Data collection for demographic

1
Data collection for demographic vital statistics

• (Session 01)

2
Module Objectives

• At the end of this module, you will be able to
• Explain basic concepts of routine demographic and
  epidemiological data collection procedures
• Describe and correctly use several commonly
  occurring data summary measures
• Discuss the value of various techniques for
  organising specialised studies in demography
  epidemiology
• Undertake some data summarisation and description
  tasks, and
• Understand, and look critically at, reports of
  demographic and epidemiological studies

3
Learning Objectives

• At the end of this session, you will be able to
• Recognise the broad scope of demography, and its
  dependence on basic measures
• Explain stock and flow measurement
• Critically discuss some data collection quality
  issues, and
• List some of the definitional difficulties faced
  in demographic studies

4
What is demography?

• Demography is the scientific study of human
  populations, primarily with respect to their
  size, their structure and their development (UN
  Multinational Demographic Dictionary 1958).
• Major themes birth/fertility, death/mortality, as
  well as marriage, social mobility, geographic
  distribution migration.
• Numerical summaries, but also social scientific
  theories to answer questions like, When and why
  do birth rates fall?

5
Demographic data sources

• Population census
• Measures population STOCK.
• Often decennial aims to count entire popul.n
  plus some inform.n about each member. Large
  size means study must be kept very simple many
  temporary enumerators, everyone able to answer
  the questions.
• e.g. often replaces household by people who
  slept here on 30th June
• Find and look at form for your last Census.

6
Demographic data sources

• 2. Vital registration
• Measures population FLOW/CHANGE.
• Usually continuous vital registration covers
  births, marriages and deaths. Often problems as
  to completeness.
• International migration data often collected by
  different bodies few countries can routinely
  keep track of internal population movements.
• How good is vital registration where you live?
  What encourages/discourages compliance?

7
Demographic data sources

• 3. Special studies usually surveys.
• Measure stocks or flows in more specialised
  circumstances that are inadequately covered in
  routine data collection systems
• Usually more complex questions, so need very well
  trained interviewers.
• Thus usually relatively small sample sizes.
• Sometimes hard to ascertain if individuals
  qualify as part of special population being
  sampled suggest some examples.

8
Nature of demographic data

• Age some count variables used e.g. no of live
  births a woman aged x has had.
• Very many demographic statistics use binary data
  e.g. individual died while aged x i.e. 0 or
  lived from exact age x to exact age (x 1) i.e.
  1. A binary datum contains the least amount of
  information that we can measure it is
  hydrogen atom of data collection needs a big
  sample size to provide weight of evidence.
• Suggest demographic examples of variables of each
  type.

9
Large samples

• Nationally, a census or even a demographic survey
  usually has very big sample size, so sampling
  error/standard deviation of national estimate is
  often so small it can be ignored. National
  estimates often published with no measure of
  accuracy for this reason.
• However non-sampling errors may be substantial
  and serious. Three main categories are frame
  errors, non-response errors, and measurement
  errors.

10
Frame Errors

• Example an out-of-date urban-biased sampling
  frame
• Omits a disproportionate number of relatively new
  households undocumented in-migrants homeless,
  nomadic and peripheral (e.g. forest dependent)
  people
• What might this do to estimates of
• (i) population age distribution? (ii) access to
  health services data? (iii) length-of-stay data?

11
Non-response Errors

• Example when interviewer leaves an item blank
  it could mean
• he forgot to ask this question
• the response was zero
• the respondent refused to answer
• the respondent could not understand or could
  not convey an answer that the interviewer
  understood.
• How should this be avoided?

12
Measurement Error

• Example Q. How many family members are
  disabled?
• This is weakly conceptualised who counts as a
  family member e.g. orphans taken in?
• What constitutes disabled e.g. temporary
  incapacity?
• Bad questions usually occur along with poor
  training of enumerators, and careless completion.
  All leads to useless data!
• Discuss how you would improve on above Q.

13
The term demographics

• Many surveys collect data such as individuals
  age and sex. Set of data classifying a sample
  unit referred to as the demographics.
• The term is used more widely e.g. in business
  survey basic characteristics such as turnover,
  number of employees, number of sales outlets may
  be referred to in this way.
• Often these define rows and columns in tables
  e.g. tables by age and sex.
• Suggest other variables, not very interesting in
  themselves, but needed to subdivide results.

14
Primary importance of demographics

• The demographic variables that are in use
  constantly - to define almost every output table
  - are very important. They must not be missing,
  and must be accurate.
• Example in a business survey, respondents may
  not know, or willingly tell, some size measure.
  If that is used in all tables, the rest of the
  responses will never be used in a case where the
  size is missing!
• Discuss problems with marital status Qs.

15
Difficult concepts

• Much of demography (and related social science)
  centres on the household or the family. Either
  must be very carefully defined having regard to
  local social patterns and study needs e.g.
  someone employed as daytime primary carer of
  household children may be close or distant
  relative how to count? Some temporary
  migrants would be household and family members,
  but are away include or not?
• Find out about/discuss variants used in NSS.

16
Difficult information to elicit

• Example asking 45-year-old woman her complete
  fertility history including all pregnancies,
  spontaneous and induced abortions, still- and
  live-births, multiple births e.g. twins, one
  stillborn, and survival of live-born children.
• She may need event calendar prompts to remember
  all events and dates may be emotionally
  difficult complex to record accurately.
• Think of/discuss other difficulties like this.

17
Practical work follows to ensure learning
objectives are achieved