Back to main page

Sampling and Resampling

Conceps

1. Sampling: an active process of gathering observations with the intent of estimating a population variable
2. Resampling: a methodology of economically using a data sample to improve the accuracy and quantify the uncertainty of a population parameter

Sampling types

• Bernoulli samples: independent Bernoulli trials on population elements, trials decide whether the element becomes part of the sample
• Cluster samples: divide population into groups, then a random sample is chosen from the cluseters
• Systematic sampling: select elements form an ordered frame
• SRS: select items completely rendomly, so each element has the same probability of being chosen as any other element
• Strtified sampling: divide population into homogneous subgroups before getting sample, and sample each subpopulation independently
• etc.

Resampling types

• Bootstrap: sampling with replacement from the original data
• Jackknife: systematically leaving out each observation from a dataset and calculating the estimate and then find the average of these calculations

References:

1. A gentle introduction to statistical sampling and resampling
2. Statisticshowto: Sampling in statistics: different sampling methods, types & error
3. Wikipedia: Sampling(statistics)
4. Wikipedia: Resampling(statistics)
5. Wikipedia: Bootstrapping(statistics)
6. Wikipedia: Jackknife resampling