Designing a Statistically Sound Sampling Plan

Designing a Statistically Sound Sampling PlanPresented by:Steven WalfishPresident, Statistical Outsourcing and ObjectiveszObjective:zDefine different types of Sampling including random, stratified and and justify your Sampling for Sampling and measurement the relationship between sample size, statistical power and statistical precisionzSampling plans for attribute PlanszDecisions are often based on our analysis of a we conduct a sample is very biaszRepresentative samplezSufficient PlanszSimple Random SamplezEach Sampling unit has an equal probability of being sampled with each perform simple random Sampling if:zEnumerate every unit of the populationzRandomly select n of the numbers and the sample consists of the units with those IDszOne way to do this is to use a random number table or random number generatorSampling PlanszStratified Random Sampling :zPopulation strata which may have a different distribution of must be known, non-overlapping and together they comprise the entire :zMeasuring Heights: Stratify on Gender Strata are Male, FemalezClinical study: stratify on stage of cancerzMeasuring Income: Stratify on education or years of experienceSampling PlanszComposite Sampling :zSample n units at randomzForm a composite of n/k units for k composite-samples; mix wellzTake the measurement on each of the k composite-sampleszFor binary outcome (positive or negative; success or failure; yes or no, etc) with rare probability of one of the tw

Sampling Plans zSimple Random Sample zEach sampling unit has an equal probability of being sampled with each selection. zCan perform simple random sampling if: zEnumerate every unit of the population zRandomly select n of the numbers and the sample consists of the units with those IDs zOne way to do this is to use a random number table or random number generator

Loading..

Tags:

  Sampling

Information

Related search queries