sampling methods

2
Sampling Method Random Sampling Systematic Sampling Stratified Sampling Random Sampling Details A sample in which each member of the population has an equal chance of being selected This is done by: Drawing lots Random Number Sampling A random starting point is selected, then the items are picked at regular intervals from the population The population is divided into non-overlapping representative groups or strata according to one or more criteria. Items are selected randomly from each stratum, where sample size being proportional to the relative size of the stratum The population is divided into non- overlapping subgroups, but sample is not random as sampler selects his sample Criteria 1. Random numbers can be generated by computer 2. Sampling frame must be complete and updated 3. Population size must be manageable 1. Starting point must be random (Use a computer to get the starting point) 2. To be used only when populatio n is likely to be homogeneo us (all other factors the same) 3. Sampling frame must be updated and complete 4. Must first have an ordered list of populatio n 1. Stratas must not overlap 2. Stratas must be exhaustiv e (mutually exclusive ) 3. Each of the strata must be treated separatel y, so sampling is easier and more accurate 4. Sampling frame must be updated and complete 1. No sampling frame is require 2. Informatio n gathered must be treated with caution as it does not represent the whole population and biasness exists Advantages -Data collected is free from bias -Sample is more evenly spread -Can obtain sufficient -Cost of sampling is low

Upload: vernon

Post on 26-Nov-2014

627 views

Category:

Documents


5 download

DESCRIPTION

H2 Mathematics Sampling Methods Notes

TRANSCRIPT

Page 1: Sampling Methods

Sampling Method

Random Sampling Systematic Sampling Stratified Sampling Random Sampling

Details A sample in which each member of the population has an equal chance of being selected

This is done by:Drawing lotsRandom Number Sampling

A random starting point is selected, then the items are picked at regular intervals from the population

The population is divided into non-overlapping representative groups or strata according to one or more criteria. Items are selected randomly from each stratum, where sample size being proportional to the relative size of the stratum

The population is divided into non-overlapping subgroups, but sample is not random as sampler selects his sample

Criteria 1. Random numbers can be generated by computer

2. Sampling frame must be complete and updated

3. Population size must be manageable

1. Starting point must be random (Use a computer to get the starting point)

2. To be used only when population is likely to be homogeneous (all other factors the same)

3. Sampling frame must be updated and complete

4. Must first have an ordered list of population

1. Stratas must not overlap

2. Stratas must be exhaustive (mutually exclusive)

3. Each of the strata must be treated separately, so sampling is easier and more accurate

4. Sampling frame must be updated and complete

1. No sampling frame is require

2. Information gathered must be treated with caution as it does not represent the whole population and biasness exists

Advantages -Data collected is free from bias

-Sample is more evenly spread out over the population-Easier to conduct than other types of sampling

-Can obtain sufficient sample points to support a separate analysis of any subgroup-Ensures a better coverage of the population as it represents the population proportionately

-Cost of sampling is low-Faster to gather information-Does not require a sampling frame

Disadvantages

-Difficult to identify every member of population, especially if population is huge-May not be able to get access to some members who have been chosen for sample

-There is a bias caused by the periodicity of the population (eg. Sampling of the use of a particular MRT station in a period of time, every 7th day selected for sampling = skewed results)-Not always possible to arrange members of population if sample size is large

-Difficult to identify appropriate strata-Time consuming as it is difficult to organise

-Not a good representative due to the lack of a sampling frame-Sample is non-random, so difficult to assess sampling error-Method is biased

Page 2: Sampling Methods