Sampling and Sample Designs

Aims of Sampling

Sample Statistics accurately estimate population parameters

The Population

• Population

  • Who we are interested in

• The Sampling Unit

  • Individuals in the population

• Sampling Frame/Accessible Population

  • What’s available

    • Blank Foreign Elements

      • Not there

    • Incomplete Frames

      • Not listed

• Law of Large Numbers

  • More is better

Sample Designs

• Probability and Nonprobability Sampling

  • Nonprobability sample (Unknown population)

    • Convenience Sampling

      • Whoever is available

    • Purposive Sample

      • Only certain kinds of people

        • Snowball Sampling

    • Quota Sample

      • purposive study w/ target #

  • Probability Sampling

    • Simple Random Sampling

      • Random from population

    • Systematic Sampling

      • A list of randomly selected people by number

    • Stratified Samples

      • Random selection from subgroups

    • Cluster Sample

      • Randomly select groups from the population

Source Bias/Error

• Sample/Selection Bias

  • The sample does not represent the population being generalized too

    • Solution

      • Random Sampling

Sample Size

• Standard error

• Power

Nonsampling Errors

• Not Found

• Not-at-Homes

• Refusals

• Uninterviewable