Stats Flashcards
Define opportunity sampling
A sample is taken from the first n members of the population who are available and fit the criteria.
Define a simple random sample.
A simple random sample of size n is one where every sample of size n has an equal chance of being selected
Define a census.
When you collect information from every member of a population.
Advantages of a census.
- you get accurate information about pop, as every member has been surveyed
- true representation of population
- unbiased
Disadvantages of a census.
- takes lots of time and effort for large pops
- expensive
- can be difficult to make sure all members are surveyed
- if tested items used up or damaged in some way, census is impractical
Define sample units.
Individual members of the population
Define sampling frame.
A full list of all the sampling units. Used to represent population when selecting random sample.
Advantages of sampling.
- quicker and cheaper than census, and easier to get hold of required information
- only option when surveyed items are used up or damaged
Disadvantages of sampling.
- variability between samples, each possible sample will give different results, so could just happen to select one which doesn’t accurately reflect pop
- samples can easily be affected by sampling bias.
Advantage of simple random sampling.
- every member has equal chance of being selected, unbiased
Disadvantage of simple random sampling.
- can be inconvenient if pop spread over large area
Define systematic sampling.
Using a sampling frame, every nth member of the pop is sampled
Advantage of systematic sampling.
- used for quality control, should give unbiased sample.
Disadvantage of systematic sampling.
- regular interval could coincide with a pattern, which would make sample biased
Advantage of opportunity sampling.
- data can be gathered quickly and easily.
Disadvantage of opportunity sampling.
- isn’t random
- can be very biased, no attempt to obtain a representative sample.
Define stratified sampling.
- divide pop into categories
- calculate total pop
- calculate number needed for each category: size of category in pop/ total size of pop x total sample size
- select sample for each category at random.
Advantage of stratified sampling.
- guarantees proportional representation
Disadvantage of stratified sample
- not useful when aren’t any obvious categories
- can be expensive
Define quota sampling.
- divide pop into categories
- give each category a quota ( number of members to sample)
- collect data until quotas are met in all categories
Advantage of quota sampling.
- can be done when isn’t full list of pop
- every sample member responds as interviewer continues until all quotas are met
Disadvantage of quota sampling.
- easily biased by interviewer
Define cluster sampling.
- divide pop into clusters covering whole pop, where no member of pop belongs to multiple clusters
- randomly select clusters to use in sample
- either use all members of selected clusters or randomly sample within each cluster to form sample
Advantage of cluster sampling.
- more practical
- can incorporate other sampling methods at either stage, so quite adaptable
Disadvantage of cluster sampling.
- results less representative of pop as whole as only sample certain clusters
- not always possible to separate pop into clusters in natural way
Define null hypothesis.
Statement about the value of population parameter.
Define alternative hypothesis.
A statement that describes the value of the population parameter if null hypothesis is rejected.
Define test statistic.
A statistic calculated from sample data which is used to decide whether or not to reject the null hypothesis.
Define critical region.
The set of all values of the test statistic that would cause you to reject the null hypothesis.
How to calculate variance
Sum of (x- mean) ^2 / n
Equation for probability of an event.
No. Of total outcomes where event happens/ total no. Of possible outcomes.
What does p(A U B) =
P(A) + P(B) - P(A n B)
Equation for P(A n B)
P(A) x P(B|A)
What is true for independent events?
P(A n B) = P(A) x P(B)
What a five conditions of binomial distribution?
- fixed number of trials
- trial results in either success or failure
- all trials independent
- probability of success is same in each trial
- variable is total number of successes in n trials
What does Z equal
X- mean/ variance
What are conditions for normal distribution
- data is continuous
- data is roughly symmetrically distributed with peak in middle
- data tails off either side of mean
