Revision notes

Chapter 3 Organisation of Data: Foundation

The key points, rules to remember, tables and diagrams from all 5 lessons, on one page for quick revision. To keep a copy, use your browser's Print and choose "Save as PDF".

Lesson 1

Raw data and the kinds of classification

To classify is to sort items into groups (classes) by a chosen rule, so that alike items sit together and are easy to find, compare and draw conclusions from.

Kinds of classification of data

  • Chronological By time: years, quarters, months, weeks
    • A series of values over time is a time series
  • Spatial By place: countries, states, districts, cities
  • Qualitative By attributes that cannot be measured
    • Presence or absence of a quality, e.g. gender, literacy, marital status
  • Quantitative By measurable characteristics grouped into classes
    • Height, weight, age, income, marks
Illustrative data: a time series
YearSales (₹ lakh)
Year 140
Year 246
Year 344
Year 455
Year 560
Illustrative two-stage classification of 200 workers
Stage 1: genderStage 2: marriedStage 2: unmarriedTotal
Male8040120
Female503080
Total13070200

Key takeaways: raw data are disorganised and hard to use; classification brings order. Data can be classified by time (chronological, giving a time series), place (spatial), attribute (qualitative, by presence or absence of a quality) or measurable size (quantitative). The purpose of the study decides which classification to use.

Read the full lesson

Lesson 2

Continuous and discrete variables; parts of a frequency distribution

Continuous versus discrete variables

  • Continuous
    • Can take any value in a range
    • Values can be broken into infinite gradations
    • Examples: height, weight, time, distance, temperature, rainfall, crop yield, age
  • Discrete
    • Takes only certain values
    • Changes by jumps
    • Examples: number of students, number on a dice, number of cars on a road, population (a count of persons)

A frequency distribution sorts the raw values of a measurable variable into classes and shows, against each class, how many values fall in it (its frequency).

The vocabulary of a frequency distribution
TermMeaningFor the class 20-40
Class frequencyNumber of observations falling in the classCounted from the data
Lower class limitLowest value of the class20
Upper class limitHighest value of the class40
Class interval (class width)Upper class limit minus lower class limit40 − 20 = 20
Class mark (mid-point)(Upper class limit + lower class limit) ÷ 2(40 + 20) ÷ 2 = 30
Illustrative data
Wage (₹)Number of workersClass intervalClass mark
200-3004100250
300-40010100350
400-50016100450
500-60012100550
600-7008100650
Total50
Wage (₹)FrequencyRelative frequency (%)
200-30048
300-4001020
400-5001632
500-6001224
600-700816
Total50100
VariableTypeReason
Area of a plotContinuousCan be any value, such as 312.75 square metres
Number appearing on a diceDiscreteOnly 1, 2, 3, 4, 5 or 6
Rainfall in a dayContinuousAny value, such as 12.6 mm
Number of cars passing a crossingDiscreteWhole numbers only
TemperatureContinuousAny value within a range

Key takeaways: continuous variables can take any value in a range; discrete variables jump between particular values. A frequency distribution groups a quantitative variable into classes with their frequencies. Class interval = upper limit − lower limit; class mark = (upper limit + lower limit) ÷ 2, and the class mark stands for the class in later calculations. Frequencies as a share of the total are relative frequencies.

Read the full lesson

Lesson 3

Forming classes: size, number, limits and the inclusive and exclusive methods

Number of classes = range ÷ size of class interval. Range = largest value − smallest value. Decide one and the other follows.

Two ways of forming class intervals

  • Exclusive method
    • Classes like 0-10, 10-20, 20-30
    • Upper limit of one class = lower limit of the next
    • A value equal to a limit is excluded from one of the two classes
    • Decide in advance: usually 'upper limit excluded', so 10 goes in 10-20
  • Inclusive method
    • Classes like 0-10, 11-20, 21-30
    • Both limits belong to the class
    • Leaves a gap between classes (10 and 11)
    • Suits discrete data such as whole-number marks

Making inclusive classes continuous

  1. Find the gap Lower limit of the second class minus upper limit of the first
  2. Halve it Gap ÷ 2
  3. Lower limits Subtract half the gap from every lower limit
  4. Upper limits Add half the gap to every upper limit
Illustrative data: adjusted classes
Original class (₹)Adjusted class (₹)FrequencyAdjusted class mark (₹)
300-349299.5-349.58324.5
350-399349.5-399.515374.5
400-449399.5-449.522424.5
450-499449.5-499.511474.5
500-549499.5-549.54524.5
Total60

Key takeaways: use equal class intervals unless the range is huge or values crowd into a small part of it. Usually six to fifteen classes; number of classes = range ÷ class size. Limits should be definite, with observations concentrated near each class mark. The exclusive method needs a rule for boundary values; the inclusive method includes both limits, and for continuous data its gaps are removed by subtracting and adding half the gap.

Read the full lesson

Lesson 4

Tally marks, loss of information and unequal classes

Illustrative data: tally marking
Class (marks)ObservationsTallyFrequencyClass mark
0-108, 6||25
10-2012, 19, 15, 10, 18||||/515
20-3027, 22, 25, 28, 21, 26, 24, 29, 20, 27, 23||||/ ||||/ |1125
30-4035, 30, 33, 36, 39, 31, 38, 34, 32||||/ ||||935
40-5041, 47, 44|||345
Total30

Grouping makes data concise but causes loss of information: the frequency tells how many observations a class has, not what they are, and all of them are represented by the class mark.

From raw data to a frequency distribution

  1. Raw data Observations in no order
  2. Decide classes Equal or unequal; number, size and limits
  3. Tally One stroke per observation, in groups of five
  4. Count frequencies Check that they add to the total
  5. Class marks Represent each class in later calculations

Key takeaways: class frequencies are counted with tally marks grouped in fives, and the frequencies must add up to the number of observations. Grouping causes loss of information because every value in a class is treated as the class mark. Where observations crowd into a few classes, narrower (unequal) classes give class marks that represent the data better.

Read the full lesson

Lesson 5

Frequency arrays and bivariate frequency distributions

A frequency array is the classification of data of a discrete variable: each value the variable takes is listed with its frequency.

Illustrative frequency array
Number of childrenNumber of households
01
13
210
36
43
51
61
Total25

Univariate and bivariate

  • Univariate distribution
    • One variable
    • One column of classes and one of frequencies
    • Example: marks of 30 students
  • Bivariate distribution
    • Two variables for each unit
    • Classes of one variable in rows, of the other in columns
    • Each cell: number of units in that row class and that column class
    • Used later to study correlation
Illustrative bivariate frequency distribution of 12 shops
Advertisement (₹ thousand)Sales 50-60 (₹ lakh)Sales 60-70 (₹ lakh)Sales 70-80 (₹ lakh)Total
10-203104
20-300314
30-400134
Total35412

Steps in making a bivariate table

  1. Collect pairs Two values from each unit
  2. Classes for each variable One set for rows, one for columns
  3. Place each pair in a cell Row class of the first value, column class of the second
  4. Count cell frequencies Tally if there are many pairs
  5. Total rows and columns Both must add to the number of units

Key takeaways: a frequency array lists each value of a discrete variable with its frequency. A univariate distribution covers one variable; a bivariate distribution covers two, with one in rows and the other in columns, each cell giving the number of units in both classes. Row and column totals both equal the number of units, and the pattern of cells hints at a relationship between the variables.

Read the full lesson

Privacy · Terms · Refunds · Delivery