Sociology 303 Statistics - Lecture 2: Sampling

Instructor: Dr. Alvarez

Key Objectives

• Understand the difference between sample statistics and population parameters.
• Define inference and its role in statistics.
• Differentiate between sample types.

Introduction to Sampling

• Sampling: Method of collecting data and identifying study participants.
• Importance: Selection impacts data and study outcomes.
• Highlights real-world implications, e.g., political polling.

Understand Population vs. Sample

• Population: Entire group of interest defined by the researcher.
  • Examples include households, businesses, students, etc.
• Sample: Subset from the population used to infer the population's characteristics.
  • Must ideally be a representative sample to accurately reflect the population.

Challenges with Population Data

• Population data is expensive and collected infrequently (e.g., U.S. Census every 10 years).
• Samples are used instead due to cost and timeliness issues.

Inference in Statistics

• From samples, make inferences about populations.
• Sample Statistic: Data derived from the sample.
• Population Parameter: Estimated characteristic based on the sample statistic.

Types of Sampling

Non-Probability Sampling

• Convenience Samples: Easy to collect, but not representative.
  • Snowball Sample: Uses social networks to find participants.
  • Purposive Sampling: Recruiting based on researcher judgment.
• Common in qualitative research for in-depth understanding but lacks generalizability.

Probability Sampling

• Probability/Randome Samples: Best for representativeness.
• Types:
  • Simple Random Sample: Equal probability for each member, ideal but difficult.
  • Systematic Random Sample: Selects every kth member; risks periodicity bias.
  • Stratified Random Sample: Divides population, samples each subgroup.
    • Proportionate vs. Disproportionate: Matches population distribution versus oversampling for detailed analysis.
    • Weighting: Corrects bias in disproportionate samples.
  • Multi-Stage Cluster Sample: Cost-effective, geographically grouped sampling; combines with stratification.

Importance of Representative Samples

• Reduces bias and errors in estimates.
• Ensures findings can generalize to the entire population.

Descriptive vs. Inferential Statistics

• Descriptive: Describe sample characteristics.
• Inferential: Use sample data to infer population characteristics.

Summary

• Importance of using probability sampling to reduce bias and improve representativeness.
• Overview of sampling types, pros, and cons.
• Email Dr. Alvarez for questions or clarifications on sampling techniques.