Big picture
Data are numbers that represent measurements from the real world, for example rainfall at a station, the population of a district or crop yields. A single number is a datum. Geographers use data to describe places, compare regions and find patterns: is rainfall related to crop output? Is population growing faster in cities?
Raw data must first be collected from reliable sources and then organised (compiled) so that they can be analysed and shown on graphs and maps. This chapter explains where data come from and how to compile them.
Learning goals
Explain the meaning of data and why they are needed.
Distinguish primary and secondary sources of data.
Describe methods of collecting primary data.
Describe published and unpublished secondary sources.
Tabulate and classify data, and express data as ratios and percentages.
Construct a frequency distribution with class intervals and cumulative frequency.
Core ideas
1. Need for data
Data help us understand phenomena, compare places and see changes over time.
Information without data is often vague: "Delhi is hot in summer" becomes precise when we say "the mean maximum temperature in June was 40°C".
Data must be organised and presented properly to be useful.
2. Sources of data
Type | Meaning |
|---|---|
Primary sources | Data collected for the first time by an individual or group, directly from the field |
Secondary sources | Data that have already been collected and published by others, and are used by someone else |
Methods of collecting primary data:
Method | Description |
|---|---|
Personal observation | The investigator collects information by direct observation in the field, for example by surveying the land use of a village. The observer needs theoretical knowledge and an unbiased approach; instruments like measuring tapes, compasses and cameras may be used |
Interview | The researcher gets information directly from the respondent through dialogue and conversation. A list of questions should be prepared, the purpose explained, the respondent's confidence gained, and sensitive questions avoided |
Questionnaire | A set of simple questions with answer options sent by post or email to respondents, who fill it in themselves. It covers a wide area, but works only with literate respondents and gets fewer responses |
Schedule | Similar to a questionnaire, but filled in by the enumerator (the person collecting data) while asking the respondent; useful with illiterate respondents |
Other methods | Measurement of soil and water quality using kits, data from instruments and field measurements |
Secondary sources:
Type | Examples |
|---|---|
Published sources: government publications | Census of India, reports of the National Sample Survey (NSS), Indian Meteorological Department weather reports, State Statistical Abstracts, reports of commissions |
Semi-government publications | Reports of urban development authorities and municipal corporations |
International publications | Publications of UN agencies: UNESCO, UNDP (Human Development Report), WHO, FAO; the World Population Prospects and the Demographic Year Book of the UN |
Private publications | Yearbooks, surveys and research reports by newspapers and private organisations |
Newspapers and magazines | Daily newspapers, weekly and monthly magazines |
Electronic media | Websites and online databases |
Unpublished sources | Government records and documents not published, such as revenue records kept by patwaris and records of village officers |
3. Tabulation and classification of data
Data collected from sources are usually raw: an unorganised list of values. To make sense of them, we tabulate them: arrange them in rows and columns in a table.
A table should have a title, column and row headings, the unit of measurement, and the source of the data.
Absolute data: data in their original form, as integers, also called raw data. For example, the population of each state.
Ratio and percentage: data are often expressed as ratios (such as the sex ratio, females per 1,000 males) or percentages, which make comparison easier.
Index numbers: show change in a value relative to a base year (set as 100).
Percentage example. A village has 1,200 hectares, of which 840 hectares are cultivated. Percentage of land cultivated = 840 ÷ 1,200 × 100 = 70%.
4. Grouped data and frequency distribution
When there are many observations, we group them into classes.
Class interval: a range of values, such as 40–50.
Lower limit and upper limit: the smallest and largest values of a class; class width = upper limit − lower limit.
Frequency (f): the number of observations in a class.
Frequency distribution: a table showing classes and their frequencies.
Tally marks help count observations: each observation is a stroke, and every fifth stroke crosses the previous four.
Methods of forming class intervals:
Method | Rule | Example classes |
|---|---|---|
Exclusive method | The upper limit of one class is the lower limit of the next; a value equal to the upper limit is put in the next class | 30–40, 40–50, 50–60 (a value of 40 goes in 40–50) |
Inclusive method | Both limits are included in the class; classes don't overlap | 30–39, 40–49, 50–59 (a value of 40 goes in 40–49) |
Cumulative frequency: the running total of frequencies. It shows how many observations lie below the upper limit of each class. A graph of cumulative frequency is called an ogive.
5. Worked example: a frequency distribution
The marks of 20 students are:
45, 52, 38, 61, 70, 55, 48, 66, 72, 58, 41, 63, 50, 47, 69, 54, 59, 35, 75, 60
Using the exclusive method with a class width of 10:
Marks (class) | Frequency | Cumulative frequency |
|---|---|---|
30–40 | 2 (35, 38) | 2 |
40–50 | 4 (41, 45, 47, 48) | 6 |
50–60 | 6 (50, 52, 54, 55, 58, 59) | 12 |
60–70 | 5 (60, 61, 63, 66, 69) | 17 |
70–80 | 3 (70, 72, 75) | 20 |
Total | 20 |
Notice that 50 goes in 50–60, 60 in 60–70 and 70 in 70–80 under the exclusive method. The total frequency equals the number of observations, a quick check. Twelve students scored below 60.
A reliable approach to questions
Classify each source as primary or secondary, and secondary as published or unpublished.
Choose the right primary method: questionnaire for literate respondents, schedule for illiterate ones.
When forming classes, decide on exclusive or inclusive method and apply the boundary rule.
Check that frequencies add up to the total number of observations.
Compute cumulative frequency as a running total.
Connections that matter
Data processing uses frequency distributions to calculate mean, median and mode.
Graphical representation turns tables into graphs and maps.
Mathematics: statistics uses the same methods of grouping data.
Common mistakes
Calling Census data primary data for a student. For the student using it, Census data are secondary.
Using a questionnaire with illiterate respondents. A schedule, filled in by the enumerator, is suitable.
Putting a boundary value in the wrong class in the exclusive method. It goes in the class where it is the lower limit.
Forgetting the source and units in a table.
Revision checklist
I can explain the need for data.
I can distinguish primary and secondary sources and give examples.
I can describe observation, interview, questionnaire and schedule.
I can tabulate data and express them as percentages.
I can construct a frequency distribution using exclusive and inclusive methods, with cumulative frequency.