1331.0 - Statistics - A Powerful Edge!, 1996  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 31/07/1998   
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> Information Studies >> Data - Collection

DATA COLLECTION

Individuals and organisations collect data because they or someone else require information. They may want information to keep records, make decisions about important issues, or be required to pass information on to others. So, whatever the specific reason, data is collected to provide information.

But who in society wants or needs information? The answer to this question could be simply stated as ‘many people and organisations’. Some of the groups that use statistics include:

GOVERNMENTFederal, state and local governments need information on the population and economy, among other things. This information helps them to make decisions on issues such as where to build hospitals, locate services, or how much money to raise through taxation. It also allows the public to hold a government to account by measuring its performance.
BUSINESSESMost Australian businesses require information. This information may be about the economy, the profile of a local population or various social trends. It helps them to make decisions about employing people, where to market their products and where to open new offices, warehouses and factories.
COMMUNITY GROUPSThese organisations need information about a wide variety of subjects, for example: Aboriginal health and population distribution, or the number and location of people with poor English proficiency. Sporting clubs may want information about attendances at matches or the number of young people in their local area.
INDIVIDUALSEveryone, from students to pensioners, needs information at some time. It may be needed to complete an essay, a major project or simply to satisfy one’s curiosity.

TYPES OF DATA COLLECTION

There are three main types of data collection: census, sample survey, and administrative by-product. Each has advantages and disadvantages over the other. As students you may well be required to collect data at some time. The method you choose will depend on a number of factors.


CENSUS

A census refers to data collection about everyone or everything in a group or population. So, if you collected data about the height of everyone in your class that would be regarded as a class census. There are various reasons why a census may be chosen as the method of data collection:

ADVANTAGES

Accuracy:Everyone in a group has had data collected about them, resulting in a high degree of accuracy.
Detail:Detailed information about small subgroups of the population can be made available.


DISADVANTAGES

Cost:In money terms, conducting a census can be expensive for large populations.
Speed:Time taken to do a census can be long compared to a survey.


SAMPLE SURVEY

In a sample survey, only part of the total population is approached for data. So, if you collected data about the height of 10 students in a class of 50, that would be a sample survey of the class rather than a census. Reasons to select a survey include:

ADVANTAGES

Cost:A survey costs less than a census because only part of a group has had data collected about it.
Speed:Results are obtained far more quickly than for a census. Fewer people are contacted, and less data needs processing with a survey.

DISADVANTAGES

Accuracy:Depending on sample size, results have a degree of inaccuracy.
Detail:Information on small population sub-groups or small area geography is not usually obtainable, unlike a census.


ADMINISTRATIVE BY-PRODUCT

Administrative by-product data is collected as a by-product of an organisation’s day to day operations. Examples include data on: births, deaths, marriages, divorces, airport arrivals, and motor vehicle registrations. For example, prior to a marriage license being issued, a couple must provide the registrar with information about their age, sex, birthplace, whether previously married, and where they live.

ADVANTAGES

Accuracy:Data is collected about everyone who uses that organisation’s service, resulting in a high degree of accuracy.
Time series:Data is collected on an on-going basis, allowing trend analysis.
Simplicity:Administrative data may eliminate the need to design a census/survey (and associated work), and saves the public having to complete further forms.

DISADVANTAGES

Flexibility:Data items may be limited to essential administrative information, unlike a survey.
Control:The agency which controls the data may restrict access to outsiders or charge for access.


EXAMPLE

1. Census

The ABS’s largest data collection exercise is its five yearly Census of Population and Housing. The information provided by this Census is used by all four groups referred in the beginning of this section.

Before every Census of Population and Housing, the ABS invites data users to submit topics they would like included. Not every submission for new information is successful. So how does the ABS decide which topics data is gathered on by the Census?

For the 1996 Census the ABS used the following principles:
  • Whether the topic was of major national importance. The Census is a large and costly operation imposing a burden on householders who are required to answer questions. It is essential that every question has a specific purpose.
  • Whether the topic was suitable for inclusion. Census topics should not cause an adverse reaction from people by unacceptably invading their privacy.
    They should not require an overlong explanation or instruction to ensure an accurate answer, and should not refer to things people are unlikely to remember.
  • Whether the Census was an appropriate method of collecting the data. Consideration should be given to the alternatives to a Census. In some cases the information being sought may already exist; it may be collected by another organisation, or administrative records may provide the required data.
  • Need to limit the number of questions on the form so it won’t take too long for the public to complete.
Some topics have been included in every National Census since 1911. From the table below you can see what they are. For example, Name, Age, Sex, Birthplace, Citizenship, Religion and Occupation have always been included. An asterisk in the table means the topic was covered in the Census.

However, other topics have not been covered in every census. After the presented table, some topics have been chosen to explain the reasons in more detail.


SELECTED CONTENT OF CENSUSES (a), 1911 TO 1996

    TOPIC
1911
1921
1933
1947
1954
1961
1971
1976
1981
1986
1991
1996

    Name
*
*
*
*
*
*
*
*
*
*
*
*
    Age
*
*
*
*
*
*
*
*
*
*
*
*
    Sex
*
    Orphanhood
*
    Birthplace
*
*
*
*
*
*
*
*
*
*
*
*
    Birthplace of parents
*
*
*
*
*
*
*
    Citizenship
*
*
*
*
*
*
*
*
*
*
*
*
    Disability
*
    Ethnic origin
*
    Blindness, deaf-mutism
*
*
*
    Religion
*
*
*
*
*
*
*
*
*
*
*
*
    Educational qualifications
*
*
*
*
*
*
    Holidays
*
    Income
*
*
*
*
*
*
    Occupation
*
*
*
*
*
*
*
*
*
*
*
*
    Journey to work
*
*
*
*
*
*
    Mode of travel to work
*
*
*
*
*

(a) Excludes 1966 Census.


DISABILITY

A general question on the effect of disabilities was asked in 1976. Questions relating to specific disabilities among the population were included in the 1911, 1921 and 1933 censuses. The questions asked people to indicate whether they were deaf, dumb or blind.

Information on disabilities is required by Federal and State Governments to help develop policies for disabled people. For example, where to best place services and how much money to set aside.

The 1976 Census found that information on disabilities was unreliable. This has also been the experience of Censuses in other countries, and has been confirmed in recent testing in Australia. Because of the sensitivity involved in answering a question on disability, and the difficulty in deciding what constitutes being disabled, many people under-report their situation while others inappropriately identify as disabled when this is not how others would classify them.


ETHNIC ORIGIN

A question on the ethnic origin (or ancestry) of the population has only been asked once, in 1986.

Information on ethnicity is needed to identify changing patterns of cultural diversity within the population. It is also used in community relations programs, targeting government access and equity strategies, and measuring well-being of ethnic groups.

The ABS retained all ethnicity-related 1991 Census questions for the 1996 Census, including: birthplace, citizenship, birthplace of parents, language and religion.

However, the 1986 Census question, “What is each person’s ancestry?” was not included. The ABS did not believe the additional ethnicity information the question provided was sufficient to justify inclusion in the 1996 Census.


INCOME

A question on income was first asked in the 1933 Census to measure effects of the Great Depression. It was re-included in 1976 and all subsequent Censuses. Income information helps locate the disadvantaged for social service planning.

Collecting data on income has its problems. There is a general tendency among respondents to under state income, particularly social welfare payments and interest earned on financial investments. Pensioners sometimes state they receive no income, as they do not regard their pension as income.


HOLIDAYS

The only census question on holidays was asked in 1976. At the time, there was interest to know if people on lower incomes took as much time off for holidays as those on higher incomes.

However, the ABS found it necessary to reduce the size of the Census questionnaire in following censuses. People find it burdensome when a questionnaire is large, so the question on holidays was dropped as it was considered to be of less national importance than the other retained topics.


OCCUPATION

A question on occupation has been asked in all Censuses of Population and Housing since 1911. Governments at all levels (Federal, State and Local) need detailed information on Australia’s occupation patterns. Information on occupation is vital to government policies and programs in the fields of education, training, immigration and industry.

The Census provides information about occupation patterns in small geographic areas, whereas sample surveys do not. For example, you can compare the proportion of working population who are tradespersons in, say, Brunswick, Bondi and North Adelaide.


CENSUS: A GLOBAL HISTORY

BC
3800BABYLONCarried out every six or seven years. It counted asses, oxen, butter, milk, honey, and wool.
2500
EGYPT
Carried out to assess the labour force available for building pyramids.
1491
ISRAEL
Carried out to count people liable for military service and taxation purposes.
550
CHINA
Carried out by Confucius to obtain information on the nation’s agricultural, industrial and commercial state.
AD
1719
PRUSSIAEurope’s first systematic census.
1790
USAAmerica’s first census.
1801ENGLAND
FRANCE
England’s and France’s first census.

CENSUS: AN AUSTRALIAN HISTORY

1788


‘Musters’ of convicts began. During this year they were held weekly because of the colony’s dependence on public stores and dread of famine. These ‘musters’ were also held on Sundays at the church parade. Families of convicts and free settlers were required to attend. Punishment of up to 500 lashes could be enforced for non-attendance.
1828

The colony’s first regular census was held in NSW.
1881

The colony’s first Australia-wide census was held. This was part of a simultaneous census of all British Empire colonies.
1911
The first census held under the Census and Statistics Act of 1905. This was followed by censuses in 1921, 1933, 1947, 1954, 1961 and thereafter at five yearly intervals.

EXAMPLE

2. Labour Force Survey.

The ABS conducts many sample surveys each year. One of these is the monthly Labour Force Survey, from which information is produced on Australia’s employed and unemployed persons. The survey obtains information from 65,000 persons each month, through a sample of about 29,000 private dwellings, and a further sample of non-private dwellings such as hotels, motels and caravan parks.

As with the Census of Population and Housing, many people in the Australian community need and use information from the Labour Force Survey. Governments need the information to assess whether their economic policies are changing the level of employment; and service providers need to know the areas where unemployment is highest, to be able to target assistance to people.

Sometimes there is a need to obtain more information than just employment and unemployment. Therefore, the ABS may ask additional questions in the Labour Force Survey. Topics can vary, and in the past have included job search experience of unemployed persons and number of people with more than one job.

Many important items of information are produced from the Labour Force Survey, for example: youth participation in the labour force (proportion of all teenagers aged 15-19 years who are either employed or unemployed). Information on this topic is shown below, and the figures are for August of each year.

Do you think the information below is useful in describing the general situation for teenagers? What other data should the survey collect for this purpose?



LABOUR FORCE PARTICIPATION RATES (%): 15-19 YEAR OLDS, AUGUST, AUSTRALIA

1967
1972
1977
1982
1987
1992
1997

    FEMALES
61.1
56.2
57.6
56.1
53.6
54.2
53.6
    MALES
64.7
58.5
62.1
62.4
57.4
54.0
52.9


METHODS OF DATA COLLECTION

PERSONAL INTERVIEW:

  • Face to face: involves trained interviewers visiting people to gain data. It is good for ensuring a high response rate to a sample survey or census, and trained interviewers should be able to gather accurate data. The ABS conducts face to face interviews for its Labour Force Survey during the respondent’s first month, and by telephone for subsequent months. However, it is costly to train interviewers, occasionally respondents are unavailable, and travel costs could be high.
  • Telephone: involves trained interviewers phoning people to gain data. It is quicker and cheaper than face to face interviewing. However, only people with phones can be interviewed, and the interviewed person can end the interview very easily!
  • Computer Assisted Telephone Interviewing (CATI): involves a phone interview as above, but with the interviewer keying respondent answers directly into a computer. This saves on time involved in processing data, but can be expensive to set up, and needs interviewers with computer and typing skills. The ABS uses CATI for its Retail Survey.

SELF-ENUMERATION:
  • Postal survey: a common method of conducting ABS economic surveys. It is a relatively inexpensive method of collecting data, and one can distribute large numbers of questionnaires in a short time. It provides the opportunity to reach difficult to contact people, and respondents are able to complete the questionnaire in their own time. Postal surveys do require an up-to-date list of names and addresses. Added to this is the need to keep the questionnaire simple and straightforward.
A major disadvantage of a postal survey is that it usually has a lower response rates than other data collection methods. This may lead to problems with data quality, and therefore reliability of results. People with a limited ability to read or write English may experience problems.
  • Hand-delivered questionnaire: a self-enumerated survey where questionnaires are hand-delivered to people and collected later. This method usually results in better response rates than a postal survey, and is particularly suitable when information is needed from several household members. This method is used for the ABS’s Census of Population and Housing.
  • Hand-delivered mail-back questionnaire: a combination of hand-delivered and postal methods, which reduces the cost of collecting completed forms. It gives a greater sense of privacy for respondents concerned with someone entering their home or business to collect forms.


ROLE OF BIAS

The effect of bias is to prejudice or unfairly influence data quality (see definition of bias). In practice, bias can be deliberate or unconscious. This is looked at in more detail in the section Information - Problems with Using.

For now, it is worth noting that the method of data collection itself can bias or unfairly influence data collection results. For instance, television program polls in which viewers call either a “yes” or “no” number to register their opinion are open to bias. The survey population may not be representative of the wider community, the program may bias callers before voting with one-sided information, and there is nothing to stop individuals voting many times to sway results.


ROLE OF DATA COLLECTORS

The role of data collectors is very important. The process of interviewing people to collect data involves a number of skills. Without these skills the quality obtained data can be reduced. Therefore, when someone is employed to collect data they should have, for example:
    • use of a car and telephone,
    • good communication skills,
    • a confident and professional appearance, and
    • the freedom to work evenings and weekends.

The ABS employs a large number of interviewers to collect data. Interviewers are trained before collecting data. This training emphasises that the interviewer’s opening remarks and manner in which they are made have a strong influence on a respondent’s reaction and willingness to co-operate. Because of this, data collectors should ensure certain things are carried out before they ask people questions, including:
    • give the respondent their name,
    • explain that a survey is being conducted and by whom,
    • provide identification and give the person time to read it,
    • explain that the respondent’s household or business has been selected in the survey sample, and
    • explain the survey’s purpose.

In addition, it is important that the data collector is familiar with correct interview technique. Such technique includes:
    • the ability to listen attentively,
    • keeping the interview short,
    • asking questions the same way for each respondent interviewed, and
    • NOT suggesting any answers for the respondent.


EXERCISES

1.Suggest reasons why data would be collected on the following topics:
a) Burglaries
b) Causes of death
c) Climate
d) Forests
e) Immigration
f) Schools
2.When collecting data, why is it sometimes better to conduct a sample survey than a census?
3.List some of the reasons why the ABS may decide not to include a question in the Census of Population and Housing on:
a) Health
b) Marital status
c) Sports participation
4.List some of the things you would need to consider when choosing a data collection method.
5.Given some of your answers to question 4, decide as a class which method of data collection you would employ to gather data on the following topics:
a) favourite rock or pop group in your class,
b) average height of your class,
c) time your parents spend each week doing housework, and
d) attitude of Australians toward the environment.
6.Are there some topics for which data should not be collected? For example, data on people’s health or political beliefs? Discuss as a class where you would ‘draw the line’ in deciding if issues are too sensitive to ask about.
    What factors would help you make your decision? For example, is the information you need from the data of national importance?

    Click here for answers



    Previous PageNext Page