Purpose and scope
This page describes the public methodological boundaries of FermatMind's Big Five content. The assessment is a structured self-report tool for self-understanding, communication, and reversible experiments with behavior. It is not a medical diagnosis, a hiring or admissions screening tool, a basis for legal or financial decisions, or a prediction of an individual's future.
How the model is expressed
Big Five research generally describes personality using broad, continuous dimensions rather than fixed identity types. Different instruments organize narrower facets differently. Keep the specific model name in view when reading a result; the facet categories of two instruments cannot simply be treated as interchangeable.
What we disclose about measurement and scoring
The FermatMind Big Five scoring guide explains how the 90- and 120-item versions score self-reported answers. It describes forward and reverse scoring, averaging items within a facet, and giving each of a domain’s six facets equal weight. Each version has its own items, scoring policy, and reference version. Check the form, score scale, and reference information before comparing numbers.
When a usable reference mean and standard deviation are available, the method derives z- and T-scores. The percentile is an estimate based on that reference and a standard normal distribution assumption. Without a usable reference, the linear 0–100 fallback shows where a score lies on the original response scale and cannot be used for population ranking. This page does not list item-by-item scoring keys or full reference distributions. The formulas explain the calculation. Evidence for specific populations and uses is a separate question.
Information currently unknown
- Product reliability coefficients: Unknown
- Product validity coefficients: Unknown
- Reference populations and sample sizes: use only the reference population and sample size displayed on that result. Treat any omitted field as unknown.
- Empirical calibration evidence for normal-approximation percentiles across reference groups: not provided on this page. A reference status label alone cannot establish sample representativeness or empirical calibration.
- Standard error of measurement: Unknown
- Measurement equivalence across groups: Unknown
- Predictive accuracy: Unknown
Known limitations
Self-report answers can be affected by how a person understands an item, their current context, reference group, response style, and desire to present an ideal self. A result is a time-bound observation. It is not direct evidence of ability, motivation, diagnosis, fit, employability, leadership, or future outcomes. A difference on retesting may reflect measurement variation, context, different responses, or real change; this page cannot determine the cause for an individual.
Privacy boundaries
Private answers, assessment records, report links, recovery information, orders, and payment information must not enter public pages, public internal links, or identifiable analytics logs. Public methodological content does not turn private results into public information.
Public evidence index
- Goldberg (1990): Big Five factor structure supports a broad five-factor structure; it does not validate FermatMind's scoring, norms, individual results, or outcome predictions.
- Soto and John (2017): BFI-2 domains and facets supports one specifically named hierarchical model; it does not mean FermatMind implements the BFI-2 or inherits its measurement results.
- DeYoung, Quilty, and Peterson (2007): Ten aspects supports one aspect hierarchy; it does not make all Big Five facet systems equivalent.
- Roberts, Walton, and Viechtbauer (2006): Meta-analysis of trait change supports group-level patterns of change; it cannot predict whether, why, or how a particular person will change.
