Box Plot: A graphical representation of the five-number summary (minimum, Q1, median, Q3, maximum) that also identifies outliers.
Five-Number Summary: The minimum, first quartile (Q1), median (Q2), third quartile (Q3), and maximum of a dataset.
- Sort the data in ascending order
- Calculate the five-number summary
- Calculate IQR = Q3 - Q1
- Determine outlier boundaries using 1.5×IQR rule
- Draw the box plot with whiskers and outliers
Dataset: 12, 15, 18, 20, 22, 25, 28, 30, 35, 40 (already sorted)
Count: n = 10
Minimum: 12
Maximum: 40
Median: Average of 5th and 6th values = (22 + 25) ÷ 2 = 23.5
Q1: Median of lower half (12, 15, 18, 20, 22) = 18
Q3: Median of upper half (25, 28, 30, 35, 40) = 30
IQR = Q3 - Q1 = 30 - 18 = 12
Lower boundary: Q1 - 1.5×IQR = 18 - 1.5×12 = 18 - 18 = 0
Upper boundary: Q3 + 1.5×IQR = 30 + 1.5×12 = 30 + 18 = 48
Check each value: 12, 15, 18, 20, 22, 25, 28, 30, 35, 40
All values are between 0 and 48
No outliers detected
12
18
23.5
30
40
Five-Number Summary:
Minimum: 12
Q1: 18
Median: 23.5
Q3: 30
Maximum: 40
IQR: 12
Outliers: None
• Five-number summary: Min, Q1, Med, Q3, Max
• IQR calculation: Q3 - Q1
• Outlier detection: Values outside Q1 - 1.5×IQR and Q3 + 1.5×IQR
1.5×IQR Rule: A value is considered an outlier if it is less than Q1 - 1.5×IQR or greater than Q3 + 1.5×IQR.
Whisker boundaries: The whiskers extend to the furthest non-outlier values.
Dataset: 5, 8, 10, 12, 15, 18, 20, 25, 30, 75 (already sorted)
Count: n = 10
Minimum: 5
Maximum: 75
Median: Average of 5th and 6th values = (15 + 18) ÷ 2 = 16.5
Q1: Median of lower half (5, 8, 10, 12, 15) = 10
Q3: Median of upper half (18, 20, 25, 30, 75) = 25
IQR = Q3 - Q1 = 25 - 10 = 15
Lower boundary: Q1 - 1.5×IQR = 10 - 1.5×15 = 10 - 22.5 = -12.5
Upper boundary: Q3 + 1.5×IQR = 25 + 1.5×15 = 25 + 22.5 = 47.5
Check each value against boundaries (-12.5, 47.5):
5, 8, 10, 12, 15, 18, 20, 25, 30 are all ≤ 47.5
75 > 47.5 → 75 is an outlier
Lower whisker: Extends to minimum non-outlier = 5
Upper whisker: Extends to maximum non-outlier = 30
Outlier: 75 plotted separately
75
47.5
30
Five-Number Summary (non-outliers):
Minimum: 5
Q1: 10
Median: 16.5
Q3: 25
Maximum (non-outlier): 30
Outlier: 75
Whisker boundaries: 5 to 30
• Outlier detection: Values beyond Q1 - 1.5×IQR or Q3 + 1.5×IQR
• Whisker extension: Only to furthest non-outlier values
• Outlier plotting: Individual points beyond whiskers
Distribution shape: Can be inferred from the relative positions of quartiles and the presence of outliers.
Outlier validation: Confirm that the outlier actually exceeds the 1.5×IQR boundaries.
- Calculate IQR from Q1 and Q3
- Determine outlier boundaries
- Verify if the given outlier is valid
- Assess distribution shape based on quartile positions
- Interpret the overall pattern
IQR = Q3 - Q1 = 35 - 20 = 15
Lower boundary: Q1 - 1.5×IQR = 20 - 1.5×15 = 20 - 22.5 = -2.5
Upper boundary: Q3 + 1.5×IQR = 35 + 1.5×15 = 35 + 22.5 = 57.5
Given outlier: 100
100 > 57.5 → 100 is indeed an outlier
It exceeds the upper boundary by 42.5 units
Distance from Q1 to Median: 25 - 20 = 5
Distance from Median to Q3: 35 - 25 = 10
Since the upper half is larger, the distribution is right-skewed
Lower whisker: 20 - 10 = 10
Upper whisker: 35 - 25 = 10 (to the whisker end, not including outlier)
Whiskers are equal, but the outlier indicates extreme high value
Right-skewed distribution with high outlier
Most values are concentrated in the lower to middle range
One extreme high value pulls the maximum significantly higher
15
57.5
100
IQR: 15
Outlier validation: 100 > 57.5 → Valid outlier
Distribution shape: Right-skewed (median closer to Q1)
Interpretation: Most values are in the lower range with one extreme high value.
• Outlier validation: Compare value to 1.5×IQR boundaries
• Skewness assessment: Compare distances from median to Q1 and Q3
• Shape interpretation: Relative positions indicate distribution characteristics
Box Plot: A standardized way of displaying the distribution of data based on the five-number summary. It shows the center, spread, and skewness of the data.
Outlier: A data point that differs significantly from other observations. In box plots, outliers are defined as values beyond Q1 - 1.5×IQR or Q3 + 1.5×IQR.
Interquartile Range (IQR): The range between the first and third quartiles (Q3 - Q1), representing the middle 50% of the data.
Whiskers: Lines extending from the box to show the range of the data excluding outliers.
Five-Number Summary: The minimum, first quartile, median, third quartile, and maximum of a dataset.
- Data sorting: Arrange all data points in ascending order
- Five-number summary: Calculate Min, Q1, Med, Q3, Max
- IQR calculation: Q3 - Q1
- Outlier detection: Apply 1.5×IQR rule
- Whisker determination: Extend to furthest non-outlier values
- Plot construction: Draw box, whiskers, and outliers
• Outlier boundaries: Q1 - 1.5×IQR and Q3 + 1.5×IQR
• Whisker limits: Extend only to furthest non-outlier values
• Box representation: Spans from Q1 to Q3
• Median line: Divides box at median value
• Outlier plotting: Individual points beyond whiskers
Simultaneous outliers: A dataset can contain multiple outliers on the same or different sides of the distribution.
Dataset: 2, 5, 8, 10, 12, 15, 18, 20, 25, 30, 80, 85
Count: n = 12
Minimum: 2
Maximum: 85
Median: Average of 6th and 7th values = (15 + 18) ÷ 2 = 16.5
Q1: Median of lower half (2, 5, 8, 10, 12, 15) = (8 + 10) ÷ 2 = 9
Q3: Median of upper half (20, 25, 30, 80, 85) = 30
IQR = Q3 - Q1 = 30 - 9 = 21
Lower boundary: Q1 - 1.5×IQR = 9 - 1.5×21 = 9 - 31.5 = -22.5
Upper boundary: Q3 + 1.5×IQR = 30 + 1.5×21 = 30 + 31.5 = 61.5
Check each value against boundaries (-22.5, 61.5):
2, 5, 8, 10, 12, 15, 18, 20, 25, 30 are all ≤ 61.5
80 > 61.5 → 80 is an outlier
85 > 61.5 → 85 is an outlier
Lower whisker: Extends to minimum non-outlier = 2
Upper whisker: Extends to maximum non-outlier = 30
Outliers: 80 and 85 plotted separately
80, 85
61.5
30
Five-Number Summary (non-outliers):
Minimum: 2
Q1: 9
Median: 16.5
Q3: 30
Maximum (non-outlier): 30
Outliers: 80, 85
Whisker boundaries: 2 to 30
• Multiple outliers: Each value checked individually against boundaries
• Whisker consistency: Still extend only to furthest non-outlier values
• Separate plotting: All outliers plotted as individual points
Real-world context: Using box plots for business data analysis and decision-making.
Dataset: 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 5000
Count: n = 15
Minimum: 1200
Maximum: 5000
Median: 8th value = 1900
Q1: Median of lower half (1200 to 1600) = 1400
Q3: Median of upper half (2000 to 5000) = 2400
IQR = Q3 - Q1 = 2400 - 1400 = 1000
Lower boundary: Q1 - 1.5×IQR = 1400 - 1.5×1000 = 1400 - 1500 = -100
Upper boundary: Q3 + 1.5×IQR = 2400 + 1.5×1000 = 2400 + 1500 = 3900
Check highest value: 5000 > 3900
5000 is an outlier (exceeds upper boundary by 1100)
Lower whisker: Extends to 1200
Upper whisker: Extends to 2500 (maximum non-outlier)
Outlier: 5000 plotted separately
The $5000 sale is significantly higher than normal daily sales
Management should investigate the cause (special event, seasonal factor, etc.)
This could represent an opportunity to understand what drove the exceptional performance
1200
1400
1900
2400
2500
Five-Number Summary:
Min: 1200, Q1: 1400, Med: 1900, Q3: 2400, Max: 2500
Outlier: 5000
Business insight: The $5000 sale is an outlier and warrants investigation for potential replication strategies.
• Business context: Outliers may indicate special opportunities or problems
• Investigation trigger: Significant outliers deserve management attention
• Decision support: Statistical analysis guides business decisions
Box Plot: A graphical representation that displays the five-number summary of a dataset and identifies outliers using the 1.5×IQR rule.
Outlier: An observation that lies an abnormal distance from other values in a dataset, typically defined as values beyond Q1 - 1.5×IQR or Q3 + 1.5×IQR.
Interquartile Range (IQR): The difference between the third quartile (Q3) and the first quartile (Q1), representing the range of the middle 50% of the data.
Whiskers: Lines extending from the box to show the range of the data excluding outliers. They extend to the furthest non-outlier values.
Five-Number Summary: The minimum, first quartile, median, third quartile, and maximum of a dataset, forming the foundation of box plots.
- Data preparation: Sort the data in ascending order
- Five-number summary: Calculate minimum, Q1, median, Q3, and maximum
- IQR calculation: Compute Q3 - Q1
- Outlier detection: Apply 1.5×IQR rule to identify outliers
- Whisker determination: Set whisker boundaries to furthest non-outlier values
- Plot construction: Draw the box, whiskers, and plot outliers
• IQR formula: IQR = Q3 - Q1
• Outlier boundaries: Lower = Q1 - 1.5×IQR, Upper = Q3 + 1.5×IQR
• Whisker limits: Extend only to furthest non-outlier values
• Box representation: Box spans from Q1 to Q3
• Median line: Divides box at the median value
Box Plot: A standardized way of displaying the distribution of data based on the five-number summary. It's a visual representation that shows the center, spread, and skewness of the data while highlighting outliers.
Outlier Detection: The systematic process of identifying data points that differ significantly from other observations using the 1.5×IQR rule. This provides an objective criterion for identifying unusual values.
Five-Number Summary: A set of descriptive statistics that provides information about a dataset: minimum, first quartile, median, third quartile, and maximum. These values form the foundation of box plots.
Interquartile Range (IQR): A measure of statistical dispersion that represents the range of the middle 50% of the data, calculated as Q3 - Q1. It's robust to outliers.
- Data preparation: Organize and sort the data in ascending order
- Five-number summary: Calculate the five key statistical measures
- IQR calculation: Find the interquartile range
- Outlier detection: Apply the 1.5×IQR rule systematically
- Whisker determination: Establish boundaries for non-outlier values
- Plot construction: Create the visual representation
- Interpretation: Analyze the distribution characteristics
• Always sort data first: Proper quartile calculation requires ordered data
• Outlier detection is crucial: It affects whisker endpoints and data interpretation
• Visual comparison: Use multiple box plots to compare different groups or conditions
• Context matters: Always interpret box plots within the context of the problem being analyzed
• Systematic approach: Follow the 1.5×IQR rule consistently for objective outlier detection
Questions & Answers
Question: Why do we use 1.5 as the multiplier for IQR to detect outliers? Why not 1.0 or 2.0?
Answer: The 1.5×IQR rule was proposed by statistician John Tukey and provides a good balance:
- 1.0×IQR: Would identify too many values as outliers (too sensitive)
- 1.5×IQR: Provides reasonable sensitivity, identifying truly unusual values
- 2.0×IQR: Would miss many actual outliers (not sensitive enough)
For normally distributed data, the 1.5×IQR rule corresponds to approximately 2.7 standard deviations from the mean, which captures about 0.7% of values as outliers. This is a reasonable threshold that balances sensitivity with specificity.
Different fields sometimes use different thresholds based on their specific needs, but 1.5×IQR is the widely accepted standard in introductory statistics.
Question: Can a box plot have no outliers? What does that tell us about the data?
Answer: Yes, absolutely! A box plot with no outliers is completely normal and tells us several things:
- Consistent data: Values don't deviate dramatically from the main cluster
- No extreme values: The data distribution is relatively uniform
- Well-behaved dataset: The values stay within the expected range
Many datasets have no outliers, especially when the data naturally falls within a reasonable range. For example, test scores between 60-100 or heights of adults in a normal population often have no outliers.
The absence of outliers is not a problem - it simply means the data distribution is more consistent and predictable.
Question: How can I tell the shape of the distribution from a box plot? What patterns should I look for?
Answer: Distribution shape can be inferred from several aspects of the box plot:
- Symmetric: Median is centered in the box, whiskers are roughly equal length
- Right-skewed: Median closer to Q1, longer right whisker, box shifted left
- Left-skewed: Median closer to Q3, longer left whisker, box shifted right
Visual cues: Look at the position of the median line within the box and the relative lengths of the whiskers. If the median is not centered in the box, the distribution is skewed in the direction of the longer whisker.
Also examine the distance from Q1 to median versus median to Q3. In right-skewed distributions, the median-Q3 distance is larger than Q1-median distance.