MATHWITHCYE • Statistics and Probability

Measures of Dispersion: Grouped Data

Boundaries, midpoints, cumulative frequency tables, ogives, box plots, quartiles $Q_1/Q_3$, IQR, mean, and standard deviation.

1 7.1 Key Definitions of Grouped Data

When dealing with continuous or large datasets, data is organized into class intervals (e.g. $40 - 49$).

Class Midpoint ($x$)
$$x = \frac{\text{Lower Limit} + \text{Upper Limit}}{2}$$

Used to represent the class in mean and variance formulas.

Class Boundaries
$$\text{Lower} = \frac{L + U_{\text{prev}}}{2}, \quad \text{Upper} = \frac{U + L_{\text{next}}}{2}$$

For $40 - 49$, Lower Boundary $= 39.5$, Upper Boundary $= 49.5$.

Class Size ($c$)
$$\text{Size} = \text{Upper Boundary} - \text{Lower Boundary}$$

For $40 - 49$: $49.5 - 39.5 = 10$.

Cumulative Frequency
$$cf_k = \sum_{i=1}^k f_i$$

Sum of frequencies of that class and all preceding classes.

2 7.2 The Ogive, Quartiles & Box Plots

How to Construct an Ogive:

  • Horizontal Axis ($x$): Upper boundaries of each class (plus the lower boundary of the first class plotted at cumulative frequency $= 0$).
  • Vertical Axis ($y$): Cumulative frequency ($cf$).
  • Join the plotted points with a smooth S-shaped curve.

Reading Quartiles ($N = \sum f$):

Lower Quartile ($Q_1$): Value at $cf = \frac{1}{4} N$
Median ($Q_2$): Value at $cf = \frac{2}{4} N = \frac{1}{2} N$
Upper Quartile ($Q_3$): Value at $cf = \frac{3}{4} N$
Interquartile Range ($IQR$): $$IQR = Q_3 - Q_1$$

The 5-Number Summary & Box Plot:

A box plot visually portrays dispersion using five essential summary numbers: $\text{Minimum}$, $Q_1$, $\text{Median } Q_2$, $Q_3$, and $\text{Maximum}$.

|----- [ === Q2 === ] -----|
Min Q1 Median Q3 Max

3 7.3 Mean, Variance & Standard Deviation Formulas

Mean of Grouped Data ($\bar{x}$)

$$\bar{x} = \frac{\sum f x}{\sum f}$$

$f$ = frequency, $x$ = class midpoint.

Variance ($\sigma^2$)

$$\sigma^2 = \frac{\sum f x^2}{\sum f} - \bar{x}^2$$

Average squared deviation from the mean.

Standard Deviation ($\sigma$)

$$\sigma = \sqrt{\frac{\sum f x^2}{\sum f} - \bar{x}^2}$$

Primary measure of dispersion in the original units.