Study Notes on Frequency Distribution
Section 2.1 Frequency Distribution
Introduction
- The raw data can be very difficult to comprehend, necessitating a meaningful presentation and summary of the data.
- This chapter will focus on tools designed to organize, display, and describe univariate data (data concerning a single variable).
Frequency Distribution
- Definition: A frequency distribution, sometimes known as a frequency table, is a tool for organizing and summarizing data, which helps in understanding the distribution characteristics of a data set.
- The basic components of a frequency distribution include:
- Frequency: Represents counts or observations; essentially the question of "How many?"
- Classes (or Categories): Categories into which data is organized in the table.
- Frequency Count: The number of observations within each class.
- Key Terminology:
- Lower/Upper Class Limits: The smallest and largest values within a class, respectively.
- Class Width: The difference between two consecutive lower class limits.
- Class Midpoints: The average of the lower and upper class limits, calculated as
(extlowerclasslimit+extupperclasslimit)/2 - Class Boundaries: A midpoint used to separate classes without gaps, calculated as
(extupperclasslimitofthefirstclass+extlowerclasslimitofthesecondclass)/2
Example 1
- The example presents a frequency distribution that records the number of women and their pulse rates. The following calculations are needed:
- (a) Number of classes
- (b) Lower and upper class limits
- (c) Class width
- (d) Class midpoints
- (e) Class boundaries
Relative (or Percentage) Frequency Distribution
- Definition: A relative frequency distribution, or percentage frequency distribution, maintains the same class limits as a frequency distribution but utilizes relative frequencies instead of actual frequencies.
- Calculation formulas:
- Relative Frequency:
extrelativefrequency=sum of all frequenciesclass frequency - Percentage Frequency:
percentage frequency=(sum of all frequenciesclass frequency)×100%
Cumulative Frequency Distribution
- Definition: A cumulative frequency distribution uses the frequency for each class that sums the frequency of that class and all previous classes.
- Instead of standard class limits, “less than” expressions are employed to describe new value ranges.
Example 2
- Table 2-2 illustrates relative (or percentage) and cumulative frequency distributions for pulse rates of females, including the following classes:
- Class Range 60-69: Frequency 12; Relative Frequency 0.3;
- Class Range 70-79: Frequency 14; Relative Frequency 0.35;
- Class Range 80-89: Frequency 11; Relative Frequency 0.275;
- Class Range 90-99: Frequency 1; Relative Frequency 0.025;
- Class Range 110-119: Frequency 0; Relative Frequency 0;
Constructing a Frequency Distribution
- Determine the Number of Classes: Choose a number between 5 and 20.
- Calculate the Class Width: It is obtained using the formula:
class width=number of classesmax value of data−min value of data
- Round the class width up to the nearest whole number.
- Choose the Starting Point: Select the minimum data value or a convenient value just below it as the first lower class limit.
- List the Lower Class Limits: Employ the first lower class limit and the class width to determine subsequent lower class limits.
- Construct Upper Class Limits: Follow the lower class limits and enter respective upper class limits.
- Tallying Data: For each data point, place a tally mark in the appropriate class, ultimately summing tally marks to get frequency counts.
Example 3
- Using a provided data set, steps to create a frequency distribution are outlined including the determination of:
- (a) The class width
- (b) Class boundaries
- (c) Class midpoint
- (d) Relative frequency (rounded to the nearest decimal)
- (e) Cumulative frequency distribution
Normal Distribution (Bell-Shaped Distribution)
- Characteristics of a normal distribution:
- Frequencies initially climb from low, reach a peak (one or two high frequencies), then taper off back to low.
- The distribution exhibits approximate symmetry, with frequencies before the peak being a mirror image of those following it.
Gaps in Frequency Distribution
- Gaps can indicate that data are sourced from two or more distinct populations; however, the absence of gaps does not confirm that different populations are not present.
- Example: A frequency distribution tracking the weight (in grams) of randomly selected pennies highlights gaps in classes, division suggesting variations in material composition based on minting years.
Comparison of Frequency Distributions
- Merging multiple relative frequency distributions into a single table streamlines data comparison.
- Example: A comparative analysis of drive-through lunch service times (in seconds) for McDonald's and Dunkin' Donuts exemplifies this method, revealing differences across time ranges for both fast-food restaurants.
TI-83 Instructions
- Data Entry: Follow the pathway STAT → 1: Edit → Input your data.
- Sort Data: Go to STAT → 2: Sort A → Specify which list you’re sorting (e.g., L1 or L2).
- Clear Data: Use STAT 4: Clr List or select the specific list needing clearance and follow through with the Clear command and Enter key.