yr1 chap 1 data collection

• a census observes or measures every member of a population
• a sample observes or measures a subset of a population which has been selected to represent the whole population
• advantages of a census:

  • completely accurate result
• disadvantages of a census:

  • time consuming
  • expensive
  • cannot be used if the testing process destroys the item
  • hard to process large amounts of data
• advantages of a sample:

  • less time consuming
  • less expensive
  • fewer people have to respond
  • easier to process smaller amounts of data
• disadvantages of a sample:

  • result is less accurate
  • sample may be too small to accurately represent sub-groups of the population
• the size of the sample affects the result as the larger the sample, the more accurate the result, but a larger sample requires more resources
• a sampling frame is a list of indiviually named or numbered sampling units
• random sampling means every member of the population has equal chance of being selected, this makes the sample representative and removes bias
• simple random sampling:

  requires a sampling frame, numbers are randomly generated by a calculator, computer, random number table or lottery
• the three types of random sampling are simple random, systematic, and stratified
• systematic sampling:

  requires an ordered list, items are chosen at regular intervals from a random starting place
• stratified sampling:

  requires the population to be divided into mutually exclusive strata, a random sample is taken from each stratum
• advantages of simple random sampling:

  • removes bias
  • easy and cheap for small populations and samples
  • each sampling unit has an equal chance of selection
• disadvantages of simple random sampling:

  • not suitable for large populations and samples
  • requires a sampling frame
• advantages of systematic sampling:

  • simple and quick to use
  • suitable for large populations and samples
• disadvantages of systematic sampling:

  • requires a sampling frame
  • can introduce bias if sampling frame is not random
• advantages of stratified sampling:

  • accurately reflects the population structure
  • proportional representation of groups
• disadvantages of stratified sampling:

  • requires population to be divided into strata
  • selection within strata has the same issues as simpple random sampling
• the two types of non-random sampling are quota and opportunity
• quota sampling:

  population is divided into groups based on a characteristic, the researcher meets people, assesses their group and interviews them, until all groups have been filled
• opportunity sampling:

  sample is taken from the first required number of people who are available and fit the criteria
• advantages of quota sampling:

  • small sample can be representative
  • no sampling frame required
  • quick, easy and cheap
  • different groups can be easily compared
• disadvantages of quota sampling:

  • non-random so can introduce bias
  • population must be divided into groups, requires time and money
  • non-responses are ignored instead of recorded
• advantages of opportunity sampling:

  • easy
  • inexpensive
• disadvantages of opportunity sampling:

  • unlikely to be representative
  • dependent on individual researcher
• quantitative data means data associated with numerical observations
• qualitative data means data associated with non-numerical observations
• a continuous variable is a variable which can take any value in a given range
• a discrete variable is a variable which can only take specific values in a given range
• the large data set has the following variables for UK weather stations:

  • daily mean temperature in degrees celsius
  • daily total rainfall in mm
  • daily total sunshine in hours
  • daily mean wind direction
  • daily mean wind speed in knots
  • daily maximum gust in knots
  • daily maximum relative humidity as a percentage
  • daily mean cloud cover in oktas
  • daily mean visibility in decametres
  • daily mean pressure in hectopascals
• the large data set has the following variables for non-UK weather stations:

  • daily mean temperature in degrees celsius
  • daily total rainfall in mm
  • daily mean pressure in hectopascals
  • daily mean windspeed in knots
• for daily total rainfall, sometimes the data is recorded as tr which means trace, this means there was less than 0.05mm of rainfall
• for daily maximum relative humidity, a value above 95% means misty and foggy conditions
• the weather stations in the large data set are:

  • leuchars in scotland
  • leeming in yorkshire
  • heathrow in london
  • hurn in dorset
  • camborne in cornwall
• the data in the large data set is for the months may to october of 1987, and may to october of 2015
• in the large data set, n/a means not available so the data is missing
• the variables which are not suitable for a normal distribution are:

  • wind speed as the data is qualitative
  • rainfall as the data is not symettrical