Brian schnick. BASIC CONCEPTS IN SAMPLING Advantages of Sampling Sampling Error Sampling Procedure.

CH 6: BASIC ELEMENTS OF

SAMPLING

brian schnick

BASIC CONCEPTS IN SAMPLING

Advantages of Sampling Sampling Error Sampling Procedure

Advantages of Sampling Sampling is a necessity in many geographic

research problemsPopulation may be simply too big for 100% contact

Sampling is an efficient and cost-effective method of collecting informationLess time, cost, personnel, etc. required vs.

collecting data on the entire population Sampling can provide highly detailed

informationIn-depth analysis of the few vs. cursory analysis of

the many

Advantages of Samplingcontinued

Sampling allows repeated collection of information quickly and inexpensivelyPhenomena under study may

frequently/rapidly change, requiring updates Sampling can provide a high degree of

accuracyFewer data points allow for detailed study

vs. superficial look at many points

Sampling Error

Randomness must be incorporated to represent population accuratelyBut samples are never 100% accurate

Methodology may introduce systematic errorSampling elements other than the elements

of interest

Sampling Errorcontinued

Improper sampling designWhat method produces the required level of

accuracy? Operational, logistic, personnel

problemsInconsistent collection practices, errors in

recording data

Sampling Procedure

Step 1: Conceptually define target population and target areaTarget Population: the complete set of

individuals from which information is to be collected

Target Area: the entire region or set of locations from which information is to be collected

Sampling Procedurecontinued

Step 2: Designate sample population and sampled area from sampling frameSampling frame: practical structure that

contains the entire set of elements from which the sample will actually be drawn

Sampled population: set of all individuals contained in the sampling frame, from which the sample is drawn

Sampled area: set of all locations within the study area from which the sample is drawn

Sampling Procedurecontinued

Step 3: Select Sampling DesignProbability sampling preferred over

non-probability samplingTypes of probability samples include

random, systematic, stratified, cluster ,and hybrid

Spatial and non-spatial variations in sampling design exist

Sampling Procedurecontinued

Step 4: Design research instrument and operational planDirect observation, field measurement,

questionnaires, personal & telephone interviews

Establish protocols for handling anticipated problems

Logistic & procedural tasks completed before sampling

Sampling Procedurecontinued

Step 5: Conduct pretestTrial run/pilot survey of sample collection

methodCorrect discovered problemsResults may indicate sample size

Sampling Procedurecontinued

Step 6: Collect sample dataConsistency in collection methods and

procedures is essentialEnsure high level of quality control

TYPES OF PROBABILITY SAMPLING Simple Random Systematic Stratified Cluster Hybrid

Simple Random Sampling

All individuals in the sampling frame have an equal chance at selection

All individuals ordered sequentially and chosen by random number generator

Systematic Sampling

All individuals ordered sequentially and chosen by fixed interval

Stratified Sampling

Population divided into subgroups/areas then sampled

Stratified Sampling can beProportional: population ratios maintained in

subgroupsDisproportional: population rations not

maintained in subgroups

20% of a city live in apartments – do the subgroups sampled have 20% in apartments?

Cluster Sampling

Population divided into subgroups/areas Some subgroups/areas selected 100% of selected subgroups/areas

sampled

Hybrid Sampling

Combination of sampling designsExample: cluster of individuals sampled on a

systematic time frequency

SPATIAL SAMPLING

Types Point sampling in detail

Types of Spatial Sampling

Line Sampling

Area Sampling

Point Sampling

Point Sampling

Simple random point sampling

Systematic point sampling

Point Samplingcontinued

Stratified point sampling

Cluster point sampling

Point Samplingcontinued

HybridsStratified systematic unaligned point sample

(most widely used)

Disproportional stratified systematic aligned point sample

Point Samplingcontinued

HybridsCluster systematic point sample

Disproportional stratified cluster point sample

Additional Material

Experiment-Resources.com US Census Wiki YouTube

Credits

Outline and scanned graphics from An Introduction to Statistical Problem Solving in Geography - 2d Edition, McGrew & Monroe, Waveland Press, Inc., 2009