📊

Understanding Z-Scores and Percentiles

Feb 18, 2025

Lecture on Measures of Relative Standing

Introduction

  • Measures of Center & Variation: Previously discussed measures of center and variation.
    • Center: One number representing the center of a dataset.
    • Variation: One number representing the spread of the dataset.
  • Measures of Relative Standing: How individual data points compare within a dataset.
    • Compare data points from different scales.

Example of Different Scales

  • SAT Scores: Mean = 1518, Standard Deviation = 325
  • ACT Scores: Mean = 21.1, Standard Deviation = 4.8
  • Different scales make direct comparisons invalid.

Z-Score

  • Definition: Indicates how many standard deviations a data point is from the mean.
  • Formula:
    • For a sample: ( z = \frac{X - \text{mean}}{\text{standard deviation}} )
    • For a population: Use population symbols.
  • Example Calculation:
    • Dataset: [10, ..., 5 total numbers]
    • Mean = 12.4, Standard Deviation = 3.05
    • Z-score for 10: ( z = \frac{10 - 12.4}{3.05} = -0.79 )
    • Interpretation: 10 is 0.79 standard deviations below the mean.

Comparing Different Scales with Z-Scores

  • SAT vs. ACT Example:
    • SAT: Score = 1030, Z-score = -1.5
    • ACT: Score = 14, Z-score = -1.48
    • Conclusion: ACT score is better (higher z-score).

Usual vs. Unusual Values

  • Usual Values: Z-scores between -2 and 2
  • Unusual Values: Z-scores beyond ±2
  • Examples:
    • 5 might have a z-score < -2
    • 1239 might have a z-score > 2

Percentiles

  • Definition: Divides data into 100 groups.
  • Example Calculations:
    • Value 41:
      • Count values < 41: 5 out of 28
      • Percentile: ( \frac{5}{28} \times 100 = 17.86 \approx 18 )
    • Value 89:
      • Count values < 89: 27 out of 28
      • Percentile: ( \frac{27}{28} \times 100 = 96.42 \approx 97 )
  • Finding Data from Percentiles:
    • Example: 80th percentile of dataset
    • Formula: ( L = \frac{k}{100} \times n )

Quartiles

  • Definition: Divides data into four groups.
  • Quartiles:
    • Q1: 25%
    • Q2: 50% (Median)
    • Q3: 75%
  • Example:
    • Q3 = same as P75
    • Calculate using location formula

Box and Whisker Plot

  • Components:
    • Minimum, Q1, Median (Q2), Q3, Maximum
    • Visual representation on a number line
  • Example:
    • Minimum = 20, Q1 = 45, Median = 60, Q3 = 71, Maximum = 89
    • Plot illustrates data spread and center

Conclusion

  • Understanding Z-scores, percentiles, and quartiles help compare data points and analyze data distribution effectively.