Data for: Cognitive Resolution: An Integrative Construct of Graded Discrimination in Judgment, with a Scoring Rubric and a Pilot Reliability Study

Four-level scoring rubric, 24 constructed vignettes with intended-level key, all three raters' individual outputs, the consolidated ratings table, and reliability statistics for a codeability pilot of the "cognitive resolution" construct. All vignettes are fictional constructions; no real persons, events, or business data are involved. The three raters are instances of one large-language-model family, so the statistics indicate feasibility of rubric codeability, not reliability or validation. Recomputation from these files confirms the paper's reported statistics (0.917 exact agreement vs. key; 0.911 quadratic-weighted kappa; 1.00 pairwise/Fleiss kappa; r = 0.396 length-level correlation).