Class 11 Economics (Statistics) Chapter 2 Collection of Data Short Notes in English

Quick-revision summary of Chapter 2: Collection of Data (Statistics for Economics). For full explanations and solved examples, read the complete notes.

1 Key definitions

Key definitions
DataEconomic facts expressed in terms of numbers
VariableValues that vary (X, Y, Z); each value is an observation
Primary dataCollected by the investigator himself, for the first time, through an enquiry
Secondary dataCollected and processed by another agency; saves time and cost
SurveyA method of gathering information from individuals
Questionnaire / interview scheduleThe most common survey instrument
Closed-ended questionAnswer from given options: two-way (yes/no) or multiple choice
Open-ended questionAnswer in own words; hard to interpret and score
Pilot surveyPre-testing of the questionnaire on a small group
Census (complete enumeration)Every unit of the population is covered
PopulationThe totality of the items under study
SampleA representative part of the population from which information is obtained
Random samplingEvery unit has an equal chance (lottery method)
Non-random samplingSelection by judgement, purpose, convenience or quota
ParameterThe actual value of a population characteristic
Sampling errorParameter minus sample estimate
Sampling biasPlan such that some members can never be included

2 Nine rules of a questionnaire

  • Not too long; easy to understand; comfortable order
  • General to specific (electricity supply first, then charges)
  • Precise and clear; not ambiguous (“a lot of money” becomes ₹ classes)
  • No double negatives (“Don’t you think …”)
  • Not leading (“high-quality tea”); no hint of alternatives (“job or housewife”)

3 The three modes at a glance

Mode Main advantages Main disadvantages
Personal interview Highest response rate, clarification, reactions watched Most expensive, time-taking, may influence respondents
Mail Least expensive, remote areas, anonymity, sensitive questions Not for illiterates, low response, no clarification
Telephone Relatively low cost, quick, relatively high response Limited access (no telephone), reactions not watched

4 Census vs sample

  • Census: every unit; huge cost; no sampling error, but non-sampling errors possible
  • Sample: lower cost, shorter time, intensive enquiry, small trained team; most surveys are sample surveys
  • First task: identify the population
  • Example: all agricultural labourers of Churachandpur (Manipur) = population; ten per cent = sample
  • Lottery method: 300 households, slips, 30 drawn; the list of 300 = sampling frame
  • Non-random: 10 of 100 households chosen for convenience or acquaintance
  • Exit polls: random sample of voters leaving the booths; not always correct

5 Formulas and a solved example

Formulas

Sampling error = population parameter − sample estimate

5 farmers of Manipur: (500 + 550 + 600 + 650 + 700) ÷ 5 = 600; the sample (500, 600) has mean 550; error 600 − 550 = 50

Samples of 3 from 10: (10 × 9 × 8) ÷ (3 × 2 × 1) = 120

6 Non-sampling errors

  • More serious; difficult to reduce even with a large sample; occur even in a census
  • Data acquisition: table length, orange prices, 13 recorded for 31
  • Non-response: person cannot be contacted or refuses
  • Sampling bias: some members can never be included

7 Agencies and facts

  • National agencies: Census of India, NSS, CSO, RGI (Registrar General of India), DGCIS, Labour Bureau
  • Census: every 10 years since 1881; first after Independence 1951; last 2011
  • Population: 23.83 crore (1901), 102.87 crore (2001), 121.09 crore (2011); more than 97 crore added in 110 years
  • Annual growth rate: 2.2% (1971-81), 1.97% (1991-2001), 1.64% (2001-2011)
  • NSSO: continuous surveys in rounds; quarterly journal Sarvekshana; 60th round (January-June 2004) morbidity and healthcare; 68th (2011-12) consumer expenditure
  • Food grain (Table 2.1): 108 million tonnes in 1970-71 to 272 million tonnes in 2016-17
Common mistakes
  • Data collected by the investigator himself are primary, not secondary
  • A larger sample reduces only sampling error, not non-sampling error
  • Sample size = the units surveyed (50 of 200 farms = 50), not the units with some feature among them
  • The pilot survey comes after the questionnaire is ready and before the actual survey