• Foundation
  • About 13 min
  • 5 practice questions

The normal curve and deviation IQ: how scores are spread

Learn the bell-shaped normal curve, the 68-95-99.7 rule, and how a deviation IQ is worked out with a mean of 100 and a stated standard deviation.

If something looks wrong, please tell us.

When many people take an intelligence test, their scores are not spread evenly. Most people score near the middle and very few score very high or very low. This pattern is so regular that it has a name, the normal distribution, and it is the basis of the modern deviation IQ. This lesson explains the curve and shows how to read proportions from it, using only arithmetic.

The normal curve

The normal distribution is shaped like a bell. It is symmetrical, so the left half is the mirror of the right half. The mean, the median and the mode all lie at the centre, and the curve falls away smoothly on both sides. Most scores lie near the mean, and the further a score is from the mean, the fewer people have it. The usual measure of spread is the standard deviation (SD), which tells how far scores typically lie from the mean. A small SD means scores are bunched close to the mean; a large SD means they are spread out.

The 68-95-99.7 rule

Approximate shares in a normal distribution
Distance from the meanShare of people
Within 1 SD (either side)about 68 percent
Within 2 SDabout 95 percent
Within 3 SDabout 99.7 percent

Because the curve is symmetrical, half of the people outside 1 SD are above and half below. So about (100 - 68) / 2 = 16 percent lie more than 1 SD above the mean, and about (100 - 95) / 2 = 2.5 percent lie more than 2 SD above it. These shares are approximate.

Deviation IQ

Because the ratio IQ does not work for adults, modern tests give a deviation IQ: a score that shows how far a person is from the average of the person's own age group. The mean is set at 100. The SD is set by the test; many tests use 15, and some use 16, so a question will always tell you which one to use. A person's z-score is the number of SDs from the mean: z = (score - mean) / SD. Then deviation IQ = 100 + (z x SD).

Worked example 1: reading proportions

On a test with mean 100 and SD 15, find the score range of the middle 68 percent and the share scoring above 130. Working: within 1 SD means 100 - 15 = 85 up to 100 + 15 = 115, so about 68 percent score between 85 and 115. A score of 130 is (130 - 100) / 15 = 2 SDs above the mean. About 2.5 percent score above 2 SD, so about 2.5 percent score above 130, or roughly 1 person in 40. In a school of 800 pupils drawn from the general population, about 800 x 0.68 = 544 would be expected between 85 and 115. (This is a general expectation; a real school may differ.)

Worked example 2: z-scores and conversion

A test has mean 100 and SD 16. Meena scores 124. z-score: (124 - 100) / 16 = 1.5. If the same standing were reported on a test with SD 15, her score would be 100 + (1.5 x 15) = 122.5. The two numbers, 124 and 122.5, mean the same standing, because the SDs differ. This is why one must know the SD before comparing IQ scores from different tests. Another pupil scores 70 on a test with SD 15: z = (70 - 100) / 15 = -2, which is 2 SDs below the mean, a position shared by about 2.5 percent of people.

z = (score - mean) / SD. Deviation IQ = 100 + z x SD. About 68, 95 and 99.7 percent lie within 1, 2 and 3 SD of the mean.

Key takeaways

  • The normal curve is symmetrical, bell-shaped, with mean, median and mode at the centre.
  • The standard deviation measures spread, and the 68-95-99.7 rule gives approximate shares.
  • Deviation IQ compares a person with others of the same age, with mean 100 and the test's stated SD.
  • Always check the SD before comparing scores from different tests.

Check your understanding

  1. 1. On a test with mean 100 and SD 15, what score is 1 SD above the mean?

    1. 85
    2. 130
    3. 115
    4. 101
    Show answer

    Answer: C. 100 + 15 = 115.

  2. 2. About what percentage of people score within 1 SD of the mean in a normal distribution?

    1. 68 percent
    2. 50 percent
    3. 95 percent
    4. 99.7 percent
    Show answer

    Answer: A. The rule gives about 68 percent within 1 SD.

  3. 3. A test has mean 100 and SD 15. What is the z-score for 70?

    1. 2
    2. -1
    3. -3
    4. -2
    Show answer

    Answer: D. (70 - 100) / 15 = -2.

  4. 4. Roughly what share of people lie more than 2 SD above the mean?

    1. 5 percent
    2. 2.5 percent
    3. 16 percent
    4. 34 percent
    Show answer

    Answer: B. About 95 percent lie within 2 SD, so (100 - 95) / 2 = 2.5 percent lie above.

  5. 5. Why must the SD be known before comparing IQ scores from two tests?

    1. The same score can mean different standings
    2. The mean must be recalculated for each pupil
    3. The z-score is always equal to zero here
    4. SD is just another name for mental age
    Show answer

    Answer: A. A score of 124 with SD 16 equals z = 1.5, while 124 with SD 15 is a higher standing.

Privacy · Terms · Refunds · Delivery