Percentiles and Quartiles
Master percentiles, quartiles, and the five-number summary. Learn to interpret relative standing and identify outliers using the IQR method.
On This Page
What Are Percentiles?
A percentile indicates the percentage of data values that fall at or below a given value. Percentiles describe relative standing within a dataset.
If your score is at the 75th percentile:
- 75% of scores are at or below yours
- 25% of scores are above yours
- You performed better than approximately 3/4 of test-takers
Common applications:
- Standardized test scores (SAT, GRE)
- Growth charts for children
- Income distributions
- Performance rankings
Calculating Percentiles
Finding the Percentile of a Value
To find what percentile a value represents:
Data (sorted): 52, 58, 63, 67, 71, 74, 78, 82, 85, 91
What percentile is a score of 74?
Values ≤ 74: 52, 58, 63, 67, 71, 74 → 6 values Total values: 10
A score of 74 is at the 60th percentile.
Finding a Value at a Given Percentile
To find the value at the th percentile:
Where:
- = position (location) in the sorted data
- = desired percentile
- = number of data points
Data (sorted): 12, 18, 23, 27, 31, 35, 42, 48, 55, 62 (n = 10)
Find the 40th percentile:
Position 4.4 means: 40% of the way between the 4th and 5th values.
- 4th value = 27
- 5th value = 31
The 40th percentile is 28.6.
Quartiles: Special Percentiles
Quartiles divide data into four equal parts:
| Quartile | Percentile | Meaning |
|---|---|---|
| Q1 | 25th | First quartile - 25% below |
| Q2 | 50th | Second quartile = Median |
| Q3 | 75th | Third quartile - 75% below |
|-------|-------|-------|-------|
Min Q1 Q2 Q3 Max
(25%) (50%) (75%)
↓ ↓ ↓
25% 50% 75%
below below below
Each section contains 25% of the data. Calculating Quartiles
Method 1: Using Percentile Formula
- Q1 = 25th percentile
- Q2 = 50th percentile (median)
- Q3 = 75th percentile
Method 2: Median of Halves
- Find the median (Q2)
- Q1 = median of the lower half
- Q3 = median of the upper half
Data (sorted): 3, 7, 12, 15, 18, 21, 24, 28, 33 (n = 9)
Step 1: Find Q2 (median)
- Middle position = (9+1)/2 = 5th value
- Q2 = 18
Step 2: Find Q1 (median of lower half)
- Lower half: 3, 7, 12, 15
- Q1 = (7 + 12)/2 = 9.5
Step 3: Find Q3 (median of upper half)
- Upper half: 21, 24, 28, 33
- Q3 = (24 + 28)/2 = 26
Quartiles: Q1 = 9.5, Q2 = 18, Q3 = 26
Data (sorted): 5, 8, 12, 15, 19, 22, 26, 31 (n = 8)
Step 1: Find Q2 (median)
- Q2 = (15 + 19)/2 = 17
Step 2: Find Q1 (median of lower half)
- Lower half: 5, 8, 12, 15
- Q1 = (8 + 12)/2 = 10
Step 3: Find Q3 (median of upper half)
- Upper half: 19, 22, 26, 31
- Q3 = (22 + 26)/2 = 24
Quartiles: Q1 = 10, Q2 = 17, Q3 = 24
The Five-Number Summary
The five-number summary provides a compact description of a distribution:
- Minimum (Min)
- First Quartile (Q1)
- Median (Q2)
- Third Quartile (Q3)
- Maximum (Max)
Data: 12, 18, 23, 27, 31, 35, 42, 48, 55, 62, 71, 85
| Measure | Value |
|---|---|
| Minimum | 12 |
| Q1 | 25 |
| Median | 38.5 |
| Q3 | 58.5 |
| Maximum | 85 |
Interpretation:
- Range: 85 - 12 = 73
- Middle 50% of data is between 25 and 58.5
- The distribution appears slightly right-skewed (larger gap above median)
Interquartile Range (IQR)
The IQR measures the spread of the middle 50% of data.
Given Q1 = 25 and Q3 = 58.5:
The middle 50% of values spans 33.5 units.
Why Use IQR?
| Measure | Affected by Outliers? | Best For |
|---|---|---|
| Range | Yes (very sensitive) | Quick overview |
| Standard Deviation | Yes (somewhat) | Symmetric distributions |
| IQR | No (resistant) | Skewed data, outliers present |
Detecting Outliers with IQR
An outlier is an observation unusually far from other values. The IQR method defines outliers as:
Values outside these fences are outliers.
Data: 2, 15, 18, 21, 24, 27, 30, 33, 36, 42, 98
Step 1: Find quartiles
- Q1 = 18, Q3 = 36
- IQR = 36 - 18 = 18
Step 2: Calculate fences
- Lower fence = 18 - 1.5(18) = 18 - 27 = -9
- Upper fence = 36 + 1.5(18) = 36 + 27 = 63
Step 3: Identify outliers
- Below -9? No values
- Above 63? 98 is an outlier!
Also note: 2 is low but not an outlier (2 > -9)
Extreme Outliers
Some analysts distinguish mild and extreme outliers:
- Mild outlier: Beyond 1.5 × IQR but within 3 × IQR
- Extreme outlier: Beyond 3 × IQR
Comparing Distributions with Percentiles
Percentiles allow fair comparison across different scales.
Student A: 720 on Math SAT (75th percentile) Student B: 28 on ACT Math (75th percentile)
Different tests, different scales, but same relative performance! Both scored better than 75% of test-takers.
Box Plots: Visualizing the Five-Number Summary
A box plot (box-and-whisker plot) displays the five-number summary graphically.
Min Q1 Q2 Q3 Max
|---------|========|=========|---------|
12 25 38.5 58.5 85
|_______ IQR _______|
= 33.5Components:
- Box: From Q1 to Q3 (contains middle 50%)
- Line inside box: Median
- Whiskers: Extend to min and max (within fences)
- Points beyond whiskers: Outliers
Interpreting Box Plots
| Feature | Interpretation |
|---|---|
| Box width | Spread of middle 50% (IQR) |
| Box position | Where the middle of data lies |
| Median line position | Symmetry within middle 50% |
| Whisker lengths | Spread of data tails |
| Points beyond | Outliers |
Skewness from box plots:
- Symmetric: Median centered, equal whiskers
- Right-skewed: Median closer to Q1, longer right whisker
- Left-skewed: Median closer to Q3, longer left whisker
Deciles and Other Quantiles
Quantiles divide data into equal parts:
| Name | Divisions | Common Percentiles |
|---|---|---|
| Quartiles | 4 parts | 25th, 50th, 75th |
| Quintiles | 5 parts | 20th, 40th, 60th, 80th |
| Deciles | 10 parts | 10th, 20th, …, 90th |
| Percentiles | 100 parts | 1st through 99th |
The 3rd decile (D3) = 30th percentile
Meaning: 30% of values fall below this point
Real-World Application
A child’s weight is at the 60th percentile for their age.
Interpretation:
- 60% of children that age weigh less
- 40% weigh more
- The child is slightly above average but within normal range
Concern thresholds:
- Below 5th percentile: May indicate underweight
- Above 95th percentile: May indicate overweight
Pediatricians track percentiles over time—the trend matters more than a single measurement.
U.S. household income percentiles (approximate):
| Percentile | Income |
|---|---|
| 25th (Q1) | $35,000 |
| 50th (Median) | $70,000 |
| 75th (Q3) | $125,000 |
| 90th | $200,000 |
| 99th | $500,000+ |
IQR = 35,000 = $90,000
The large gap between median and mean (mean ≈ $95,000) indicates right skew—high earners pull the mean up.
Summary
In this lesson, you learned:
- Percentiles indicate relative standing (percent at or below)
- Quartiles divide data into four equal parts (Q1, Q2, Q3)
- The five-number summary includes Min, Q1, Median, Q3, Max
- IQR = Q3 - Q1 measures spread of middle 50%
- Outliers are beyond 1.5 × IQR from quartiles
- Box plots visualize the five-number summary
- Percentiles enable comparison across different scales
- Median and IQR are resistant to outliers
Practice Problems
1. For the data: 5, 8, 12, 15, 18, 22, 25, 29, 33, 40
- Find Q1, Q2, Q3, and IQR
- Are there any outliers?
2. A student’s standardized test scores:
- Math: 85th percentile
- Reading: 70th percentile
- Science: 92nd percentile
In which subject did they perform best relative to other students?
3. Given the five-number summary: 10, 25, 45, 60, 95
- Calculate the IQR
- Find the outlier fences
- Would a value of 8 be an outlier? What about 100?
Click to see answers
1.
- Q1 = (8 + 12)/2 = 10
- Q2 = (18 + 22)/2 = 20
- Q3 = (29 + 33)/2 = 31
- IQR = 31 - 10 = 21
Outlier fences:
- Lower: 10 - 1.5(21) = -21.5
- Upper: 31 + 1.5(21) = 62.5
No outliers (all values between -21.5 and 62.5)
2. Science (92nd percentile) - they performed better than 92% of students, compared to 85% in Math and 70% in Reading.
3.
- IQR = 60 - 25 = 35
- Lower fence = 25 - 1.5(35) = 25 - 52.5 = -27.5
- Upper fence = 60 + 1.5(35) = 60 + 52.5 = 112.5
- 8 is NOT an outlier (8 > -27.5)
- 100 is NOT an outlier (100 < 112.5)
Next Steps
Now that you understand percentiles and quartiles:
- Z-Scores and Standardization - Another way to measure relative standing
- Data Visualization - Create box plots
- Standard Deviation Calculator - Compute IQR and more
Was this lesson helpful?
Help us improve by sharing your feedback or spreading the word.