Full technical article is currently available in English, Simplified Chinese, Traditional Chinese, Japanese, and Korean. English fallback is shown below.

Scoring Specification

How Scoring Works

Exact arithmetic for OpenJung’s 32-item Full and 8-item Quick modes

Last updated:

Evidence and validation status

English fallback — this evidence note is available in English, Simplified Chinese, Traditional Chinese, Japanese and Korean.

Traceable source: the full test adapts the 32-item OEJTS 1.2 by Open Psychometrics. The questionnaire PDF documents that version’s items, scoring rules and license. The development page provides item-selection background, not validation of this adaptation.

Not yet published: internal consistency, test–retest reliability and factor validation for this OpenJung version, its 8-question quick test, or its translations. We do not present unsupported reliability coefficients for this implementation.

Some wording and item directions have been adapted. The quick test uses only 2 questions per dimension; scaling it to 8–40 for display does not give it the precision of the full test. Percentages describe a position on a scale, not accuracy or diagnostic probability.

Use results for reflection and communication. Near-midpoint results do not establish a fixed type. Do not use them for hiring, clinical or other high-stakes decisions.

OEJTS questionnaire items and adaptations are licensed under CC BY-NC-SA 4.0, not public domain. The MIT license covers original OpenJung code, not the questionnaire.

Current assessment version: openjung-oejts-1

Pipeline Overview

32 Items (Full)
8 per axis · native 8–40 sum
8 Items (Quick)
2 per axis · native 2–10 × 4
4 Preference Axes
E/I, S/N, T/F, J/P
16 Letter Profiles
Cut at midpoint 24 (50%)

Full Mode (32 Items) vs. Quick Mode (8 Items)

Both assessment lengths map answers onto the same 8–40 raw dimension scale centered at 24. Choosing between them trades response granularity for speed.

32-Item Full Assessment

Answers across all 8 bipolar items per axis are summed directly into an 8–40 score. Moving one answer by one point changes the raw score by 1 point, or 1 / 32 = 3.125% of the scale. Full results display concrete pole percentages plus a 1-step sensitivity range.

8-Item Quick Assessment

Each axis uses 2 items (EI: 3, 15; SN: 24, 32; TF: 22, 14; JP: 9, 13) yielding a 2–10 native sum, which is multiplied by 4 to project onto the 8–40 scale. Because one answer step shifts the projected score by 4 points (12.5%), Quick mode classifies leans into three bands rather than foregrounding single-percent precision.

Quick Mode Preference Bands

45–55%
Near midpoint
Close to 24; one step on one item can flip the letter.
55–75%
Slight lean
Moderate tilt toward one pole, still sensitive to 1-step swings.
75–100%
Clear lean
Both items pull consistently toward the same preference pole.

Step 1: Bipolar Question Structure

Every item anchors two contrasting everyday habits on a 1–5 scale, where 3 marks neutral.

Sample EI item (rating = 4):

Energized by socializing
1 2 3 4 5
Energized by solitude
Rating Meaning
1 Strongly prefer left trait
2 Somewhat prefer left trait
3 Neutral / no clear preference
4 Somewhat prefer right trait
5 Strongly prefer right trait

Step 2: Summing Answers in the 32-Item Worked Example

In the full test, OpenJung sums 8 items per axis. Below is a realistic worked example producing near-midpoint boundary cases:

Extraversion (E) vs. Introversion (I)

E – I

Items: 3, 7, 11, 15, 19, 23, 27, 31

4 3 5 2 4 3 4 3 = 28

Sensing (S) vs. Intuition (N)

S – N

Items: 4, 8, 12, 16, 20, 24, 28, 32

2 3 4 2 3 4 3 4 = 25

Feeling (F) vs. Thinking (T)

F – T

Items: 2, 6, 10, 14, 18, 22, 26, 30

3 2 4 3 2 3 4 3 = 24

Judging (J) vs. Perceiving (P)

J – P

Items: 1, 5, 9, 13, 17, 21, 25, 29

4 5 3 4 5 4 3 4 = 32

Scale span: minimum 8 (eight 1s) to maximum 40 (eight 5s), midpoint 24 (eight 3s).

Step 3: Cutoff Threshold & Midpoint 24 Tie Convention

Each 8–40 sum is checked against threshold 24: scores at or below 24 select the left pole (E, S, F, J), and scores strictly above 24 select the right pole (I, N, T, P). Note that TF assigns low scores (8–24) to F and high scores (25–40) to T.

Because a raw score of exactly 24 gives 50.0% / 50.0%, the letter assigned at the tie (E, S, F, or J) is purely a tie-breaking convention—it indicates a balanced near-midpoint axis, not a clear tilt.

E – I Score: 28
E · Extraversion Introversion · I
8 24 (50%) 40
28 > 24 → I
S – N Score: 25
S · Sensing Intuition · N
8 24 (50%) 40
Near midpoint (53.1% / 46.9%) 25 > 24 → N
F – T Score: 24
F · Feeling Thinking · T
8 24 (50%) 40
Exact midpoint 50% / 50% tie → convention picks left pole F 24 ≤ 24 → F
J – P Score: 32
J · Judging Perceiving · P
8 24 (50%) 40
32 > 24 → P

Worked Example Result (with near-midpoint flags)

I N F P

Result reads I–N–F–P, where T–F sits right on the 50% / 50% midpoint (raw 24 → convention selects F) and S–N sits just above it at 53.1% N / 46.9% S (near midpoint).

Step 4: Percentage Calculation (E + I = 100%)

Percentages locate your raw score along the 32-point span between 8 and 40. Both poles are exact complements totaling 100%:

Right pole % = ((score - 8) / 32) × 100
Left pole % = 100% - Right pole % (so E% + I% = 100%)
Dimension Raw Score Arithmetic Resulting Poles
E – I 28 ((28 - 8) / 32) × 100 62.5% I / 37.5% E
S – N 25 ((25 - 8) / 32) × 100 53.1% N / 46.9% S
F – T 24 ((24 - 8) / 32) × 100 50.0% F / 50.0% T
J – P 32 ((32 - 8) / 32) × 100 75.0% P / 25.0% J

Inspectable Scoring Source

Every weighting constant and tie-breaking line lives in the open-source scoring module.

View openjung/core on GitHub →

Continue Reading