What is IQ?
IQ compares a person with a group. It does not name one talent.
[Test items] → [Raw correct] → [Age group]
→ [Scaled score near 100]
What the number compares
An IQ score starts as a count of correct answers on a named test. That raw count is then compared with people of a similar age who took the same edition.
The publisher rescales the ranks so the middle of that group lands near 100. On many Wechsler-style batteries the spread is a standard deviation of about 15.
A higher number means a higher rank in that reference group, not a larger brain region with a public size. Two people can share a score and still differ in school skills, music, or social judgment.
The score also belongs to a date. Norms age, and a number from a 1990 manual is not the same object as a number from a current manual.
Because the scale is a rank, the gap from 100 to 115 is not the same kind of thing as fifteen extra kilograms. You cannot add two IQ scores and get a family total that means anything on the test.
Two kinds of thinking tests often mix
Researchers often split broad ability into fluid reasoning and crystallized knowledge. Fluid reasoning is spotting a new pattern when the words are unfamiliar.
Crystallized knowledge is using facts and skills you already stored, such as vocabulary. A matrix puzzle leans on the fluid side.
A vocabulary item leans on the crystallized side. A full battery can include both, which is why one short puzzle cannot stand in for the whole score.
Culture and language move crystallized scores more easily than a well-made picture matrix. That is a reason picture tests exist, not a reason every picture quiz on the internet is fair.
┌────────────────────┬────────────────────┐
│ Fluid reasoning │ Crystallized skill │
│ new patterns │ stored facts │
│ matrices, rules │ words, procedures │
└────────────────────┴────────────────────┘
How to read a score without inflating it
A useful score names the test. If the page only says IQ and a percent sign, you do not yet know which yardstick was used.
Ask which edition was given. The fourth and fifth editions of a battery do not share every item, and their norm tables are rebuilt.
Ask the age band. A scaled number is a comparison with people near that age, not a raw count of correct marks.
Ask the setting. A noisy room, a translated instruction, or a phone screen can change the raw count before any scaling happens.
The bell curve is a picture of many scores piled near 100 and fewer scores far away. It is a model of the norm group, not a portrait of one person.
Distance from 100 is often described in standard deviations. On an SD 15 scale, 115 is about one step above the middle, and 85 is about one step below.
Those steps are ranks. They do not measure liters of ability the way a jug measures water.
Fluid items reward a fresh rule. Crystallized items reward knowledge that had time to settle. A single hobby can lift one side and leave the other still.
Schools sometimes see the crystallized side more clearly because lessons store words and procedures. A quiet puzzle hour leans the other way.
Neither side is a moral grade. A low crystallized score can mean a new language, not a closed mind.
A high fluid score can mean comfort with pictures, not wisdom about people.
When someone quotes only the highest number they have ever seen, ask whether that number was the full scale or one narrow index.
Indexes inside a battery can disagree. Working memory and processing speed are not the same index as verbal comprehension.
That spread is often more informative than the single average. The average hides which part carried the day.
Retesting too soon recycles memory of the items. A gain of a few points across a short gap is a familiar finding in manuals.
A gain across years, with schooling or health in between, is a different story and needs its own record.
Games that feel similar to matrix items train the look of those items. They do not automatically rewrite every index.
If you want the concept in one line, say this: IQ is a ranked result on a named test, dated and compared with an age group.
Anything shorter starts to sound like a label for a whole life. The test was never that large.
Keep one more distinction nearby. A percentile says how many people in the norm group scored lower. An IQ number of 100 is built so it sits near the 50th percentile on that test, and the conversion is not a second mystery.
If a report gives both, they should tell the same story. If they fight, someone mixed scales.
Teachers, coaches, and friends are allowed to notice skills the test never asked about. The score does not veto those observations. It only ranks the sitting that was actually given.
Trio IQ in this story
Trio IQ asks you to find sets of cards whose attributes follow one shared rule. The screen shows a clock and a result for that round. It does not print an IQ score, and it is not a clinical instrument.
Takeaway
[Read the rule] → [Try one example] → [Play one round]
| Step | Do this |
|---|---|
| Name the idea | Say it in one sentence before you trust a chart. |
| Check the source | Prefer a published test manual over a nameless quiz. |
| Practice a rule | Open Play and use the rule on cards. |
Further reading
- American Psychological Association, intelligence topics: https://www.apa.org/topics/intelligence
- Mensa International, what the society is: https://www.mensa.org/
- A public overview of Raven matrices: https://en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
Score ranges · Practice and scores
Learn home · Privacy · Play