Assessment methodology, scoring and limitations
What we assess
16Mindsets is an independent 16-type self-reflection tool, not the official MBTI assessment. Questions cover E/I, S/N, T/F and J/P preferences. The assessment does not measure cognitive-function stacks, intelligence, mental health or ability.
70 questions and five answer levels
Each question presents two descriptions, A and B. Strong A, lean A, equal, lean B and strong B contribute 1, 0.75, 0.5, 0.25 and 0 points toward the first preference on that dimension. The other preference receives the remaining share.
Calculating percentages
E/I has 10 questions; S/N, T/F and J/P each have 20. Each dimension is calculated independently: first-preference points divided by its question count, multiplied by 100. The second preference receives the remaining percentage. Raw totals are not compared across dimensions. The higher score determines the letter; an exact tie is shown as X.
Example: balanced responses
If every answer on a dimension is equal, each side receives 50% and that position is X. Displayed percentages may be rounded. Letters use the underlying scores, so a displayed 50% does not necessarily indicate an exact tie in the underlying calculation.
Intended use and evidence limits
This is a five-level adaptation of a self-report assessment. We have not published peer-reviewed reliability, validity, representative norming or diagnostic-accuracy research. Results should not be treated as validated predictions of behaviour. Report suggestions support reflection and should not alone determine hiring, medical or major life decisions.
Payment and result protection
After all questions are answered, the server calculates and stores the result. Personal type, scores and the complete report unlock only after successful payment is confirmed. Retaking creates another result and does not rewrite an existing purchased report. Public profiles and articles do not contain your personal assessment result.
English and Traditional Chinese
The Traditional Chinese version translates the same questions and uses the same order and scoring. We have not published cross-language measurement-equivalence research and do not claim complete psychometric equivalence between the versions.