Sampling

• sample needs to represent the population so needs to be big enough to be reliable and they need to be randomly chosen.

• Random sample

  • assign a number to each member of the population

  • use a computer or a random number generator to select a random selection of numbers

  • pick the members corresponding to these numbers

• Biased sample

  • not representative

  • to assess if it is biased, think whether any groups of people were excluded from the sample

  • small samples are likely to be biased

Types of data

• primary data - data collected yourself

• secondary data - data from any other source

• qualitative data

  • descriptive or categorical

  • uses words rather than numbers

  • e.g. colours and names

• quantitative data

  • measured in numbers

  • represents quantities

  • e.g. heights and shoe size

  • discrete if the numbers can only take certain values (e.g. only whole numbers)

  • continuous if the numbers can take any value

• grouped data

  • data that has been sorted into categories for comparison

  • e.g. exam grades based on marks

• ungrouped data

  • data that hasn’t been sorted into categories

Sampling Techniques

Census

• sample which includes every member of the population

• true measure of population

• work well for populations that are heterogeneous and have a lot of variation

• disadvantages

  • may take a long time to collect and process

  • may be expensive to collect that much data

  • can be labour-intensive to gather data

Sample

• includes members from a proportion of the population

• easier to carry out - less data collected

• less time-intensive, less effort and cheaper than census

• work well for populations that homogeneous and have less variation

• disadvantages

  • less representative - only includes some members of the population

  • ability of a sample to effectively represent the population depends on multiple factors

Simple random sampling

• each sample of the same side has an equal chance of being selected

• can be done by numbering each member of the population, and using a random number generator to pick the members to be part of the sample

Stratified sampling

• divide the population into groups called strata

• take a proportionate simple random sample from each stratum

Quota sampling

• non-random form of stratified sampling

• population divided into groups

• sample taken from each stratum is picked by random, but by specific properties

Systematic sampling

• randomly select a starting point and take every nth piece of data from a listing of the population

• frequently chosen because it is simple to do

Opportunity/Convenience sampling

• non-random sampling

• involves collecting data that is easily accessible and requires minimal effort

  • e.g. interviewing people walking down a street

• results may be highly biased