Library/Personality / IPIP‑NEO‑120 Map (Postgres-First)
All articles

Personality / IPIP‑NEO‑120 Map (Postgres-First)

This folder is organized to resemble a Postgres schema + seed data so schema gaps are visually obvious before we implement the real database.

Layout

  • raw/: original source CSV/TS artifacts (kept for provenance / diffing)
  • postgres/schema.sql: proposed Postgres DDL
  • postgres/seeds/: CSV seed files matching the tables in postgres/schema.sql

Tables (Seeds)

Core model

  • postgres/seeds/ffm_domains.csv
  • postgres/seeds/ffm_aspects.csv
  • postgres/seeds/ffm_facets.csv
  • postgres/seeds/ffm_facet_aspect_loadings.csv

Evidence

  • postgres/seeds/citations.csv

Domain routing (interview question selection)

  • postgres/seeds/psychology_domains.csv
  • postgres/seeds/psychology_domain_facet_affinity.csv
  • postgres/seeds/facet_domain_preference.csv

AB5C

  • postgres/seeds/ab5c_facet_blends.csv
  • postgres/seeds/ab5c_derived_correlations.csv

Rules

  • postgres/seeds/cross_facet_rules.csv
  • postgres/seeds/cross_facet_rule_conditions.csv
  • postgres/seeds/aspect_rules.csv
  • postgres/seeds/aspect_rule_conditions.csv
  • postgres/seeds/red_flags.csv
  • postgres/seeds/red_flag_conditions.csv
  • postgres/seeds/red_flag_contraindications.csv
  • raw/ffm_aspects_CALIBRATION.md — methodology for calibrating aspect rule boost_multiplier values

Cross-Model Construct Mapping

  • postgres/seeds/cross_model_constructs.csv — model-agnostic psychological constructs
  • postgres/seeds/scale_construct_mapping.csv — maps model-specific scales to shared constructs

Instrument (IPIP‑NEO‑120) — gap placeholders

  • postgres/seeds/ipip_instruments.csv (present)
  • postgres/seeds/ipip_items.csv (seeded for ipip_neo_120)
  • postgres/seeds/ipip_item_facet_key.csv (seeded for ipip_neo_120)

Cross-Model Construct Mapping

The scale_construct_mapping table links model-specific scales (FFM facets, HEXACO facets, BFAS aspects) to model-agnostic constructs. The mapping_type column disambiguates three distinct relationship types:

| mapping_type | Meaning | Loading interpretation | |--------------|---------|------------------------| | primary | The scale operationalizes this construct | Always 1.0 (definitional identity) | | cross_model | Empirical correlation with a construct from another model | Published correlation coefficient | | aspect_loading | Factor loading from CFA/EFA (BFAS aspect structure) | Loading from DeYoung et al. (2007) |

Sources:

  • FFM facets: Definitional (loading = 1.0)
  • HEXACO-FFM correlations: Ashton & Lee (2014), Lee & Ashton (2004)
  • BFAS factor loadings: DeYoung, Quilty, & Peterson (2007) Table 2

Known Gaps (Intentional)

  • Only the ipip_neo_120 instrument is seeded; other instruments and versions still need item banks + keys.
  • Norms / percentile transforms are not modeled yet (e.g., scoring tables per instrument version, population, role, etc.).
  • Interview-question content itself is not seeded here; raw/registry_questions_analysis.csv is analysis output, not a source-of-truth table.

Loading (Optional)

  • Create tables: run postgres/schema.sql in a local Postgres instance.
  • Seed data: COPY/\copy from postgres/seeds/*.csv into the matching personality.* tables.
← Back to library