
+      "intercept": -18.86597016699005,
+      "mae": 11.767375193181332
+    },
+    "Fairness": {
+      "n": 132,
+      "spearman_r": 0.6225605362565393,
+      "spearman_p": 1.5959953666830447e-15,
+      "pearson_r": 0.8129609595820121,
+      "pearson_p": 2.530684016591177e-32,
+      "slope": 16.7264766737861,
+      "intercept": -19.131907744766767,
+      "mae": 11.111987094673093
+    },
+    "Loyalty": {
+      "n": 132,
+      "spearman_r": 0.6926099209091109,
+      "spearman_p": 3.599007373381228e-20,
+      "pearson_r": 0.7484483078065683,
+      "pearson_p": 6.0343078480636855e-25,
+      "slope": 14.538458435939967,
+      "intercept": -15.378022367181732,
+      "mae": 9.309441437396758
+    },
+    "Authority": {
+      "n": 132,
+      "spearman_r": 0.3856396780981898,
+      "spearman_p": 4.961061693936076e-06,
+      "pearson_r": 0.6875303659859135,
+      "pearson_p": 8.65179834421809e-20,
+      "slope": 13.428260762938889,
+      "intercept": -13.722907418808013,
+      "mae": 11.683327561037107
+    },
+    "Sanctity": {
+      "n": 132,
+      "spearman_r": 0.6196021834312778,
+      "spearman_p": 2.365393748673303e-15,
+      "pearson_r": 0.8887367634786868,
+      "pearson_p": 7.21924815710244e-46,
+      "slope": 18.316412771225654,
+      "intercept": -17.576719463687116,
+      "mae": 6.263818224972179
+    },
+    "Liberty": {
+      "n": 132,
+      "spearman_r": 0.6020785985113376,
+      "spearman_p": 2.2394672471928682e-14,
+      "pearson_r": 0.8139335364560988,
+      "pearson_p": 1.865088580341189e-32,
+      "slope": 14.712894877552804,
+      "intercept": -14.163679534840798,
+      "mae": 8.163126583084338
+    },
+    "SocialNorms": {
+      "n": 132,
+      "spearman_r": -0.4272322719509285,
+      "spearman_p": 3.2400850542942107e-07,
+      "pearson_r": -0.3207249716209886,
+      "pearson_p": 0.00017729490832067168,
+      "slope": -9.865750625757574,
+      "intercept": 53.56969710715305,
+      "mae": 18.84631540933629
+    },
+    "wrongness": {
+      "n": 132,
+      "spearman_r": 0.7618659176897857,
+      "spearman_p": 2.7640404209932235e-26,
+      "pearson_r": 0.8292329749373769,
+      "pearson_p": 1.1959738514994125e-34,
+      "slope": 0.6992902585652869,
+      "intercept": 0.1468375358206399,
+      "mae": 0.4377854374641585
+    }
+  },
+  "dominant_accuracy": 0.803030303030303,
+  "n_vignettes": 132
+}