GRE Standard Deviation: Meaning, Comparisons, and Examples

Standard deviation is a measure of spread. On the GRE, many useful questions can be solved by understanding how data values sit around their mean rather than by calculating a long formula. A set whose values cluster closely together has a small standard deviation. A set with values far from the mean has a larger standard deviation. The mean describes the center; the standard deviation describes dispersion around that center.

What standard deviation tells you

Imagine two groups with the same average score of 70. In one group, nearly everyone scores between 68 and 72. In another, several people score near 40 and several near 100. The means can match even though the second group is much more spread out. Its standard deviation is larger because its values are typically farther from the mean.

Standard deviation is expressed in the same units as the original data. If scores are measured in points, standard deviation is measured in points. Variance, by contrast, is expressed in squared units. A question may mention variance, but many comparison problems can be answered without computing either statistic exactly.

Standard deviation does not tell you whether a data set is good, bad, or representative. A small spread can mean consistent values, but it does not imply a high mean. A large spread does not reveal how many values are above or below the mean unless the distribution's shape and other details are provided. Do not infer more than the statistic supports.

The distance-from-the-mean idea

Conceptually, standard deviation summarizes deviations from the mean. A deviation is a value minus the mean. Positive and negative deviations balance, so simply averaging signed deviations would always produce zero. Squaring deviations makes them nonnegative; averaging them yields a variance, and taking the square root returns to the original measurement units. For GRE reasoning, this helps explain why unusually distant values increase spread.

You usually do not need to compute a complicated square root on a GRE comparison item. The test often supplies a relationship or asks how a simple transformation affects the statistic. Use the structure of the values. If every value is shifted equally, each value remains the same distance from the shifted mean. If all distances are scaled, the standard deviation scales with them.

Transformation rules to know

For example, converting temperatures from Celsius to Fahrenheit uses F = (9/5)C + 32. The constant 32 shifts every value and does not affect standard deviation. Multiplication by 9/5 scales distances, so the standard deviation in Fahrenheit is 9/5 times the standard deviation in Celsius. The new units matter: a spread of 5 degrees Celsius becomes a spread of 9 degrees Fahrenheit.

If all values are multiplied by −3, their order reverses and signs change, but distances from the mean become three times as large. Standard deviation cannot be negative, so the multiplier is its absolute value, 3. A transformation such as x → −x changes the mean's sign but preserves the standard deviation.

Original example: compare two sets

Set A contains 4, 5, 6. Set B contains 1, 5, 9. Both have mean 5. The distances from 5 in Set A are 1, 0, and 1. In Set B they are 4, 0, and 4. Set B is more spread out, so its standard deviation is larger. You can answer without choosing a population or sample formula because both sets have the same number of observations and the comparison is clear either way.

Now add 10 to every entry. The sets become 14, 15, 16 and 11, 15, 19. Their means shift from 5 to 15, but the distances from their means stay exactly the same. Each standard deviation is unchanged. This is a useful check: if your calculation says adding a constant changes standard deviation, you have confused center with spread.

Original GRE-style practice

Question 1: a shift

A data set has mean 12 and standard deviation 3. Each value is increased by 7. What are the new mean and standard deviation?

1. Mean 19; standard deviation 3
2. Mean 19; standard deviation 10
3. Mean 12; standard deviation 10
4. Mean 84; standard deviation 21

Answer: A. Adding 7 to every value raises the mean by 7, to 19. It does not change any distance from the mean, so standard deviation remains 3. The standard deviation is not added to the shift.

Question 2: a scale change

A list of measurements has standard deviation 4. Every measurement is multiplied by 2.5. What is the new standard deviation?

1. 1.6
2. 4
3. 6.5
4. 10

Answer: D, 10. Multiplying all values by a positive factor multiplies every distance from the mean by that factor. Thus the new standard deviation is 2.5 × 4 = 10. Dividing by 2.5 would reverse the transformation rather than apply it.

Set X has values 2, 6, 10. Set Y is formed by multiplying every value in X by −2 and then adding 5. Which statement is true?

1. Y has mean 6 and the same standard deviation as X.
2. Y has mean −5 and twice the standard deviation of X.
3. Y has mean −7 and twice the standard deviation of X.
4. Y has mean 5 and half the standard deviation of X.

Answer: C. The mean of X is 6. Multiplying by −2 gives a mean of −12, then adding 5 gives −7. The spread is multiplied by |−2| = 2, so the standard deviation doubles. Applying the transformation to each value verifies the result: 2, 6, 10 become 1, −7, −15, whose mean is −7.

Question 4: identifying greater spread

Which set has the greater standard deviation? Set P is 20, 21, 22, 23, 24. Set Q is 18, 20, 22, 24, 26.

1. P
2. Q
3. They are equal
4. Cannot be determined

Answer: B. Both means are 22. Set P's values range from 20 to 24; Set Q's range from 18 to 26 and places its values farther from 22. Because both are symmetric with the same number of values and same center, Q has the larger standard deviation. In other problems, range alone is not enough to determine standard deviation, but here the listed values show the full pattern.

Standard deviation and outliers

A value far from the mean tends to increase standard deviation because its squared deviation contributes substantially to the total. Consider 3, 4, 5, whose mean is 4, compared with 3, 4, 15, whose mean is 22/3. The added high value changes both the center and spread. Do not simply compare the distance of 15 from the old mean; the mean moves too. Still, the widely separated observation makes the second data set much more dispersed.

If a question asks what happens when an extreme value is replaced by one closer to the mean, the spread will generally decrease when the rest of the data are held fixed and the replacement moves toward the center. But be careful with wording: replacing a value can also shift the mean. If the problem supplies exact values, calculate or reason about the new deviations rather than assuming the old mean stays fixed.

What standard deviation does not establish

The same standard deviation can occur in data sets with different shapes. One set might cluster around the mean with a few distant points; another might be spread evenly. Standard deviation alone does not provide the minimum, maximum, median, or exact proportion of observations within a particular interval.

Do not assume that a fixed percentage of values lies within one standard deviation unless a distributional model is stated. For normally distributed data, familiar empirical-rule percentages apply approximately, but a GRE problem must provide enough information to justify that model. For arbitrary data, a mean and standard deviation do not determine the shape.

A standard deviation of zero has a precise meaning: every observation equals the mean, so every value in the data set is identical. If at least two values differ, standard deviation is positive. This gives a quick way to assess statements that claim a nonconstant data set has zero spread.

Comparing distributions with different means

A higher mean does not imply a larger standard deviation. Compare these two lists: 10, 11, 12 and 100, 101, 102. The second has a much higher mean, but the lists have identical spreads because one is obtained by adding 90 to every value in the other. This is one of the most common conceptual traps.

Likewise, a data set with a lower mean can be more variable. Separate the questions: first find or compare the centers, then inspect the distances around each center. If the problem applies the same additive shift to every value, immediately recognize that spread is preserved.

Standard deviation versus range and variance

Range is maximum minus minimum. It depends only on two values; standard deviation reflects all observations. Two sets can have the same range and different standard deviations if one clusters near the middle while the other places more values near the endpoints. The range can be a helpful first clue, but it is not a substitute for analyzing the complete data set.

Variance is the square of standard deviation. If standard deviation is 6, variance is 36 in squared units. If values are scaled by 3, standard deviation triples and variance is multiplied by 9. GRE questions sometimes switch between these terms; keep the square relationship clear.

A strategy for quantitative comparison

1. Check whether the sets have the same center. A shift in center does not by itself change spread.
2. Look for a transformation applied to every value. Addition preserves standard deviation; multiplication scales it by the absolute factor.
3. If no transformation is stated, compare distances from each set's own mean, not from zero or a shared external value.
4. Use exact calculation only when the listed values or information require it. A conceptual comparison can be faster and more reliable.
5. Reject conclusions about distribution shape or fixed proportions unless the prompt supplies that information.

In a GRE quantitative comparison, decide whether the relationship is always true for the stated values. If the problem gives only a mean, there may be many possible sets with different standard deviations. Constructing two examples that satisfy the same mean but have different spreads can show that the relationship cannot be determined.

A cannot-be-determined example

Suppose a question states that four positive integers have mean 10 and asks whether their standard deviation is greater than 0. It is greater than 0 because four distinct positive integers are not all equal. If instead the values are four real numbers with mean 10, they could all equal 10, giving standard deviation 0, or they could differ, giving a positive standard deviation. The domain of the numbers changes what can be concluded.

For comparison questions, use the stated constraints literally. Words such as integer, positive, distinct, or consecutive can narrow the possibilities. If no constraint rules out equal values, do not assume the data must vary.

Practice plan

Practice transformations before formula calculations. Make small sets and shift them, scale them, reverse their signs, or replace one value. Predict the new mean and standard deviation, then verify using distances. This hands-on reasoning makes the rules memorable and prepares you for unfamiliar wording.

For each missed item, record whether the error concerned center, spread, scale, or an unsupported assumption about distribution. Rework the item by stating the mean transformation and spread transformation separately. ETS's free mathematics review can refresh statistical foundations; use original mixed problems afterward so you can transfer the concepts to tables, data descriptions, and comparisons.

How often standard deviation appears

Statistics is part of quantitative reasoning, but ETS does not publish a guaranteed count of standard-deviation questions for each administration. The current GRE includes 27 quantitative questions over two sections. Prepare this concept as part of a broader foundation in statistics and data interpretation instead of planning around a predicted item quota.