ACT Statistics: Distributions, Spread, Standard Deviation, and Data Comparisons

Statistics questions on ACT Math are rarely about grinding through a long standard-deviation formula.

They are usually about understanding what a data set is doing.

The current ACT Math description puts Statistics & Probability at roughly 12–15% and explicitly includes:

• center;
• spread;
• data collection;
• bivariate relationships;
• probabilities;
• sample spaces.

The expert habit is:

separate center from spread

Two data sets can have the same mean but very different variability.

Center: mean and median

Mean:

$$\boxed{\bar x=\frac{\text{sum of values}}{\text{number of values}}}$$

Median:

middle value after ordering.

Example:

$$2,4,5,7,12$$

Mean:

$$\frac{30}{5}=6$$

Median:

$$5$$

They measure center differently.

Outliers affect mean more than median

Data:

$$10,11,12,13,14$$

Mean:

$$12$$

Median:

$$12$$

Replace 14 with 100:

$$10,11,12,13,100$$

Median remains:

$$12$$

Mean rises sharply.

Therefore median is more resistant to extreme values.

Spread: what does it mean?

Spread measures how dispersed the data are.

Common measures include:

• range;
• interquartile range;
• standard deviation.

A data set tightly clustered around its mean has low spread.

A data set scattered far from its mean has high spread.

Range

$$\boxed{\text{range}=\max-\min}$$

Example:

$$3,5,8,12$$

Range:

$$12-3=9$$

Simple, but sensitive to outliers.

Interquartile range

$$\boxed{IQR=Q_3-Q_1}$$

It describes spread of the middle 50% of the data.

Because it ignores extremes, IQR is more resistant to outliers than range.

Standard deviation: conceptual meaning

Standard deviation measures typical distance from the mean.

You usually need to know:

• larger SD → more spread;
• smaller SD → more clustered.

If all values are identical:

$$5,5,5,5$$

then:

SD=0

because every value equals the mean.

Compare same mean, different spread

Set A:

$$8,9,10,11,12$$

Set B:

$$0,5,10,15,20$$

Both means:

$$10$$

But B is much more spread out.

Therefore:

$$\boxed{\text{SD}_B>\text{SD}_A}$$

No full formula needed.

Adding a constant to every value

Suppose every value increases by 7.

Effects:

• mean increases by 7;
• median increases by 7;
• quartiles increase by 7;
• range unchanged;
• IQR unchanged;
• standard deviation unchanged.

Why?

All pairwise distances and distances from mean remain the same.

This is a very high-value ACT pattern.

Multiplying every value by a constant

If every value is multiplied by 3:

• mean × 3;
• median × 3;
• range × 3;
• IQR × 3;
• standard deviation × 3 in magnitude.

Scaling changes spread.

Translation does not.

Negative scaling

If every value is multiplied by $-2$:

center changes sign and scale.

Standard deviation multiplies by:

$$|-2|=2$$

because spread is nonnegative.

Box plots

A box plot summarizes:

• minimum;
• $Q_1$;
• median;
• $Q_3$;
• maximum

or uses whiskers based on an outlier convention.

The box itself spans:

$$Q_1\text{ to }Q_3$$

so its length represents IQR.

Compare box plots

Look separately at:

Center

median line.

Middle spread

box width/length = IQR.

Overall spread

whisker/extreme extent.

Skew/outliers

asymmetry and isolated points if shown.

Do not say:

one distribution is larger

without specifying whether you mean center or spread.

Histograms

A histogram shows frequencies across intervals.

Use it to analyze:

• shape;
• center;
• spread;
• skew;
• possible modes.

But you may not know exact raw values.

ACT often asks qualitative comparisons rather than exact calculations.

Symmetric distribution

In a perfectly symmetric unimodal distribution:

• mean and median are often similar.

If the distribution is strongly skewed, mean is pulled toward the long tail.

Right-skewed distribution

Long tail to right.

Usually:

$$\text{mean}>\text{median}$$

because large values pull the mean upward.

Left-skewed distribution

Long tail to left.

Usually:

$$\text{mean}<\text{median}$$

Normal distribution

A normal distribution is symmetric and bell-shaped.

Typical empirical-rule approximation:

• about 68% within 1 SD;
• about 95% within 2 SD;
• about 99.7% within 3 SD.

Use only when the problem establishes an approximately normal distribution.

Z-score

$$\boxed{z=\frac{x-\mu}{\sigma}}$$

It tells how many standard deviations a value lies above/below mean.

Example:

mean $=70$, SD $=5$, value $=80$:

$$z=\frac{80-70}{5}=2$$

So value is:

2 SD above mean.

Compare scores from different distributions

Student A:

• score 85;
• class mean 75;
• SD 5.

$$z_A=2$$

Student B:

• score 92;
• class mean 80;
• SD 8.

$$z_B=1.5$$

Even though B has higher raw score, A is farther above their class mean.

Standardization allows comparison.

Outliers and standard deviation

An extreme outlier generally increases standard deviation because it creates a large distance from the mean.

It can also shift the mean itself.

Do not assume median and IQR react as strongly.

Sample vs population

ACT may use:

• sample data;
• population data.

A sample statistic estimates a population parameter.

You usually do not need advanced notation, but you should understand that conclusions depend on how the sample was collected.

See the broader ACT Statistics and Probability guide for sampling and probability.

Bivariate data

When data involve pairs:

$$(x,y)$$

you may see:

• scatterplots;
• association;
• line of best fit.

Statistics is not only one-variable distributions.

ACT explicitly includes relationships in bivariate data.

Correlation is not causation

If two variables move together, that does not prove one causes the other.

Possible explanations:

• lurking variable;
• reverse causality;
• coincidence;
• selection bias.

This is as much reasoning as calculation.

Worked example: adding constant

Original mean:

$$20$$

Original SD:

$$4$$

Add 10 to every value.

New mean:

$$30$$

New SD:

$$\boxed{4}$$

Worked example: multiplying

Original mean:

$$12$$

SD:

$$3$$

Multiply every value by 5.

New mean:

$$60$$

New SD:

$$\boxed{15}$$

Worked example: outlier

Set A:

$$4,5,5,6,5$$

Set B:

$$4,5,5,6,50$$

B has:

• larger mean;
• much larger range;
• larger SD.

Median changes little.

Worked example: z-score

Mean:

$$100$$

SD:

$$15$$

Value:

$$70$$ $$z=\frac{70-100}{15}=-2$$

So:

2 SD below mean.

Common ACT statistics traps

Trap 1

Confusing mean and median.

Trap 2

Assuming higher mean means higher spread.

Trap 3

Adding a constant and changing SD.

Trap 4

Multiplying data and leaving SD unchanged.

Trap 5

Calling correlation causation.

Trap 6

Comparing raw scores across different distributions instead of z-scores.

Fast ACT distribution routine

When comparing two distributions, ask separately:

1. Which has higher center?
2. Which has greater spread?
3. Which has outliers?
4. Is one skewed?
5. Is a transformation applied?
6. Does the question ask raw value or standardized position?

Standard-deviation estimation without formula

If two sets have same mean, compare distances from mean.

Set A:

$$9,10,10,10,11$$

Set B:

$$0,5,10,15,20$$

B clearly has much larger deviations.

No need to calculate variance term by term.

Current ACT relevance

Current ACT Math explicitly includes:

• center and spread of distributions;
• data collection;
• bivariate relationships;
• probability and sample spaces.

Statistics & Probability is about 12–15% of current Math content.

Practice set

1

Data:

$$2,2,2,2$$

Standard deviation:

$$\boxed{0}$$

2

If 5 is added to every score, SD:

unchanged.

3

If every score is doubled, SD:

doubled.

4

Mean 50, SD 10, score 70:

$$z=\boxed{2}$$

How to review mistakes

Bad note:

statistics question wrong.

Better:

I treated a shift of +8 as changing spread, but translation changes center only.

Or:

I compared raw scores instead of how many standard deviations above the mean they were.

Compare distributions without raw data

ACT may show two histograms or box plots rather than lists.

You can still compare:

Center

Where is the distribution located?

Spread

How wide is it?

Shape

Symmetric, skewed, bimodal?

Outliers

Are there isolated extreme values?

Do not invent exact means from a graph if the information does not support them.

Mean and standard deviation under linear transformation

If:

$$Y=aX+b$$

then:

$$\mu_Y=a\mu_X+b$$

and:

$$\sigma_Y=|a|\sigma_X$$

This summarizes the earlier rules.

Adding $b$:

shifts center.

Multiplying by $a$:

scales center and spread.

Example: temperature conversion

Suppose Fahrenheit is:

$$F=\frac95C+32$$

If Celsius distribution has:

• mean $20$;
• SD $4$;

then Fahrenheit mean:

$$\frac95(20)+32=68$$

SD:

$$\frac95(4)=7.2$$

The +32 does not affect spread.

Weighted mean

If groups have different sizes, you cannot average group means equally unless group sizes match.

Example:

• 10 students average 80;
• 30 students average 90.

Overall mean:

$$\frac{10(80)+30(90)}{40}$$ $$=\frac{3500}{40}=87.5$$

Not:

$$85$$

This connects distribution reasoning with weighted averages.

Median from a cumulative count

If 21 values are ordered, median is 11th.

If a frequency table shows:

• values 1–5 appear 4 times;
• 6–10 appear 9 times;
• 11–15 appear 8 times;

the 11th observation falls in the 6–10 interval.

Even without raw data, position reasoning can identify the median interval.

Percentiles

A percentile tells relative position.

If score is at 80th percentile:

about 80% of the reference group is at or below that score.

It does not mean:

80% of questions were answered correctly.

Do not mix percentile rank with percent correct.

Standard deviation and normal distributions

If normal with mean:

$$100$$

and SD:

$$15$$

then about 68% lie between:

85 and 115

About 95% lie between:

70 and 130

This is useful for quick probability approximations when the normal model is explicitly appropriate.

Same IQR, different range

Two box plots can have equal box lengths but very different whiskers.

That means:

• same middle-50% spread;
• different overall spread.

Be precise about which spread measure the question asks.

Outlier rules and IQR

A common outlier convention marks values below:

$$Q_1-1.5(IQR)$$

or above:

$$Q_3+1.5(IQR)$$

If ACT provides quartiles and asks about potential outliers, this rule may be relevant.

Bivariate residual idea

For line-of-best-fit questions:

$$\text{residual}=\text{observed}-\text{predicted}$$

Positive residual:

actual point lies above predicted line value.

Negative:

below.

This is another form of "distance from a model," conceptually related to spread but not the same as standard deviation.

Review checklist

Classify statistics misses as:

• center vs spread;
• mean vs median;
• transformation effect;
• outlier effect;
• z-score;
• percentile meaning;
• weighted mean;
• graph interpretation;
• correlation/causation.

That prevents treating Statistics as one vague topic.

FAQ: ACT standard deviation

Do I need to calculate standard deviation by hand?

ACT more often rewards conceptual understanding of spread, transformations, and comparisons, though a direct calculation can still be possible.

Does adding a constant change standard deviation?

No.

Does multiplying every value change standard deviation?

Yes, by the absolute value of the multiplier.

Does an outlier affect mean or median more?

Usually mean.

What does a z-score of -1.5 mean?

The value is 1.5 standard deviations below the mean.

Related MastaPrep guides

• ACT Statistics and Probability
• ACT Math Formulas
• ACT Functions
• Enhanced ACT 2026–27 Guide

Final takeaway

Separate:

center

from:

spread

Then remember the transformation rule:

• add/subtract constant → center changes, spread does not;
• multiply/divide by factor → both center and spread scale.

That single framework solves a surprising number of ACT statistics questions.

---

Official ACT references:
https://www.act.org/content/act/en/products-and-services/the-act/test-preparation/description-of-math-test.html
https://www.act.org/content/act/en/college-and-career-readiness/standards/mathematics-standards.html
https://www.act.org/content/dam/act/unsecured/documents/R2519-Design-Framework-for-the-ACT-Enhancements-2026-02.pdf

Trademark note: ACT® is a registered trademark of ACT Education Corp. MastaPrep is independent and is not affiliated with or endorsed by ACT Education Corp.