Our methodology Mazarion guides
How Mazarion’s assessments work
Our assessments summarize what you report. This guide explains how answers become scores, what the percentages mean, and where our interpretation begins.
Research context
These are custom self-reflection questionnaires. Mazarion has not established them as validated psychological instruments; a numerical score does not provide diagnostic accuracy or population norms.
Five questionnaires, one response scale
Each item takes a response from 1 to 5. Items are assigned to a category or preference pole; their points are summed within that group. Question order may be shuffled, but the item’s assigned group stays the same.
| Assessment | Items | Scored groups | Maximum per group |
|---|---|---|---|
| Attachment patterns | 20 | 4 categories | 25 |
| Love languages | 25 | 5 categories | 25 |
| MBTI-style preferences | 32 | 8 poles | 20 |
| Enneagram | 45 | 9 categories | 25 |
| Archetypes | 48 | 12 categories | 20 |
The MBTI-style quiz compares eight poles in four pairs. The other quizzes rank category totals. These calculations describe completed questionnaires with one answer per item.
A percentage of points, not a probability
For a category or MBTI pole, we divide its total by its maximum possible points, multiply by 100, and round to a whole number.
Worked example · five questions
4 + 3 + 5 + 2 + 4 = 18 points.
18 ÷ 25 × 100 = 72%.
This means 72% of the available points in that question group. It is not a 72% chance that the label is correct, and not a 72nd percentile among other people.
Because the lowest answer is 1 rather than 0, answering 1 to every item in a complete group produces 20%. Separate category percentages need not add to 100. The quizzes do not use comparison samples to turn these values into population percentiles.
Rankings, ties, and generated descriptions
Attachment, love-language, Enneagram, and archetype labels come from the largest group totals. Equal totals retain a fixed category order, so even equal responses can produce a leading label. For MBTI pairs, ties choose E, S, T, and J.
An Enneagram wing is selected from adjacent group scores; growth and stress themes are predefined. An archetype shadow is mapped from the leading archetype. MBTI cognitive stacks and love-language expression descriptions are also generated from the result, rather than measured independently.
When totals are close, focus on the answers rather than the rank. A repeat result that changes by one letter or position does not by itself establish a meaningful psychological change.
The attachment indices use different formulas
Our security index is the secure-group total divided by the sum of all four group totals, multiplied by 100. It measures that group’s share of quiz points; it does not mean “percent psychologically secure.”
The anxiety index combines the anxious and fearful group totals, divides by 50, and multiplies by 100. The avoidance index similarly combines avoidant and fearful totals. For an all-neutral response of 3 to every question, all four category scores are 60%, the security index is 25%, and both axes are 60%.
These custom indices do not use the ECR-R’s items or scoring. Published ECR-R cutoffs or norms should not be applied to them. Compare the original measure’s scoring guidance.
What would stronger accuracy claims require?
Validation would require studies of these exact questionnaires and their intended uses, including score consistency, structure, comparison with relevant measures, and performance in the populations using them. We have not established that evidence for Mazarion’s custom quizzes.
For contrast, the BFI-2 development paper reports research on its particular instrument. It is an example of instrument-specific evidence, not evidence for our app. Soto & John, 2017.
Use our results for optional reflection. Neither a quiz score nor a compatibility score estimates the probability of relationship success, establishes a diagnosis, or measures a person’s worth.
Take it into your day
A question to sit with
Before interpreting a score, ask: what was counted, what is it being compared with, and what conclusion can that actually support?
Sources & evidence
- Soto & John (2017) · Development and assessment of the BFI-2
Primary validation research for a specific Big Five instrument. Mazarion does not administer the BFI-2.
- R. Chris Fraley · ECR-R scoring and interpretation
The measure developer’s guidance on anxiety, avoidance, and the limitations of categorical cutoffs.
Research about a framework does not validate Mazarion’s custom questionnaire. Scoring descriptions on this page were checked against our implementation.