Item Development & Provenance
How our questions were authored, mapped to constructs, and structured into Quick and Deep assessment forms.
1. The 60-Item Master Bank
Our research-beta item bank contains exactly 60 original statements authored specifically for DualLens. No items were copied from copyrighted commercial assessments such as the official MBTI® or commercial NEO inventories.
- • Extraversion: 8 items (E01–E08)
- • Agreeableness: 8 items (A01–A08)
- • Conscientiousness: 8 items (C01–C08)
- • Negative Emotionality: 8 items (N01–N08)
- • Open-Mindedness: 8 items (O01–O08)
- • Energy (E/I): 5 items (EI01–EI05)
- • Information (S/N): 5 items (SN01–SN05)
- • Decisions (T/F): 5 items (TF01–TF05)
- • Structure (J/P): 5 items (JP01–JP05)
2. The 36-Item Quick Map Subset
The Quick Map is not a separate toy quiz. It is a strictly controlled 36-item subset drawn directly from the master bank:
- 20 Big Five items: Exactly 4 items per domain.
- 16 Type items: Exactly 4 items per preference axis (2 forward, 2 reverse).
Because both assessments draw from the identical item architecture, a user who completes Quick Map can immediately transition into Deep Map without having to re-answer the first 36 questions.
3. Selective Shared-Item Architecture & Construct Independence
A naive mistake in multi-model assessments is pretending two frameworks are either 100% identical or completely isolated. In psychometric reality, personality constructs share valid variance without being identical: related, not equivalent.
In DualLens, the Big Five Trait Lens is scored strictly from its primary Big Five items to preserve pure domain definitions. The Type Lens combines dedicated type-preference items with a limited set of explicitly supported provisional secondary mappings from overlapping Big Five items (e.g., Extraversion items informing E/I, Openness informing S/N, Agreeableness informing T/F, Conscientiousness informing J/P). Negative Emotionality has zero cross-mappings and functions as a strictly independent cross-cutting modifier.
This selective architecture preserves discriminant coverage without artificially walling off theoretical overlap, ensuring that synthesis rules reflect genuine multidimensional dynamics rather than circular measurement artifacts.