Statistics

• Quantative


  data that is number based
• Qualitative


  data that is not number based
• Continuous


  data that can take any value
• Discrete


  data that can only take certain values
• Categorical


  data that has been classified e.g. gender
• Ordinal


  written in order or a scale
• Bivariate


  a pair of related data
• Multivariate


  data that has three or more variables
• Primary data


  collected by the person using it
• Secondary data


  has already been collected by someone else
• Population


  everything or everyone that could be involved
• Census


  survey of the whole population
• Sample


  a small number from the population
• Sampling frame


  a list of people that are going to be sampled
• Biased sample


  not representative of the population
• Bias


  favoring one thing over another
• Extraneous variables


  an unwanted variable that might affect the results
• Control group


  in the same conditions but stay untreated
• Explanatory variable


  the thing you investigate
• Response variable


  measures an outcome of a study
• raw data


  The original unorganized data collected.
• Simulation


  computer generated
• Observation


  natural experiment where you have little control over the variables.
• CPI


  is the speed of inflation
• RPI


  cost of mortgage/rent and is used to calculate the cost of living
• Interpolation


  estimating the value within a known range (more reliable)
• Extrapolation


  estimating values beyond a known range
• Double mean


  find the mean of each set of data and plot it
• Spearmans rank correlation coefficient (SRCC)


  runs on a scale between -1 and 1
• Pearson's product-moment correlation coefficient (PMCC)

  calculates the vertical distance between each point and the line of best fit.
• Random sampling method:

  1. number each member of the population
  2. select numbers using a random number generator
  3. pick the people who correspond
• Random sampling is when each member of the population has an equal chance of being picked.
• Judgement sampling is when the person choosing the sample has an element of choice which leads to bias.
• Stratified sampling means you choose a random sample that represents your population.
• Convenience sampling

  • picking what ever is easiest (e.g. the first people who arrive)
  • likely to be biased and unrepresentative
  • but its quick and easy
• Systemic sampling is when there are no particular subgroups and you choose people / things from a list at regular intervals.
• Quota sampling:

  • is used when there are constraints (e.g. picking certain ages / genders to ask rather than a fair representation)
  • ensures results are relevant but can be biased
• Estimating the median:

  1. find the group the median is in
  2. use n+1 / 2 to find the middle position
  3. calculate how far in the group it is ans / total of group x group width
  4. add onto start of category