• ↑
  • STA258 Lecture 01 Raw

Sample Data

• It is usually difficult to measure every unit in a population.
• So you get measurements on all units in a sample.
• But the sample needs to be representative of the population. So we can make a generalization on the population.
  • $\text{Sample}\stackrel{\text{Generalize}}{\to }\text{Population}$
• Types of Samples

Issues with Collecting Sample Data

• Collecting sample data is expensive (time and money).
• Data might be biased:
  • Selection Bias
    • What if the sample isn't representative of the population?
  • Nonresponse Bias
    • What if participants refuse to cooperate?
  • Measurement Error Bias
    • What if the response from a unit is measured incorrectly?
• Ideal to get a Random Sample