AURA Basic
48 progressive-matrix items focused on non-verbal pattern reasoning. Responses are scored server-side against the configured answer key and converted using the assessment's configured norm table.
General factor loadingg=0.81How AURA works
Methodology should be visible, not implied.
AURA is being developed with psychometric input. This page describes the assessment structure, task mechanics, score presentation and the evidence claims currently attached to each product.
AURA Full
12 distinct subtests across six cognitive domains, including timed processing-speed tasks and interactive visual-spatial and working-memory measures.
General factor loadingg=0.92What AURA measures
AURA Basic is deliberately narrow: it focuses on progressive-matrix reasoning. AURA Full separates performance into six domains and then measures each domain using task formats chosen for that construct rather than forcing the whole battery into one generic multiple-choice format.
VCIVerbal Comprehension
FRIFluid Reasoning
VSIVisual Spatial
QRIQuantitative Reasoning
WMIWorking Memory
PSIProcessing Speed
Task structure
VCI · Verbal Comprehension
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
FRI · Fluid Reasoning
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
VSI · Visual Spatial
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
QRI · Quantitative Reasoning
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
WMI · Working Memory
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
PSI · Processing Speed
Multiple task formats are used within this domain. Specific subtest names and administration details are shown to paid examinees.
The Full battery combines multiple-choice, sequential-memory, timed processing-speed and interactive visual-spatial formats. Task mechanics are selected to match the cognitive construct rather than forcing every domain into the same response format.
Scoring
AURA converts subtest performance to a common score scale so results from different task formats can be combined into domain and overall profiles. The conversion layer is intentionally isolated so empirical norm tables can replace the current QA calibration without redesigning the report.
Domain scores combine the subtests within each cognitive domain, and the overall Full score combines the six domain results. Percentiles are calculated from the resulting standard-score profile.
This is a temporary QA/calibration model, not an empirical population norm. It is deliberately isolated in the scoring layer so observed medians, standard deviations, age bands and representative norm tables can replace it later without redesigning the report.
General-factor loading
The project's current psychometric work reports a general-factor loading of g = 0.81 for AURA Basic and g = 0.92 for AURA Full. In practical terms, a higher general-factor loading indicates that performance on the assessment is more strongly associated with the broad general cognitive factor identified in the underlying psychometric analysis.
Before public scientific claims are expanded beyond these figures, the supporting sample, analysis method and study documentation should be published here so users can inspect the evidence directly.
Norms and percentiles
AURA presents scores on the familiar IQ reference scale centred at 100 with a standard deviation of 15. AURA Basic uses its configured raw-score conversion table. AURA Full currently uses the temporary 50%-at-100 development anchor described above; empirical subtest medians and standard deviations should replace that assumption once representative calibration data are available.
Scoring integrity
Payment verification for AURA Full and final score calculation are server-side. Development-only subtest routes are separately gated and automatically disabled on Vercel Production. This allows QA and pentesting without weakening the paid production route.