Sampling Methods Study Guide

1.2: Sampling Methods

1.2.1 Introduction to Sampling

  • When studying a population, it is often impractical or impossible to examine the entire group due to:
    • Cost: High expenses in terms of time and money.
    • Practical Limitations: Example, testing all batteries for lifetime would deplete the inventory.
  • A sample allows researchers to infer characteristics of the larger population.
  • A well-chosen sample should ideally reflect the population’s diversity:
    • For a painkiller study, include:
    • Various body types (fat, skinny)
    • Different ages (old, young)
    • Health statuses (healthy, not healthy)
    • Genders (male, female)

1.2.2 Limitations of Sampling

  • No sampling method guarantees a perfectly representative sample.
  • Example of potential bias:
    • Random sample of a population with equal males and females may accidentally consist only of females.
  • Techniques that can yield reasonable approximations of the population include:
    • Simple Random Sample (SRS):
    • Every possible sample of size n has the same likelihood of selection.
    • Every individual has an equal chance of being included.
    • Difficulties:
    • Obtaining a complete list of individuals may be challenging.

1.2.3 Types of Sampling Techniques

1. Simple Random Sample (SRS)

  • They can be achieved by:
    • Drawing names from a hat.
    • Assigning numbers and using random number generators.
    • Example: Number each student in a classroom and use a random generator to select students.

2. Stratified Sampling

  • Defined as dividing the population into homogeneous groups called strata and conducting random sampling within each group.
  • Example Applications:
    • Musical preference surveys grouped by age.
    • Pricing surveys grouped by academic major.
  • Limitations: Risk of excessive subdivision, should ideally keep one stratification.

3. Systematic Sampling

  • Starts with a random selection of a starting point and measures every k-th individual.
  • Examples:
    • Selecting every 5th item on an assembly line.
    • Choosing every 10th name on a list.
  • Risks: May miss patterns in data due to the cycles of selection.

4. Cluster Sampling

  • Involves dividing the population into clusters (groups) and then randomly selecting entire clusters.
  • Examples:
    • Polling all businesses within randomly selected sections of a city.
    • Assessing tree health in randomly chosen forest sectors.
  • Common Confusion: Differentiate between stratified (all groups, some individuals) and cluster sampling (some clusters, all individuals).

5. Convenience Sampling

  • This technique utilizes individuals who are easy to reach/respond, leading to potential biases.
  • Example:
    • Polling people near a court house to gather opinions on the judicial system, likely not representative of the general population.
  • Recommendation: Avoid convenience samples for serious research as they do not yield reliable data.

1.2.4 Census

  • A census measures every individual in the population of interest.
  • Example: Banner Health may use a census method to obtain data on surgical complications.

1.2.5 Application of Sampling Techniques in Research

  • Various sampling techniques applied to a healthcare setting:
    • Split patients by surgery type and use stratified sampling.
    • Use random number tables for simple random sampling of patients.
    • Randomly select a subset of facilities for a cluster sample of all surgery patients.
    • Implement systematic sampling by tracking complications in every 100th surgery.
    • Convenience sampling results were poor due to selective data from facilities.

Homework Problems: Classification of Sampling Techniques

  1. Cholesterol levels in heart attack patients:
    • a. Cluster Sample
    • b. Systematic Sample
    • c. Convenience Sample
    • d. Simple Random Sample
    • e. Stratified Sample
  2. Quality control officer at a manufacturing plant:
    • a. Stratified Sample
    • b. Convenience Sample
    • c. Systematic Sample
    • d. Stratified Sample
    • e. Cluster Sample

Key Takeaways

  • When selecting sampling techniques, consider the population characteristics, the desired accuracy, and the feasibility of the method. Costs, practical constraints, and the nature of the data play vital roles in determining the best technique.

Definitions

  • Simple Random Sample (SRS): A method of selecting a sample so that every possible sample has an equal chance of being chosen.
  • Stratified Sample: A sampling method that divides the population into strata and randomly selects from each.
  • Systematic Sample: A sample obtained by selecting every k-th individual from a list after a random starting point.
  • Cluster Sample: A sampling method where entire clusters are selected and surveyed.
  • Convenience Sample: A non-probability sample that consists of individuals that are easy to reach.
  • Census: A complete enumeration of a population.