Sampling in Epidemiology Quiz

Sampling in Epidemiology

• Why study only a portion of the population?
  • To avoid studying the entire population
  • To reduce bias in results
  • To save time and money
  • Measurements may be more accurate in a sample than the population
  • To ensure feasibility in conducting research

Sample Selection

• Sampling is the process of selecting a sample of participants with the right characteristics and of appropriate size.
• A sample should be representative of the population to maintain validity.
  • Consider factors like age, sex, location, and time to represent the target population
  • Representativeness is essential to be able to make generalizations from the sample to the population
• The process of selecting a sample should ensure representativeness and this should be confirmed after data collection
• The process of selecting a sample consists of:
  • Start with the target population
  • Select a sampling population
  • And finally, the sample

Definitions

• Population: Groups of people (or things) sharing one or more characteristics
• Sample: A representative subgroup of a larger population
  • Used to estimate population characteristics
  • Must closely resemble the population in the characteristic under study
• Sampling Frame: A complete list of sampling units in the target population.
  • Must be accurate and complete for successful study
• Sampling Unit: The individual units (people or things) that make up the sampling frame.
• Sampling Error: The difference between sample values and the population value. Occurs due to random sampling.
• Sampling Variation: Differences between samples from the same population due to chance.

How to Sample?

• Two requirements:
  • An available sampling frame, or create one
  • A suitable sampling method to draw a sample from the frame

Sampling Design Process: A Step-by-Step Process

• Define the Population
• Determine the Sampling Frame
• Select appropriate Sampling Technique(s)
• Determine Sample Size
• Execute the Sampling Process

Sampling Methods

• Probability Sampling:

  • Each unit in the population has a known probability of selection
  • Simple Random Sampling: Each unit has an equal chance of selection. Techniques include lottery method and using random number tables
    • Advantage: Most representative sample
    • Disadvantage: Identifying all population members can be difficult
  • Systematic Random Sampling: Uses a system to select the sample. For Example: selecting every 5th person in a list.
    • Advantage: Quick & efficient
    • Disadvantage: Not entirely bias free and each item does not have equal chance of selection
    • The system used to select subjects may introduce systematic error
    • Generalizations may be limited to the population sampled
  • Stratified Random Sampling: Divides the population into strata based on certain characteristics.
    • Random sampling occurs within each stratum
    • The sample size from each stratum should be proportional to its size in the population
    • Advantage: More representative in terms of control variables
    • Disadvantage: Selection of appropriate strata can be difficult. It may be difficult to identify every member of the population
  • Cluster (Area) Random Sampling: Randomly selects groups (clusters) and all members of the group are included in the sample
  • Multistage Random Sampling: Combinations of simple random and cluster sampling.
    • Example: Stage 1: random sampling of clusters (schools). Stage 2: Random sampling of individuals from the selected schools

• Non-Probability Sampling:

  • Not all units in the population have a known probability of selection
  • Deliberate (Quota) Sampling: Similar to stratified sampling, but uses non-random methods to select subjects meeting pre-established quotas.
    • Disadvantage: Selection bias
    • Difficult to establish quotas for all relevant characteristics
  • Convenience Sampling: Accessible subjects are selected. It's convenient but can be biased.
    • Examples: Using volunteers or intact classes
  • Purposive Sampling (Criterion-based Sampling): Uses criteria to select subjects.
    • Example: Studying smokers but not knowing all smokers, a random sample of people can be taken and then those who don't smoke are removed.
  • Snowball Sampling: Uses a starting group of subjects to identify other participants, expanding the sample like a snowball

Sample Research Examples

• Study on genital ulcer disease and HIV infection: First 20 patients were enrolled at the STD department each week for 40 weeks
• Study on prevalence of eye disease in north London: 17 general practice groups were selected randomly, with 750-850 participants in each.
  • Participants were 65 years or older and registered with the practices.
  • One-third of each practice was selected for the survey.

Conclusion

• For any research, the appropriate random sampling technique should be selected based on the study design and objectives. This is so that the findings can be properly generalized to the population from which the sample was drawn.

This quiz will test your understanding of the principles of sampling in epidemiology. You will learn why sampling is essential, how to select a representative sample, and the factors that influence sample size and characteristics. Prepare to assess your knowledge on the accuracy and feasibility of sampling methods.