Comparing Center and Spread Across Distributions

Master comparing distributions: mean, median, standard deviation, range, IQR with 5 detailed exercises and comprehensive rules.

Exercise Solutions 1-3
1 Comparing Centers
Exercise 1
Two datasets represent test scores from Class A and Class B:
Class A: {85, 90, 78, 92, 88, 82, 87}
Class B: {75, 80, 85, 90, 95, 80, 85}
Compare the centers using mean and median.
Key Definitions:

Mean: Average of all values = Σ(values)/n

Median: Middle value when arranged in order

Center: Measure representing the typical value in a distribution

Comparison Method:
  1. Calculate mean for each dataset
  2. Calculate median for each dataset
  3. Compare the measures of center
  4. Draw conclusions about the distributions
Step 1: Organize data and calculate means

Class A: Mean = (85 + 90 + 78 + 92 + 88 + 82 + 87) ÷ 7 = 602 ÷ 7 = 86

Class B: Mean = (75 + 80 + 85 + 90 + 95 + 80 + 85) ÷ 7 = 590 ÷ 7 = 84.3

Step 2: Find medians

Class A (sorted): {78, 82, 85, 87, 88, 90, 92} → Median = 87

Class B (sorted): {75, 80, 80, 85, 85, 90, 95} → Median = 85

Step 3: Compare centers

Class A has higher center (mean = 86, median = 87) than Class B (mean = 84.3, median = 85)

Class A: Mean = 86, Median = 87
Class B: Mean = 84.3, Median = 85
Class A has higher center values
Final Answer:

Class A has a higher center than Class B. Both mean and median are greater for Class A, indicating better overall performance.

Applied Rules:

• Mean calculation: Sum all values and divide by count

• Median calculation: Arrange in order and find middle value

• Center comparison: Higher values indicate better performance

2 Comparing Spreads
Exercise 2
Heights of basketball players (in cm) from two teams:
Team X: {180, 185, 190, 195, 200}
Team Y: {175, 180, 185, 190, 195}
Compare spreads using range and IQR.
Key Definitions:

Range: Maximum - Minimum values

IQR (Interquartile Range): Q3 - Q1

Spread: Measure of variability in the data

Spread Calculation Method:
  1. Calculate range for each dataset
  2. Find quartiles (Q1, Q3) for each dataset
  3. Calculate IQR for each dataset
  4. Compare the measures of spread
Step 1: Calculate ranges

Team X: Range = 200 - 180 = 20 cm

Team Y: Range = 195 - 175 = 20 cm

Step 2: Find quartiles for Team X

Data: {180, 185, 190, 195, 200}

Q1 (25th percentile): 182.5, Q3 (75th percentile): 197.5

IQR = 197.5 - 182.5 = 15 cm

Step 3: Find quartiles for Team Y

Data: {175, 180, 185, 190, 195}

Q1 (25th percentile): 177.5, Q3 (75th percentile): 192.5

IQR = 192.5 - 177.5 = 15 cm

Team X: Range = 20cm, IQR = 15cm
Team Y: Range = 20cm, IQR = 15cm
Both teams have identical spread measures
Final Answer:

Both teams have the same spread. The range and IQR are identical (20cm and 15cm respectively), indicating similar variability in heights.

Applied Rules:

• Range: Maximum value minus minimum value

• IQR: Difference between 75th and 25th percentiles

• Spread comparison: Larger values indicate greater variability

3 Comprehensive Comparison
Exercise 3
Daily temperatures (°F) over a week in two cities:
City A: {65, 68, 70, 72, 69, 67, 71}
City B: {60, 65, 70, 75, 80, 70, 65}
Compare both center and spread comprehensively.
Key Definitions:

Standard Deviation: Average distance from the mean

Comparative Analysis: Examining both center and spread together

Comprehensive Analysis Method:
  1. Calculate center measures (mean, median)
  2. Calculate spread measures (range, IQR, std dev)
  3. Compare both center and spread
  4. Interpret results in context
Step 1: Calculate centers

City A: Mean = (65+68+70+72+69+67+71)÷7 = 482÷7 = 68.86°F

City B: Mean = (60+65+70+75+80+70+65)÷7 = 485÷7 = 69.29°F

Step 2: Calculate spreads

City A: Range = 72-65 = 7°F, IQR = 70.5-67.5 = 3°F

City B: Range = 80-60 = 20°F, IQR = 72.5-65 = 7.5°F

Step 3: Comprehensive comparison

Centers are similar (City B slightly higher), but City B has much greater spread

City A: Mean = 68.86°F, Range = 7°F, IQR = 3°F
City B: Mean = 69.29°F, Range = 20°F, IQR = 7.5°F
City B warmer but more variable
Final Answer:

City B has a slightly higher average temperature (69.29°F vs 68.86°F) but significantly greater variability (range 20°F vs 7°F, IQR 7.5°F vs 3°F). City B experiences more temperature fluctuations.

Applied Rules:

• Center comparison: Look at mean and median differences

• Spread comparison: Compare range, IQR, and standard deviation

• Contextual interpretation: Consider practical implications

Measures of Center and Spread
Mean = \(\frac{\sum x_i}{n}\)
Measure of Center
Standard Deviation = \(\sqrt{\frac{\sum(x_i - \bar{x})^2}{n}}\)
Measure of Spread
Center Measures
Mean
Average of all values
Center Measures
Median
Middle value when ordered
Spread Measures
Range
Max - Min values
Spread Measures
IQR
Q3 - Q1
Mean Formula
\(\bar{x} = \frac{\sum x_i}{n}\)
Arithmetic average
Standard Deviation
\(\sigma = \sqrt{\frac{\sum(x_i - \bar{x})^2}{n}}\)
Average distance from mean
Quartiles
Q1, Q2, Q3
25%, 50%, 75% percentiles
When to use Mean: Symmetrical distributions without outliers
When to use Median: Skewed distributions or presence of outliers
When to use IQR: Robust measure of spread resistant to outliers
When to use Std Dev: Normal distributions with minimal outliers
Exercise Solutions 4-5
4 Outlier Impact
Exercise 4
Test scores before and after including an outlier:
Without outlier: {85, 88, 90, 92, 95}
With outlier: {85, 88, 90, 92, 95, 40}
Analyze impact on center and spread.
Key Definitions:

Outlier: Value significantly different from others

Robustness: Resistance to influence by outliers

Outlier Impact Analysis:
  1. Calculate measures without outlier
  2. Calculate measures with outlier
  3. Compare changes in center and spread
  4. Identify robust vs sensitive measures
Step 1: Without outlier calculations

Mean: (85+88+90+92+95)÷5 = 450÷5 = 90

Median: 90 (middle value)

Range: 95-85 = 10

Step 2: With outlier calculations

Mean: (85+88+90+92+95+40)÷6 = 490÷6 = 81.67

Median: (90+92)÷2 = 91 (average of middle values)

Range: 95-40 = 55

Step 3: Compare impacts

Mean decreased significantly (90→81.67), but median barely changed (90→91)

Range increased dramatically (10→55)

Without outlier: Mean = 90, Median = 90, Range = 10
With outlier: Mean = 81.67, Median = 91, Range = 55
Mean and range affected by outlier, median robust
Final Answer:

The outlier significantly affects the mean (decreases from 90 to 81.67) and range (increases from 10 to 55), but the median remains relatively stable (90 to 91). This demonstrates that the median is a more robust measure of center.

Applied Rules:

• Mean sensitivity: Highly affected by extreme values

• Median robustness: Resistant to outliers

• Range sensitivity: Always affected by outliers

5 Real-world Application
Exercise 5
Daily commute times (minutes) for two routes:
Route A: {25, 28, 30, 32, 35}
Route B: {20, 25, 30, 35, 40}
Compare and recommend based on center and spread analysis.
Key Definitions:

Reliability: Consistency indicated by low spread

Predictability: Ability to estimate outcomes

Decision-Making Process:
  1. Calculate all relevant statistics
  2. Consider both center and spread
  3. Interpret in real-world context
  4. Make recommendation with reasoning
Step 1: Calculate statistics for Route A

Mean: (25+28+30+32+35)÷5 = 150÷5 = 30 min

Median: 30 min

Range: 35-25 = 10 min

IQR: 32-28 = 4 min

Step 2: Calculate statistics for Route B

Mean: (20+25+30+35+40)÷5 = 150÷5 = 30 min

Median: 30 min

Range: 40-20 = 20 min

IQR: 35-25 = 10 min

Step 3: Compare and recommend

Both routes have same average time (30 min), but Route A has lower spread

Route A is more predictable and reliable

Route A: Mean = 30min, Range = 10min, IQR = 4min
Route B: Mean = 30min, Range = 20min, IQR = 10min
Recommend Route A for reliability
Final Answer:

Both routes have the same average commute time (30 minutes), but Route A is more reliable with less variability (Range: 10 min vs 20 min, IQR: 4 min vs 10 min). I recommend Route A because it provides more predictable travel times with less variation.

Applied Rules:

• Equal centers: Compare spread measures for decision-making

• Lower spread: Indicates greater reliability and predictability

• Real-world context: Consider practical implications of variability

Comprehensive Guide: Comparing Center and Spread Across Distributions
Center = \(\bar{x}, Median\)
Measures of Central Tendency
Spread = \(Range, IQR, \sigma\)
Measures of Variability
Key Definitions:

Center: Typical value that represents the middle of a distribution

Spread: How far apart values are from each other

Distribution: Pattern of values in a dataset

Complete Comparison Methodology:
  1. Calculate center measures: Mean and median for each distribution
  2. Calculate spread measures: Range, IQR, and standard deviation
  3. Visualize distributions: Box plots, histograms, or dot plots
  4. Compare and contrast: Note similarities and differences
  5. Interpret findings: Consider practical significance
Tip 1: Always compare both center AND spread together for complete analysis.
Tip 2: Use median when distributions are skewed or have outliers.
Tip 3: Lower spread indicates more consistent and predictable outcomes.
Tip 4: Consider context - sometimes higher spread is beneficial (e.g., diverse investment portfolio).
Common errors: Focusing only on center while ignoring spread, misinterpreting the impact of outliers, confusing measures of center with measures of spread.
Exam preparation: Practice calculating all measures, understand when to use each measure, interpret results in context.
Essential Formulas:

• Mean: \(\bar{x} = \frac{\sum x_i}{n}\)

• Range: \(Max - Min\)

• IQR: \(Q_3 - Q_1\)

• Standard Deviation: \(\sigma = \sqrt{\frac{\sum(x_i - \bar{x})^2}{n}}\)

Measure Definition When to Use Sensitivity
Mean Average of all values Symmetric distributions High to outliers
Median Middle value Skewed distributions Low to outliers
Range Max - Min Quick spread estimate Very high to outliers
IQR Q3 - Q1 Robust spread measure Low to outliers

Questions & Answers

Question: How do I decide whether to use mean or median when comparing centers of distributions? What factors should I consider?

Answer: Great question! The choice between mean and median depends on the characteristics of your distributions:

  • Use Mean when: The distribution is roughly symmetric with no significant outliers. The mean uses all data values and is sensitive to changes in the dataset.
  • Use Median when: The distribution is skewed or contains outliers. The median is resistant to extreme values and represents the middle value.

For example, if comparing household incomes in two neighborhoods where one has a few extremely wealthy residents, the median would be more appropriate since the mean would be inflated by those high values.

Always examine the shape of your distributions before deciding. If the mean and median are close, either can be used. If they differ significantly, investigate why and choose the more appropriate measure.

Question: When comparing spreads, how do I know whether to focus on range, IQR, or standard deviation? Are some more important than others?

Answer: Each measure of spread provides different insights:

  • Range: Quick and easy to calculate, shows total span of data. However, it's highly sensitive to outliers and doesn't reflect the distribution of values within the range.
  • IQR: Focuses on the middle 50% of the data, making it robust against outliers. It's excellent for comparing the core variability between distributions.
  • Standard Deviation: Measures average distance from the mean, using all data points. Most informative for normally distributed data.

In practice, I recommend using multiple measures. If you suspect outliers, prioritize IQR over range. For symmetric distributions, standard deviation provides the most complete picture. Always consider your specific context and the nature of your data.

Question: I often get confused when comparing two distributions with different centers but similar spreads, or vice versa. How do I interpret these scenarios?

Answer: These scenarios require careful interpretation:

  • Different centers, similar spreads: One distribution is shifted relative to the other, but both have similar consistency. For example, two manufacturing processes might produce parts with different target dimensions but similar precision.
  • Similar centers, different spreads: Both distributions have similar typical values, but one is more variable than the other. This indicates different levels of consistency or reliability.

Think of it this way: center tells you where the distributions are located, while spread tells you how consistent they are. Both aspects are crucial for a complete comparison. In many applications, having a favorable center is important, but low spread often indicates quality and reliability.