Methodology: test sources and scoring
The scale or theory behind each mind.io test, where its items come from, how scores are computed, and a dated log of changes that affected results.
How A Test Page Usually Gets Structured
A mind.io test page usually starts with the same basic question: what is this trying to measure? From there we explain the response format, expected time, and how the outcome is grouped or summarized for the reader.
Scoring details such as question counts, reverse scoring, score ranges and tie handling come straight from the real code and data, so a result can be inspected, not just consumed. The test-by-test summary is below, under “Sources and scoring for each test”.
How We Turn Scores Into Reader-Facing Language
Scores are numbers, but readers are really looking for meaning. Our job is to translate that score into language that helps someone reflect on patterns, tradeoffs, or tendencies they might notice in real life.
At the same time, we try not to smuggle extra certainty into that translation. If the data only supports a measured description, we would rather sound modest than push the result into a bigger claim than it can carry.
Where We Draw Interpretation Limits
mind.io is focused on making responses easier to reflect on, not on turning them into a permanent label, a prediction, or a justification for overconfident decisions.
That is why we avoid diagnostic, prophetic, legal, or financial framing in our guides. We keep repeating the same point on purpose: a result only becomes useful when it is read alongside real context.
How We Revisit And Update Content
When question sets, scoring logic, or recommendation copy change, the explanations change with them. Changes that affected results are listed with their dates below, under “Changes that affected results”.
We also revisit pages when we spot mistranslations, awkward wording, overclaiming, or broken links. In practice, that means rereading the page and replacing brittle phrasing with something clearer and more careful.
Sources and scoring for each test
For each test: the scale or theory behind it, where its items come from, how you answer and how the score is computed, matched to the actual code and data. Select a test to see the details.
Personality & temperament
MBTI
Theory-based · own items32 itemsThe four MBTI preference pairs (E/I, S/N, T/F, J/P)- Based on
- The four MBTI preference pairs (E/I, S/N, T/F, J/P). Katharine Cook Briggs and Isabel Briggs Myers built the MBTI on Carl Jung’s theory of psychological types.
- Items
- MBTI-style statements that mind.io wrote separately around the same four pairs, not items from the official MBTI® assessment: eight per pair, half of them keyed toward each side.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- With 3 as neutral, agreeing adds the answer’s distance from 3 (1 or 2 points) to the side the statement points to, and disagreeing adds it to the opposite side. The difference between the two sides becomes a percentage for each pair, and a pair at exactly 50% goes to I, N, F or P, as in the MBTI convention.
- Permission
- MBTI® is a registered trademark of The Myers-Briggs Company. The mind.io test is an unofficial self-reflection aid, not affiliated with the official MBTI® assessment.
Big Five
Theory-based · own items30 itemsThe Big Five model of personality- Based on
- The Big Five model of personality: openness, conscientiousness, extraversion, agreeableness and emotional sensitivity (neuroticism in research terms).
- Items
- Written by mind.io: six statements per trait, three of them reverse-worded.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- Reverse-worded items count as 6 minus the answer; each trait’s mean is then placed on 0–100 (a mean of 1 is 0, a mean of 5 is 100). Up to 33 reads as low, 34–66 as middle and 67 or more as high. It is not a percentile against other people.
Enneagram
Theory-based · own items45 itemsThe Enneagram’s nine types and their core motivations- Based on
- The Enneagram’s nine types, read through their core motivations. It grew out of a self-reflection tradition, so it is not a measurement tool validated by research the way the Big Five is.
- Items
- Written by mind.io: five statements per type, three keyed forward and two reversed.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- After reversed items are flipped, each type’s sum is placed on 0–100 between its lowest and highest possible totals. When the top two types are within 5 points, both are shown as candidates.
DISC Behaviors
Theory-based · own items24 itemsThe four DISC behavior styles- Based on
- The DISC framework, which reads behavior at work and in communication as four styles: Dominance (D, labeled Decision on this site), Influence (I), Steadiness (S) and Conscientious (C). It has little academic validation.
- Items
- Workplace scenarios written separately on the DISC framework, not a paid official assessment. Each option maps to one style, and the options are shuffled for every question.
- Response format
- Pick the one of four behaviors closest to yours.
- Scoring
- Each style is the share of all your answers that picked it. An even spread is about 25% each, so 40% or more reads as a clear tendency and under 20% as a rare style. It is not validated for hiring or evaluation.
Sensitivity (HSP) test
Theory-based · own items18 itemsThree factors of sensory-processing sensitivity (Smolewska et al., 2006)- Based on
- The three factors of sensory-processing sensitivity (Smolewska, McCabe & Woody, 2006): ease of excitation, aesthetic sensitivity and low sensory threshold.
- Items
- Everyday statements mind.io wrote with the research in mind. None is taken from the HSP Scale (Aron & Aron, 1997); there are six per factor and none is reverse-worded.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- The mean of each factor and of all items is placed on 0–100. The three bands (0–39, 40–64, 65–100) are reading aids, not cut-offs set in research or clinical practice.
Relationships & love
Attachment Style
Theory-based · own items28 itemsTwo axes of attachment: anxiety and avoidance- Based on
- A two-axis view of attachment in close relationships: anxiety (worry about being left) and avoidance (discomfort with leaning on others and getting close).
- Items
- Written by mind.io: 14 statements per axis (9 forward, 5 reversed), with the two axes alternating.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- Each axis’s mean, with reverse-worded items counted the other way, is converted to 0–100, and whether each axis is above 50 gives secure, anxious, avoidant or fearful-avoidant. Scores within 5 points of 50 are flagged as borderline. 50 is the scale’s midpoint, not the average person.
Love Languages
Theory-based · own items30 itemsGary Chapman’s five love languages (1992)- Based on
- The five categories counselor Gary Chapman introduced in his 1992 book: Words of Affirmation, Quality Time, Receiving Gifts, Acts of Service and Physical Touch.
- Items
- mind.io’s own items, with no commercial test copied: 30 pairs, in which any two languages meet exactly three times and each language appears twelve times.
- Response format
- Pick the one of two statements that speaks to you more.
- Scoring
- Each language scores the share of the times it was offered that you picked it, on 0–100. Because every pair forces a choice, the ranking can look sharper than it really is.
Love style test
Theory-based · own items18 itemsJohn Alan Lee’s six main love styles (1973)- Based on
- The six main styles of loving sociologist John Alan Lee described in his 1973 book Colours of Love: Eros (passion), Ludus (play), Storge (friendship), Pragma (practical), Mania (all-consuming) and Agape (selfless).
- Items
- Statements mind.io wrote from the theory: three per style, all worded toward their style, so none is reverse-scored. They are not taken or adapted from the Love Attitudes Scale used in research, and have no standardized norms or reliability studies.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- Each style’s score is the mean of its three answers (1–5) converted to 0–100 and rounded (steps of about 8.3). The highest-scoring style is the main style; when two or more share the top score, they are shown together. All six scores are shown, and no style is ranked as better.
Empathy & emotion style test
Published scale · edited37 itemsFour public IPIP scales (Barchard, 2001)- Based on
- Four public IPIP (International Personality Item Pool) scales measuring components potentially related to emotional intelligence (Barchard, 2001): Empathic Concern, Responsive Joy, Emotion-Based Decision-Making and Attending to Emotions.
- Items
- The published items: English exactly as published, other languages translated by mind.io. Empathic Concern’s two items about criminals, which ask for a view on criminal justice more than for everyday concern, are left out, keeping four forward and four reversed items.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree); the IPIP’s own scale runs from very inaccurate to very accurate.
- Scoring
- Reversed items count as 6 minus the answer, and each scale’s mean is placed on 0–100. With no published norms, scores are read against the midpoint of 50, not as percentiles. Empathy warmth averages the Empathic Concern and Responsive Joy means.
- Permission
- IPIP items are in the public domain, and the IPIP allows editing and translating them.
Career & strengths
RIASEC Interests
Theory-based · own items42 itemsJohn Holland’s theory of vocational interests (RIASEC)- Based on
- John Holland’s theory of vocational interests: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E) and Conventional (C).
- Items
- mind.io’s own items, for reference rather than diagnosis: seven per type (five forward, two reversed).
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- After reversed items are flipped, each type’s mean is placed on 0–100, and the three highest types form the three-letter code. A tie at third place comes with a note that the code could change.
VIA Strengths
Theory-based · own items48 itemsThe 24 character strengths of positive psychology (VIA)- Based on
- The 24 character strengths classified in positive psychology (VIA).
- Items
- Short items written by mind.io: two per strength (one forward, one reversed). None is taken from the VIA Institute’s official survey (VIA-IS).
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- After reverse-worded items are flipped, each strength’s mean is converted to 0–100. With two items per strength, scores move in steps of about 12.5 (0, 13, 25, 38, 50 …). Equal scores share a rank, and several strengths tied at the top are shown together as signature strengths.
Wellbeing check-ins
Self-esteem test
Published scale10 itemsThe Rosenberg Self-Esteem Scale (Rosenberg, 1965)- Based on
- The Rosenberg Self-Esteem Scale (Rosenberg, 1965).
- Items
- The scale’s ten items in a widely used order, in mind.io’s own wording.
- Response format
- Strongly agree · agree · disagree · strongly disagree (four options).
- Scoring
- Answers score 3 to 0, with the five negatively worded items (2, 5, 6, 8 and 9) reversed, and the total runs 0–30. The three bands (0–14, 15–25, 26–30) are mind.io’s reading aids, not clinical cut-offs.
- Permission
- The scale is in the public domain.
Self-compassion test
Published scale12 itemsSelf-Compassion Scale – Short Form (SCS-SF; Raes et al., 2011)- Based on
- The Self-Compassion Scale – Short Form (SCS-SF; Raes, Pommier, Neff & Van Gucht, 2011, Clinical Psychology & Psychotherapy 18:250-255), derived from Neff’s (2003) Self-Compassion Scale.
- Items
- The twelve published items in their published order. English follows the published wording; Korean, Japanese and French are mind.io’s translations.
- Response format
- Five-point scale from 1 (almost never) to 5 (almost always).
- Scoring
- The six self-critical items count as 6 minus the answer, and the score is the mean of all twelve (1.00–5.00). Following Neff’s ranges, up to 2.49 is low, 2.50–3.50 moderate and 3.51 or more high. As the short form’s authors recommend, the three components get no separate scores.
- Permission
- Used with citation under the permission Kristin Neff has posted; this implies no endorsement by the authors.
Burnout self-check
Published scale13 itemsCopenhagen Burnout Inventory (CBI; Kristensen et al., 2005)- Based on
- The Copenhagen Burnout Inventory (CBI; Kristensen, Borritz, Villadsen & Christensen, 2005, Work & Stress 19(3):192-207).
- Items
- Only the personal burnout (6 items) and work-related burnout (7 items) scales; the client-related scale is left out. English follows the published wording; other languages are mind.io’s translations.
- Response format
- Five options on how often (always to never or almost never) or how much (to a very high degree to a very low degree).
- Scoring
- Answers score 100, 75, 50, 25 or 0, the one positively worded item reversed, and each scale is the mean of its items. Under 50 is low, 50–74 moderate and 75 or more high: a scale used in research, not a diagnostic cut-off.
- Permission
- Released by its authors for free use.
Depression self-check (PHQ-9)
Published scale9 itemsPHQ-9 (Kroenke, Spitzer & Williams, 2001)- Based on
- The Patient Health Questionnaire-9 (PHQ-9; Kroenke, Spitzer & Williams, 2001).
- Items
- English is the original wording; Korean, Japanese and French were checked item by item against Pfizer’s official translations on phqscreeners.com. The form’s tenth, unscored question (how difficult things have been) is not asked.
- Response format
- How often over the last two weeks, four options scored 0–3.
- Scoring
- The nine answers add up to 0–27, read with the published bands (0–4, 5–9, 10–14, 15–19, 20–27). From 10 we suggest talking with a professional, and any answer above zero on item 9 shows where to find help first, whatever the total.
- Permission
- Developed with an educational grant from Pfizer Inc.; no permission is required to reproduce, translate, display or distribute it.
Anxiety self-check (GAD-7)
Published scale7 itemsGAD-7 (Spitzer, Kroenke, Williams & Löwe, 2006)- Based on
- The Generalized Anxiety Disorder-7 (GAD-7; Spitzer, Kroenke, Williams & Löwe, 2006).
- Items
- English, Korean and French were checked item by item against the official forms on phqscreeners.com (the Korean one is the version used in the Korean validation study, Ahn et al. 2019). That site has no Japanese version, so the Japanese items are not presented as an official translation.
- Response format
- How often over the last two weeks, four options scored 0–3.
- Scoring
- The seven answers add up to 0–21, read with the published bands (0–4, 5–9, 10–14, 15–21). From 10 we suggest talking with a professional.
- Permission
- Developed alongside the PHQ-9 with an educational grant from Pfizer Inc.; no permission is required to reproduce, translate, display or distribute it.
Just for fun
Teto vs Egen
For fun · own items20 itemsThe internet slang “Teto” and “Egen” (not a psychological test)- Based on
- The internet slang “Teto” and “Egen”, taken from testosterone and estrogen. It is not a psychological theory or test, and it has nothing to do with real hormone levels.
- Items
- For-fun items built around the meme: five each on talk, love, work and daily life, ten leaning Teto and ten leaning Egen.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- With 3 as neutral, agreeing adds the distance from 3 to the side the statement leans to and disagreeing adds it to the other side, giving a Teto percentage: 80 or more is strong Teto, 60–79 Teto, 41–59 balanced, 21–40 Egen and 20 or less strong Egen. The optional male/female label only changes the result’s name, never the score.
Personality animal test
For fun · own items15 itemsThree everyday tendencies, read lightly for fun- Based on
- A for-fun test that lightly reads three everyday tendencies: energy (outgoing or reserved), relating (harmonious or candid) and daily life (planned or spontaneous).
- Items
- Statements mind.io wrote for fun: five per tendency, with no standardized norms or reliability studies.
- Response format
- Five-point scale from 1 (strongly disagree) to 5 (strongly agree).
- Scoring
- After reverse-worded statements are flipped, each tendency’s sum is converted to 0–100 (in steps of 5). Above 50 takes the first side and below 50 the second; at exactly 50, the first of that tendency’s answers, in question order, that is not “Neutral” decides (all “Neutral” gives the first side). The three sides together pick one of eight animals.
Saju
Reading, no questionnaireFour pillars and five elements, with astronomically computed solar terms- Based on
- Traditional saju, read from the four pillars of birth year, month, day and hour and their five elements. A cultural reading for self-understanding, not a verdict on your fortune.
- Items
- No questions: you enter a Gregorian or Korean lunar birth date and, if you know it, the birth time (24-hour, KST).
- Scoring
- The solar terms that start each month and year are computed astronomically from the Sun’s position in Korean standard time, and their dates all match the Korea Astronomy and Space Science Institute’s published dates for 1950–2030. The whole day of a term counts as the new month. A lunar birth date is converted with a table based on the Institute’s data (1900–2050). The five elements are counted over the characters of the four pillars (three without a birth time). Overseas times are not converted and no true-solar-time correction is applied.
- Permission
- The lunar table is published under the MIT License (kor-lunar); its notice ships at /third-party-notices.txt.
Daily Fortune
Reading, no questionnaireA daily reflection card set by the date- Based on
- Not a prophecy recalculated from star signs or saju each day, but a reflection card that picks a question to look back on and a small action for the day.
- Items
- No questions: the card picks from text mind.io wrote in advance. Entering a birth date is optional.
- Scoring
- The combination is set by this device’s date together with the birth date in the settings and the saved attachment, Big Five or saju results you leave switched on, so the same day and settings always show the same card. Those saved results also change how strongly each line is weighted, and you can turn them off in the settings.
Face Analysis
Reading, no questionnaireA for-fun reading from the proportions of a face photo- Based on
- A for-fun reading that borrows the language of traditional face reading. A face cannot tell anyone’s personality or fortune.
- Items
- No questions: upload one front-facing photo or take one with the camera.
- Scoring
- A face-finding model is downloaded to your browser and runs on this device, and only values it actually measured, such as eye spacing and left-right balance, go into the reading. Anything it could not detect is left out rather than made up. The photo is never sent anywhere.
Palmistry
Reading, no questionnaireA for-fun reading from the creases of the palm- Based on
- A for-fun reading that borrows the language of traditional palmistry. Palm lines cannot tell lifespan or destiny.
- Items
- No questions: upload one photo of your palm or take one with the camera.
- Scoring
- A hand-finding model is downloaded to your browser and runs on this device; it follows the main creases of the palm to see how long, clear and unbroken the life, heart, head and fate lines are. The photo is never sent anywhere.
Changes that affected results
Changes since October 2026 to items, keying or calculation rules that could change a result, recorded from the actual code history. Use it to decide whether to retake a test you took earlier.
Big Five: emotional sensitivity keyed the right way round, six items replaced
From the test’s launch in January 2026, the worry and calm items of emotional sensitivity (neuroticism) were scored the opposite way, so the score actually measured emotional stability. The keys are now fixed. The same month, six items that mixed in other traits or asked about something other than a personality trait were replaced, and each trait now has three forward and three reverse-worded items.
Good to know: Results saved earlier and old share links are corrected automatically when opened: emotional sensitivity is shown as 100 minus the old score. Older results may therefore show a different emotional sensitivity score, explanation and comparison than before.
Saju: day-pillar reference fixed, solar terms computed astronomically, lunar birth dates
From the launch in February 2026, day pillars were counted from the wrong reference day (1 January 1900 taken as 甲辰 instead of 甲戌), so the day pillar came out wrong. The fix also corrected the hour stem for births between 11 p.m. and midnight. Solar terms are now computed astronomically from the Sun’s position in Korean standard time, which moved 97 term boundary days between 1900 and 2100, mostly one day later. Korean lunar birth dates can now be entered too.
Good to know: Saved and shared charts are recomputed from the stored birth date and time under the current rules. If you were born near a solar term, your month pillar may have changed. If you once entered a lunar birthday as a Gregorian date, choose the lunar option and check again.
MBTI: 32 items, five replaced, and a rule for exact 50:50
The test grew to 32 items, eight per pair with half keyed each way, and each pair is now scored from the items actually answered. Five items that measured social anxiety, indecision or stress tolerance, or nearly duplicated another item, were later replaced, and a pair at exactly 50% now goes to I, N, F or P, as in the MBTI convention.
Good to know: Saved results and share links that carry scores keep the type they had. If an exact 50:50 pair once gave you E, S, T or J, retaking the test may now give the other letter.
DISC: shuffled options, scores as a share of answers
To stop the order of options from steering answers, each question now shows its options in a fixed shuffled order, and the label that revealed each option’s style is hidden. Scores changed from being scaled so the top style reads 100 to the share of all answers that picked each style. Four options that read as another style were later rewritten.
Good to know: Results saved before the change are now shown as shares too, recomputed from the saved answer counts (or from the old scores when a record has none). Their options were ordered and labeled differently, though, so retake the test to compare with today’s result.
Show 9 more changes
Love languages: balanced pairs, now 30
Pairs were rearranged so each language appears equally often, and scores changed from being scaled so the top language reads 100 to the share of times each language was picked when offered. Eight sentences that blurred two languages were reworded, and five pairs were added, making 30 in which any two languages meet exactly three times.
Good to know: Links answered with the earlier 25 pairs are scored the same way, so their scores stay the same. Answers in progress were reset because the questions changed.
RIASEC: eleven reverse and twelve forward items replaced
Eleven reverse items that did not state low interest in their own type, most of them describing another type’s interests, were replaced with statements of low interest in their own type’s activities. Two Enterprising items, for example, had Conventional content and lowered the Enterprising score of people leaning Conventional. Twelve forward items (two per type) that asked about beliefs, abilities or jargon were then rewritten as concrete activities a student can picture. The scoring is unchanged.
Good to know: Saved results keep their scores. Comparing a result from before the forward items changed with a newer one adds a note that the items partly changed. Answers in progress were reset.
Enneagram: seven items replaced
Five items were replaced, among them one that is typical of type 4 but was scored in reverse, one nearly everyone agrees with, and type 9 items that forced types 8 and 9 apart. Type 8’s two reverse items, which actually described type 9, were replaced next.
Good to know: Saved results and score links stay as they were. If your result used to hover between types 8 and 9, it is worth retaking. Answers in progress were reset.
Attachment style: nine items replaced
Nine items were replaced, including one keyed the wrong way and one that counted a healthy time-out as avoidance, and anxiety and avoidance items now alternate.
Good to know: Saved results and score links keep their scores. Answers in progress were reset.
Teto vs Egen: rebuilt with 20 items, four replaced
The twelve work-focused items became twenty across talk, love, work and daily life, with the lean shown in five levels. Four items were later replaced, and an exact 50:50 now says so.
Good to know: Results and links from the twelve-item version still open, but the breakdown by area appears only when every current item was answered.
VIA strengths: four items replaced, shared top strengths
Four items for honesty, humility, leadership and zest that measured something else were replaced. When several strengths share the top score, they are now shown together as signature strengths instead of the list order picking one.
Good to know: Saved results and score links keep their scores. Answers in progress were reset.
HSP: nine items replaced
Nine items that asked about several situations at once, asked you to compare yourself with others, or that almost anyone would agree with were rewritten as everyday statements that ask one thing each. Scoring and bands are unchanged.
Good to know: Saved results keep their scores. Comparing an earlier result with a newer one adds a note that some items changed, mostly in the sensory facet.
PHQ-9 and GAD-7: wording aligned with the official translations
Checked item by item against Pfizer’s official translations, the Korean GAD-7 instruction and answer options, the French GAD-7 items and some Japanese and French PHQ-9 items now follow the official wording. The Korean PHQ-9 already matched it.
Good to know: Scoring and bands are unchanged, so earlier results read on the same scale.
New tests
The self-esteem test (Rosenberg scale), the sensitivity (HSP) test, the burnout self-check (CBI), the depression (PHQ-9) and anxiety (GAD-7) self-checks, the empathy and emotion style test (IPIP), the self-compassion test (SCS-SF), the personality animal test and the love style test opened one after another.
Good to know: Each test’s sources and scoring are described above.
Self-esteem testSensitivity (HSP) testBurnout self-checkDepression self-check (PHQ-9)Anxiety self-check (GAD-7)Empathy & emotion style testSelf-compassion testPersonality animal testLove style test