Collection, Organisation and Presentation of Data

The One Thing

Economic data become useful evidence only when they are collected carefully, organised systematically and presented accurately through tables, diagrams and graphs. The method selected must suit the investigation’s purpose, while reliable conclusions depend on data that are relevant, adequate, accurate and as unbiased as possible.

Who and What

• Data: Facts, figures or information collected for analysis and decision-making.
• Primary Data: Original data collected directly by an investigator for a specific purpose. Collection methods include personal interviews, telephone interviews, mailed or online questionnaires, schedules filled by enumerators, observation and experiments.
• Secondary Data: Data already collected and processed by another person, organisation or agency. Sources include government publications, official websites, international organisations, research institutions, books, journals, newspapers and internal records of organisations. Its suitability, adequacy and reliability should be assessed, including its purpose, method, unit, time period and area of collection.
• Census Method: Collection of information from every unit of the population. It provides complete coverage but is generally expensive, time-consuming and difficult when the population is very large.
• Sample Survey: Collection of information from a representative part of the population. It is usually quicker and less costly than a census, but its results depend on proper sample selection and representative coverage.
• Population or Universe: The complete group of individuals, items or observations being studied.
• Sample: A selected subset of the population used to obtain information about the whole population.
• Random Sampling: A method in which every unit of the population has an equal or known chance of selection.
• Sampling Error: Error arising because conclusions are drawn from a sample rather than the entire population. It can be reduced through proper sampling design and an adequate sample size.
• Non-Sampling Error: Error caused by inaccurate responses, poor questioning, recording mistakes or non-response.
• Questionnaire: A written set of questions used to collect information from respondents.
• Schedule: A list of questions filled in by an enumerator after asking the respondent.
• Classification: The systematic arrangement of data into groups or classes according to common characteristics. It may be qualitative, quantitative, chronological or geographical.
• Qualitative Classification: Grouping according to qualities or attributes, such as literacy or gender.
• Quantitative Classification: Grouping according to numerical characteristics, such as income or age.
• Chronological Classification: Arranging data according to time, such as years, months or quarters.
• Geographical Classification: Arranging data according to location or region.
• Frequency: The number of observations falling into a particular category or class.
• Frequency Distribution: A table showing values or class intervals and the number of observations corresponding to each.
• Class Interval: The range of values included in a class, such as 10–20 or 20–30.
• Inclusive Method: A classification method in which both class limits are included, such as 10–19 and 20–29.
• Exclusive Method: A classification method in which the upper limit of one class is excluded and included in the next class, such as 10–20 and 20–30.
• Tally Marks: Short strokes used to count observations while preparing a frequency table.
• Tabulation: Presenting classified data systematically in rows and columns.
• Components of a Table: A statistical table generally contains a table number, title, captions, stubs, body, headnote, units and, where required, a source note.
• Bar Diagram: A diagram using separate rectangular bars of equal width, where bar length or height represents the value. It is suitable for discrete or categorical data.
• Multiple Bar Diagram: A diagram using two or more bars for each category to compare related series.
• Component Bar Diagram: A bar divided into parts to show the components of a total.
• Pie Diagram: A circle divided into sectors to represent parts of a whole.
• Histogram: A graph of a continuous frequency distribution made of adjoining rectangles. Unlike ordinary bar diagrams, its rectangles touch because the data are continuous.
• Frequency Polygon: A line graph formed by joining the midpoints of the tops of histogram bars or class intervals.
• Frequency Curve: A smooth curve representing the general shape of a frequency distribution.
• Ogive: A cumulative frequency curve, either less-than or more-than, used to study cumulative totals and locate the median.
• Time Series: Data arranged according to time, such as yearly production or monthly prices.
• Line Graph: A graph in which values are plotted against time or another variable and joined by straight lines.
• Variable: A characteristic that can take different values, such as income, age or quantity demanded.
• Discrete Variable: A variable that takes separate, countable values, such as number of children.
• Continuous Variable: A variable that can take any value within a range, such as height or weight.
• Percentage: The proportion of a part relative to the total, calculated as:

Percentage = (Part / Total) × 100.

• Sector Angle: The angle used for a sector in a pie diagram, calculated as:

Sector angle = (Component value / Total value) × 360°.

• Frequency Density: The adjusted height used for a histogram with unequal class intervals, calculated as:

Frequency density = Frequency / Class width.

• Class Midpoint: The central value of a class interval, calculated as:

Class midpoint = (Lower class limit + Upper class limit) / 2.

• Cumulative Frequency: The total obtained by successively adding frequencies. Less-than cumulative frequency uses upper class limits, while more-than cumulative frequency uses lower class limits.
• Statistical Investigation: A process generally involving deciding the objective, collecting data, organising and classifying data, presenting data, analysing data and interpreting results.
• Graphing Conventions: The horizontal axis is generally the x-axis, and the vertical axis is the y-axis. Graphs and diagrams require a suitable title, clearly marked axes, a correct scale, units, a legend where needed and accurate plotting.

Causes and Consequences

• The objective, availability and required reliability of an investigation determine the collection method. Consequently, primary data may be preferred when original information for a specific purpose is needed, whereas secondary data may be used when existing information is suitable, adequate and reliable.

• Primary data can be collected directly through interviews, questionnaires, schedules, observation and experiments. This enables an investigator to design collection around a specific purpose, but the process may require greater time and resources.

• Secondary data are obtained from existing publications, websites, institutions and organisational records. Before use, the investigator must examine the source’s purpose, method, unit, time period and area of collection, because unsuitable or unreliable data can produce defective conclusions.

• A census covers every unit of the population. It therefore provides complete coverage, but it is generally expensive, time-consuming and difficult when the population is very large.

• A sample survey examines only a selected part of the population. It is quicker and less costly than a census, but its conclusions are dependable only when the sample is representative, sufficiently large, scientifically selected and appropriate to the investigation.

• Random sampling gives every unit an equal or known chance of selection. This can reduce selection bias, while sampling error remains possible because the sample does not include the entire population.

• Inaccurate responses, poor questioning, recording mistakes and non-response create non-sampling error. These problems are not eliminated merely by increasing the sample size; they require careful design and administration of the investigation.

• Raw data are difficult to interpret when they are unclassified and unorganised. Classification into qualitative, quantitative, chronological or geographical groups reduces complexity and makes comparisons clearer.

• Frequency distributions record the number of observations in categories or class intervals. Tally marks assist counting, while the choice between ungrouped, discrete and continuous distributions depends on the nature of the data.

• Inclusive and exclusive classification methods determine how class limits are treated. The inclusive method includes both limits, as in 10–19 and 20–29, whereas the exclusive method excludes the upper limit of one class and includes it in the next, as in 10–20 and 20–30.

• Tabulation presents classified data in rows and columns. A clear title, labelled rows and columns, suitable units, logical ordering and a source note where necessary allow exact numerical information to be examined efficiently.

• Different diagrams and graphs suit different kinds of data. Bar diagrams are appropriate for discrete or categorical data; histograms are appropriate for continuous grouped data; pie diagrams show parts of a whole; and multiple and component bar diagrams support comparisons or the analysis of totals.

• Bars in an ordinary bar diagram have gaps, while histogram rectangles touch. This distinction follows from the difference between separate categories and continuous data.

• Unequal class intervals can distort a histogram if raw frequencies are used as heights. Frequency density must therefore be used: Frequency density = Frequency / Class width.

• Midpoints are required when constructing a frequency polygon. They are calculated as: Class midpoint = (Lower class limit + Upper class limit) / 2.

• Cumulative addition of frequencies produces cumulative frequency. Less-than cumulative frequency uses upper class limits, and more-than cumulative frequency uses lower class limits; an ogive can then be used to study cumulative totals and locate the median.

• Pie diagrams represent components as sectors of a circle. The sector angle is calculated as (Component value / Total value) × 360°, while percentage is calculated as (Part / Total) × 100.

• Time-series data are arranged chronologically, such as yearly production or monthly prices. A time-series graph places time on the horizontal axis and corresponding values on the vertical axis, making increases, decreases, fluctuations and long-term trends easier to identify.

• Tables and visual presentations serve complementary purposes. Tables provide exact numerical information, whereas diagrams and graphs provide rapid visual understanding of trends, relationships and comparisons.

• Misleading scales or incomplete labels can distort interpretation. Honest graph construction therefore requires a suitable title, clearly marked axes, a correct scale, units, a legend where necessary and accurate plotting.

What Gets Asked

• Compare primary data and secondary data, including their sources, purposes, advantages and limitations.
• Distinguish between the census method and a sample survey, and explain why a representative sample must be sufficiently large, scientifically selected and suitable for the investigation.
• Explain sampling error and non-sampling error, including methods of reducing sampling error through proper sampling design and adequate sample size.
• Distinguish qualitative, quantitative, chronological and geographical classification, and explain the inclusive and exclusive methods of forming class intervals.
• Compare ordinary bar diagrams, multiple bar diagrams, component bar diagrams, pie diagrams, histograms, frequency polygons, frequency curves, ogives and line graphs in terms of purpose and data type.
• Calculate percentages, pie-diagram sector angles, frequency density and class midpoints, and explain the construction and interpretation of cumulative frequencies and time-series graphs.