Binomial conditions<span>fixed number of trials each trial must be either a success or a fail trails must be independent the value of P must remain constant</span>Binomial E(X)npBinomial Var(X)np(1-p)Poisson conditions<span>events must be independent events must occur singly in space or time event must occur at a constant rate</span>poisson E(X)£Var(X)£binomial to normal<span>n is large p is close to 0.5 N(np,np(1-p) np>10</span>binomial to poisson<span>n is large p is small Po(np) np<10</span>poisson to normal<span>n is large N(£,£)</span>populationa collection of itemscensusinformation obtained from every member of a populationsamplea selection of indvidual members from a populationpopulation parameterany characteristic of a population which is measurablefinite populationa population in whihc every individual member can be given a numberinfinite populationa population which is impossible to give a number to every individualadvantage censusevery single member of a population is used, unbiased, gives an accurate answerdisadvantage censustime consuming, costly, difficult to ensure that the whole population is surveyeddisadvantage sample<span>natural variation bias</span>advantage sample<span>sample is representative cheaper data more readily avalible</span>poisson<span>events occur randomly singly in space or time independently of each other constant rate</span>binomial<span>fixed number of trials each trail either a success or failure trails independent probability of success constant</span>significance levelprobability of incorrectly rejecting the null hypothesisstatisticrandom variable quantity calculated soley from observations in a sample does not involve any unknown parameters numerical property of a samplesampling distributionall possible values of a test statistic and their probabilitiessampling framea list of all the sampling units within a populationsampling unitsthe individual units of a populationsample surveyan investigation using a samplerandom samplingevery possible sample of size n has an equal chance of being selectedhypothesisa statement made about the value of a population parameternull hypothesishypothesis that is assumed to be correcttest statistica form of a statistic in which the evidence from a sample in a hypothesis test is summarisedcritical valuesthe values on the boundaries
