2.6.1.2.13 Systematic stratified design pros & cons

A method applied to each stratum of a target population where sample members are selected within the stratum according to a random starting point and a fixed, periodic interval. Typically, every "nth" member is selected from the stratum for inclusion in the sample population. Stratified systematic sampling is still thought of as being random, as long as the periodic interval is determined beforehand and the starting point is random.

There may often be factors which divide up the population into sub-populations (groups / strata) and we may expect the measurement of interest to vary among the different sub-populations. Stratified systematic sampling accounts for these differences by selecting a systematic sample within each of these sub-populations.

When we sample a population with several strata, we generally require that the proportion of each stratum in the sample should be the same as in the population.

Stratified systematic sampling techniques are generally used when the population is heterogeneous, or dissimilar, or where certain homogeneous, or similar, sub-populations can be isolated (strata). Systematic sampling with no strata is most appropriate when the entire population from which the sample is taken is homogeneous. Some reasons for using systematic stratified sampling over simple random sampling are:

1. the cost per observation in the survey may be reduced;
2. estimates of the population parameters may be wanted for each sub-population;
3. increased accuracy at given cost.

Tools:

Software to implement such a design is available from R project package sp.

Pros and Cons:

The following pros and cons of systematic stratified designs should assist you in determining if it is appropriate for your monitoring needs.

Systematic Stratified Samples:

Site Selection

Pros:

• May enable survey to ensure important subpopulations are included
• Can reduce cost to obtain same sampling error compared to non-stratified sample
• Relatively simple to implement for some indicators and situations (some exceptions are large and complex sampling areas)
• May provide representative sample
• Provide spatial balance

Cons: 

• May not provide representative sample. Spatial alignment of sites may correspond to spatial alignment of attributes of interest, which will lead to biased resource representation.
• Difficult to replace dropped sites and maintain spatial balance
• May create unreasonable sampling effort if the sampling rate is set too high
• Requires additional information about population to stratify
• May not know if homogeneous subgroups exist or if homogeneous subgroups are same for all variables
• Errors in the sampling frame may result in the sampling frame excluding some sites that are in the rarget population or including some sites that are not in the target population
• Sampling frames based on different GIS scales (e.g., 1:100000 and 1:24000) may result in different estimates for the target population

Statistical Inference

Pros:

• May result in unbiased estimates and variance estimates
• May reduce sampling variance when strata constructed for that purpose
• Sampling equal numbers from strata that vary widely in physical size may be used to equate the statistical power of tests of differences between strata
• Enable standard statistical methods to be used (with assumptions)
• Precision/power depends on sample size and metric variability

Cons: 

• May not provide representative sample. Spatial alignment of sites within a stratum may correspond to spatial alignment of attributes of interest, which will lead to biased resource representation.
• There is no exact estimator for variances; there are only approximations
• May increase sampling error compared to non-stratified sample if stratification does not result in homogeneous subgroups

 Next:  Spatial Design Results and Next Steps | Go Back