Experiment
Repeatable process that gives a rise to a number of outcomes for an event
Event
Set of onwards or more of these outcomes
Sample space
Set of all possible outcomes
Advantages of functional form
Probability of each outcome is explicit
Variable
A collection of values
Rules for binomial distribution
° fixed number of trials
° only 2 possible outcomes
° there is a fixed probability of success
° trials are independent of eachother
Binomial v normal distribution
B - DISCRETE
N - CONTINUOUS
Population
The whole set of items that are of interest
Sample
Some sub-set of the population intended to represent the population
Sampling unit
Each individual thing in a population that can be sampled
Sampling frames
List of all of the individually named or numbered sampling units in the population
Census
Taking data from the entire population
A and D census
A: give a representative accurate result
D: expensive, time consuming, cannot use when destruction
Sample A and D
A: easy, quick, cheap
D: may bot be accurate or representative
Simple Random sample A&D
A: no bias, easy cheap, equal chance of selection
D: not suitable for large pop, need sampling frame
Systematic sampling
A: simple, quick, suitable for large population
D: sampling frame needed, can be bias if not random
Stratified sampling
A: reflects structure of pop, proportionally represents groups in population
D: population must be clearly classified into distinct groups, sampling frame needed.
Quota sampling
A: small sample still represented, no sampling frame, quick, easy, allows comparison between groups
D: bias, costly inaccurate, non-responses not recorded
Opportunity sampling
A: easy, quick, inexpensive
D: unlikely to represent sample, biased, highly dependent on individual researcher
