Big Five personality test results: how a 60 is calculated, and why the same 60 lands in different places

How is a Big Five (OCEAN) 0–100 score calculated? Work through a 60 by hand, see why the same score lands at different percentiles on each trait, and look up your own.

SCHROE Editors

Thursday evening at a neighborhood climbing gym. Between routes, Hana and Jun sit on the crash mat and realize they both took the Big Five personality test that day. They open their results side by side, and the screens don’t look alike. Jun’s shows “People energy” with a big 60 pts and not much else. Hana’s shows the same 60 with extra boxes next to it: Percentile, Average, Sample. Hana scrolls and frowns. “My ‘Emotional sensitivity’ is also 60, and it says Higher 19%. My ‘People energy’ 60 says Lower 46%. Same number. Why so different?”

Hana and Jun are made up for this article, and their answers and scores are illustrative. The way the numbers are produced, though, is exactly what the schroe.kr Big Five (OCEAN) test uses: its scoring code, its question file and its norm file. Below we walk through how the 0–100 score is calculated, why there is a separate Higher/Lower % next to it, why the same 60 lands in different places on different traits, and a table you can use to look up your own position.

First, check which test you took

When you start, you choose one of two lengths. Here they are, labeled as on screen.

| Name on screen | Items | Items per trait | What the result includes | |---|---|---|---| | Standard Test (50 items · 5 min) | 50 | 10 | Five trait scores | | Full Test (120 items · 15 min) | 120 | 24 | Five scores + Percentile, Average, Sample + “Detailed scores” |

Jun took the Standard Test; Hana took the Full Test. That is the first reason their screens differ: the site’s code only calculates a comparison with other respondents on the Full Test. We’ll get to why.

The items come from the IPIP (International Personality Item Pool), a public-domain item bank. Each of the five traits has six facets, each facet has four items, so 5 × 6 × 4 = 120 items make up the Full Test. The Standard Test picks 10 of those per trait, not at random but by a fixed rule, always the same 10. Every facet appears at least once, and the share of reverse-keyed items (explained below) is kept close to the Full Test. Take the Standard Test twice and you’ll see the same questions.

The five names on screen, and the letters O-C-E-A-N

The result screen uses everyday names rather than textbook terms. Their initials in English spell OCEAN.

| Name on the result screen | Letter | Name in the research literature | |---|---|---| | New things | O | Openness | | Planning | C | Conscientiousness | | People energy | E | Extraversion | | Care for others | A | Agreeableness | | Emotional sensitivity | N | Neuroticism |

There’s a reason for the softer names. A word like “Neuroticism” sounds like a verdict, while what the scale actually tracks is a direction: how quickly and strongly worry and tension register for you. We’ll stick with the on-screen names here.

How one answer gets into the score: five steps, and a flip

Each item has five choices. On screen they read Strongly disagree · Disagree · Neutral · Agree · Strongly agree, and the scoring code reads them as 1 through 5.

There’s one twist. On the same “People energy” scale, some items, like “I make friends easily,” mean more people-energy when you agree, while others, like “I prefer to be alone,” mean the opposite. The second kind is called reverse-keyed. The code flips those answers with 6 − answer: 5 becomes 1, 4 becomes 2, and 3 stays 3.

Why mix in reverse-keyed items at all? If every item pointed the same way, a habit of clicking “Agree” regardless of content could push a score in one direction. The question file’s comments state this purpose directly. With items pointing both ways, habitual agreement tends to cancel out.

Jun’s “People energy 60 pts,” worked out by hand

Here are the ten “People energy” items Jun saw on the Standard Test, with his answers. Hollow circles are what he chose; filled circles are the values that went into the score.

Step by step:

Flip: there are three reverse-keyed items. “I avoid contacts with others”: Disagree (2) → 6 − 2 = 4. “I prefer to be alone”: Agree (4) → 6 − 4 = 2. “I wait for others to lead the way”: Neutral (3) → 6 − 3 = 3.
Leave the other seven as they are: 4, 4, 4, 4, 2, 3, 4.
Average: 4 + 2 + 3 + 4 + 4 + 4 + 4 + 2 + 3 + 4 = 34, and 34 ÷ 10 = 3.4.
Stretch to 0–100: the code computes (average − 1) ÷ 4 × 100 and rounds. (3.4 − 1) ÷ 4 × 100 = 60, so the screen shows 60 pts.

That last formula just maps the 1–5 scale onto 0–100. An average of 1 is 0 points, an average of 3 (Neutral) is 50, an average of 5 is 100. The scale along the top of the chart (0, 25, 50, 75, 100) is that mapping.

Why an average and not a total

A total would make the Standard Test (10 items per trait) and the Full Test (24 items) produce numbers of different sizes. An average puts both on the same 1–5 scale regardless of length, and stretching that to 0–100 gives both tests the same unit. What differs is how fine the steps are. On the Standard Test the average moves in steps of 0.1, so scores jump by 2–3 points (3.4 → 60, 3.5 → 63). On the Full Test it moves in steps of 1/24, which is much finer.

Even the rounding matters

An average of 3.5 gives (3.5 − 1) ÷ 4 × 100 = 62.5, and the screen shows the rounded 63. So a one- or two-point gap between you and a friend doesn’t say much: on the Standard Test, moving a single answer by one step can shift the score by 2–3 points.

Hana’s two 60s: “score” and “Percentile” answer different questions

Back to Hana. On the Full Test she scored 60 on both “People energy” and “Emotional sensitivity,” yet the Percentile boxes read Lower 46% and Higher 19%.

Here’s the key. The score asks, “Where on the 1–5 scale did my answers gather?” The Percentile asks, “Where does that score sit among other respondents?” Answering the second question requires a comparison group.

The site’s comparison group, in the words on screen, is “public personality-test respondents.” The norm file names the source as the Johnson IPIP-NEO-120 public sample (published on PsychArchives under CC-BY 4.0), and the denominator is 410,376 people: the valid cases kept out of 619,150 lines of raw data. The screen says so too: “Compared with 410,376 public personality-test respondents. This is not a national average.” In other words, it compares you with a published research sample, not with other schroe.kr users.

The diamonds are the averages of those 410,376 people on each trait. The same 60 lands differently depending on where the average sits.

| Trait | Sample average | Percentile value at 60 | Shown on screen | |---|---|---|---| | New things | 61.56 | 47 | Lower 47% | | Planning | 65.69 | 35 | Lower 35% | | People energy | 60.56 | 46 | Lower 46% | | Care for others | 67.02 | 28 | Lower 28% | | Emotional sensitivity | 45.37 | 81 | Higher 19% |

(Denominator for every trait: 410,376 people)

For “Emotional sensitivity,” the sample average is 45.37, lower than on the other traits. So a 60 sits about 15 points above average and lands on the higher side. For “Care for others,” the average is 67.02, so the same 60 is about 7 points below. On screen that reads: “Emotional sensitivity is about 15 points above average, placing it around Higher 19%.”

How the percentile value is built

The norm file is a lookup table with a percentile value for every score from 0 to 100. The rule: take the number of people below that score plus half the people at exactly that score, divide by 410,376, multiply by 100 and round. Counting ties as half keeps “Higher X%” from being inflated when many people share a score. Values are kept between 1 and 99.

The screen then reads the value one of two ways. At 50 or above, it shows Higher (100 − value)%; below 50, Lower value%. Hana’s “Emotional sensitivity” value is 81, so 100 − 81 = Higher 19%. Her “People energy” value is 46, so Lower 46%.

Why only the Full Test shows a Percentile

The 410,376 comparison respondents answered the 120-item test. A 10-item Standard Test score is on the same 0–100 scale, but with fewer items each answer moves it more. The code comment is explicit: it keeps Standard Test scores from being mapped onto the 120-item norms with the same confidence. The missing box on Jun’s screen wasn’t a bug. It was a choice not to compare when the basis for comparison doesn’t match.

Higher and lower are directions, not grades

Hana paused at “Emotional sensitivity: Higher 19%.” “Higher” somehow sounded like “more problems.” But follow the math and the number says one thing: on the worry-and-tension items, Hana leaned toward “Agree” a bit more than the sample average.

Every Big Five trait is a scale with a different picture at each end. Under the comparison bar on a Full Test result, the labels are “Lower side · Average zone · Higher side,” not “bad · good.”

On the higher side of Emotional sensitivity, this can show up as noticing warning signs early and preparing ahead; on the lower side, as staying steady when things get tense.
On the lower side of Care for others, it can look like stating your own judgment clearly; on the higher side, like checking in on how others feel first.
New things, Planning and People energy work the same way: each end is useful in different situations.

And a position always depends on who you’re compared with. Change the sample, and the same 60 moves. That’s why the screen spells out “This is not a national average.” This test is a tool for reflection, not a diagnosis.

Look up your own position

If you took the Full Test, your Percentile is already on screen. If you took the Standard Test, or want to compare all five traits at once, use the table below. Find the row closest to your score and read across. The values are from the site’s norm file, written the way the screen shows them.

| Your score | New things | Planning | People energy | Care for others | Emotional sensitivity | |---|---|---|---|---|---| | 40 | Lower 4% | Lower 5% | Lower 10% | Lower 3% | Lower 38% | | 45 | Lower 10% | Lower 9% | Lower 15% | Lower 6% | Higher 49% | | 50 | Lower 19% | Lower 16% | Lower 23% | Lower 10% | Higher 38% | | 55 | Lower 31% | Lower 24% | Lower 34% | Lower 17% | Higher 28% | | 60 | Lower 47% | Lower 35% | Lower 46% | Lower 28% | Higher 19% | | 65 | Higher 41% | Lower 45% | Higher 43% | Lower 39% | Higher 13% | | 70 | Higher 27% | Higher 41% | Higher 29% | Higher 45% | Higher 8% | | 75 | Higher 15% | Higher 29% | Higher 17% | Higher 29% | Higher 4% | | 80 | Higher 8% | Higher 18% | Higher 8% | Higher 15% | Higher 2% |

(Sample for every cell: 410,376 people, Johnson IPIP-NEO-120 public sample)

Two cautions. First, putting a Standard Test score into this table gives you only a rough sense, for the very reason the site doesn’t show a Percentile there. Second, 50 is not “the middle.” A 50 is the “Neutral” point on the 1–5 scale, not the middle of the sample, so depending on the trait it can be Lower 10% or Higher 38%.

If you took the Standard Test, you can also work backward from your score to your average answer.

| Score on screen | Average of ten items (after flipping) | Roughly what the answers look like | |---|---|---| | 0 | 1.0 | Every item at one end | | 25 | 2.0 | Mostly on the “Disagree” side | | 50 | 3.0 | Balanced, or lots of “Neutral” | | 60 | 3.4 | A little more on the “Agree” side | | 75 | 4.0 | Mostly on the “Agree” side | | 100 | 5.0 | Every item at the other end |

Back at the climbing gym

Before the next route, Jun decides to take the Full Test. Not for the Percentile this time, but for the “Detailed scores” the Standard Test doesn’t show. Inside a single 60, comfort with meeting people and appetite for excitement can come out quite differently.

Hana writes a note next to “Emotional sensitivity: Higher 19%”: “60 among people who average 45. The me who checks the packing list three times the night before an exam.” The number didn’t rule on her. It put a coordinate on something she already knew about herself.

In short:

The 0–100 score is the average answer on that trait (reverse-keyed items flipped with 6 − answer), stretched with (average − 1) ÷ 4 × 100 and rounded.
The Percentile (Higher/Lower %) appears only on the Full Test, from a lookup table built on a public sample of 410,376 people. Each trait has its own average, so the same score can land in different places.
50 is where “Neutral” answers sit, not the middle of the crowd.
Higher and lower are directions. Neither end is the better personality.

Curious what your five scores look like? Try the Big Five personality test.

Big Five results are for self-reflection, not diagnosis or evaluation. The calculations follow the schroe.kr Big Five test’s scoring code, question file and norm file as deployed in October 2026. Hana, Jun, their answers and scores are illustrative. Comparison statistics use the Johnson IPIP-NEO-120 public sample (410,376 people) as the denominator; no schroe.kr user results were used.

Try it yourself

Take the Big Five test