Revision notes

Unit 13 Statistical Techniques: 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

Statistics in geography: types of data and how they are collected

The four processes of statistics

  1. Collection Gather the facts, for example rainfall records of a region.
  2. Presentation Arrange them in tables, graphs or diagrams.
  3. Analysis Calculate measures such as the range and the average.
  4. Interpretation Explain what the results mean for the place being studied.

Two ways of classifying data

  • By nature
    • Qualitative: describes an item, such as tree type or language spoken
    • Quantitative: has a numerical value, such as marks, rainfall in mm
  • By source
    • Primary: collected first-hand by the investigator for the study
    • Secondary: already collected and recorded by others, published or unpublished
Methods of data collection
Primary dataSecondary data
Direct personal investigation: the investigator interviews people face to facePublished sources: census reports, government statistical handbooks, yearbooks, journals
Indirect oral investigation: information is gathered from other people who know about the subject (witnesses, local officials)Unpublished sources: records kept by offices, institutions or researchers but not printed
Questionnaire method: a printed list of questions is filled in by the respondents

Qualitative = describes; quantitative = numbers. Primary = collected first-hand (interview, indirect oral, questionnaire); secondary = taken from published or unpublished records.

Key takeaways: statistics deals with quantitative data through collection, presentation, analysis and interpretation. Geographers use it to describe, generalise, estimate probability, infer from samples, compare places and test patterns. Data are qualitative or quantitative, and primary or secondary. Primary data come from direct personal investigation, indirect oral investigation or questionnaires; secondary data come from published or unpublished sources.

Read the full lesson

Lesson 2

Organising data: range, array and frequency distribution

Terms used, with the class 10-20 as an example
TermMeaningFor 10-20
ClassA group into which the variable is divided10-20
Class limitsThe extreme values of a class: lower limit and upper limitLower 10, upper 20
Midpoint(Lower limit + upper limit) ÷ 2(10 + 20) ÷ 2 = 15
Class intervalUpper limit - lower limit20 - 10 = 10
Tally marksA stroke for each observation; every fifth stroke is drawn across the previous four|||| for 4, a crossed group for 5
Class frequencyNumber of observations in the classThe count of tallies in 10-20
Total frequencyThe sum of all class frequencies, equal to the number of observationsN
Frequency distribution of the made-up marks
ClassMidpointTally (count)Frequency
0-1058, 62
10-201512, 19, 15, 174
20-302527, 22, 26, 24, 29, 21, 27, 25, 28, 2310
30-403534, 31, 38, 33, 36, 30, 39, 32, 35, 3710
40-504545, 41, 48, 444
Total30

Range = highest - lowest. Midpoint = (lower + upper) ÷ 2. Class interval = upper - lower. Total frequency = sum of class frequencies = number of observations.

From raw data to a frequency table

  1. Raw data Figures in the order collected.
  2. Range Highest - lowest, to decide classes.
  3. Classes Equal intervals that cover the range without overlap.
  4. Tally One stroke per observation, in groups of five.
  5. Frequencies Count, then check the total.

Key takeaways: organising data reduces raw figures to a usable form. Range = highest - lowest; an array is data in ascending or descending order. A frequency distribution groups data into classes with lower and upper limits, a midpoint and a class interval, counted with tally marks into class frequencies whose sum is the total frequency.

Read the full lesson

Lesson 3

The arithmetic mean for ungrouped and grouped data

Mean X̄ = ΣX ÷ N, where ΣX is the sum of the values and N the number of items.

Made-up monthly rainfall of a coastal station (mm)
JanFebMarAprMayJunJulAugSepOctNovDec
81215306248709514021018096
Made-up data: heights of 40 saplings in a nursery
Height (cm)fMidpoint xfx
0-104520
10-20815120
20-301425350
30-401035350
40-50445180
TotalN = 40Σfx = 1,020

X̄ = A + (Σfd ÷ N) × i, where A = assumed mean, d = (x - A) ÷ i, i = class size, N = total frequency.

Step-deviation working, A = 25, i = 10
Classfxd = (x - 25) ÷ 10fd
0-1045-2-8
10-20815-1-8
20-30142500
30-401035110
40-5044528
Total40Σfd = 2

Two methods for grouped data

  • Direct method
    • Multiply each midpoint by its frequency
    • X̄ = Σfx ÷ N
    • Simple but large products
  • Step-deviation method
    • Choose A and find d = (x - A) ÷ i
    • X̄ = A + (Σfd ÷ N) × i
    • Small numbers, fewer slips

Key takeaways: an average is a single value representing the data and lies between the extremes, so it is a measure of central tendency; the types are mean, median and mode. Mean = ΣX ÷ N. For grouped data, use midpoints: X̄ = Σfx ÷ N, or the step-deviation shortcut X̄ = A + (Σfd ÷ N) × i, which gives the same answer.

Read the full lesson

Lesson 4

The median for ungrouped and grouped data

Median M = value of the ((N + 1) ÷ 2)th item of the arranged series. If this falls between two items, take the mean of those two items.

Median = l + ((N ÷ 2 - m) ÷ f) × c, where l = lower limit of the median class, N = total frequency, m = cumulative frequency of the class before the median class, f = frequency of the median class, c = class interval.

Cumulative frequency for the made-up sapling heights
Height (cm)fcf
0-1044
10-20812
20-301426
30-401036
40-50440

Steps for the median of grouped data

  1. Cumulative frequency Add a running total column.
  2. N ÷ 2 Find half the total frequency.
  3. Median class First class whose cf reaches N ÷ 2.
  4. Values Read l, m, f and c.
  5. Formula Median = l + ((N ÷ 2 - m) ÷ f) × c, and check it lies within the class.

Key takeaways: the median is the middle value of an arranged series and splits it into two equal halves. For ungrouped data it is the ((N + 1) ÷ 2)th item, or the mean of the two middle items when N is even. For grouped data, find the median class from the cumulative frequency at N ÷ 2 and use Median = l + ((N ÷ 2 - m) ÷ f) × c. The median is not pulled by extreme values as the mean is.

Read the full lesson

Lesson 5

The mode, and choosing between mean, median and mode

Made-up data: percentage of days with wind from each direction
DirectionNNEESESSWWNWCalm
Days (%)3129652815814

Mode = l + ((f₁ - f₀) ÷ (2f₁ - f₀ - f₂)) × c, where l = lower limit of the modal class, f₁ = its frequency, f₀ = frequency of the class before it, f₂ = frequency of the class after it, c = class interval.

Mean, median and mode compared
AverageStrengthWeaknessGood for
MeanUses every valuePulled by extreme valuesTemperature, data without extremes
MedianNot affected by extremesIgnores the size of other valuesIncomes, rainfall with very wet months
ModeShows the most common value; works for categoriesMay not exist or may not be uniquePrevailing wind, most common crop, shoe size

The three results for the made-up saplings

  • Mean
    • 25.5 cm
    • From Σfx ÷ N
  • Median
    • 25.71 cm
    • From the cumulative frequency
  • Mode
    • 26 cm
    • From the modal class

Key takeaways: the mode is the most frequent value; for qualitative data it is the most frequent category, such as the prevailing wind direction. For grouped data, the modal class has the highest frequency and Mode = l + ((f₁ - f₀) ÷ (2f₁ - f₀ - f₂)) × c. The mean uses all values but is pulled by extremes, the median resists extremes, and the mode shows what is most common.

Read the full lesson

Privacy · Terms · Refunds · Delivery