Statistics for Economics: Collection, Organisation and Presentation of Data
This domain tests the data-handling chain: where data come from, how they are organised into frequency distributions, and which diagram suits which data type. Definitions and small conversion exercises carry the marks.
Core rules
- Primary data are collected first-hand; secondary data are already collected by someone else and must be checked for reliability, units and coverage.
- Methods of primary collection: direct personal investigation, indirect oral interviews, information from local sources, mailed questionnaires, and schedules filled by enumerators.
- Census covers every unit; sampling covers a part. Random sampling (lottery, random number tables) reduces bias. Sampling error arises from studying a sample; non-sampling errors arise from faulty design, recording or processing.
- A good questionnaire is short, unambiguous, free of leading questions and pre-tested.
- Frequency distribution: exclusive classes (10–20, 20–30) versus inclusive classes (10–19, 20–29). Class mark = midpoint; class width = upper − lower limit.
- Presentation: bar diagrams for discrete categories, pie charts for shares, histograms and frequency polygons for continuous distributions, ogives for cumulative frequency, arithmetic line graphs for time series.
Worked example
Convert the inclusive classes 10–19, 20–29, 30–39 to exclusive form. The gap is 20 − 19 = 1, so adjust by 0.5: 9.5–19.5, 19.5–29.5, 29.5–39.5. The class mark of 19.5–29.5 is (19.5 + 29.5)/2 = 24.5, and its width is 10. For an unequal class 20–40 with frequency 30, frequency density = 30/20 = 1.5, which is what the histogram plots as height.
Traps that cost marks
- Drawing a histogram from inclusive classes without converting them.
- Using a bar diagram for continuous data or a histogram for discrete categories.
- Computing frequency density as frequency divided by number of classes.
- Accepting secondary data without checking the source and definitions.