Sampling Course

Lesson 1 • Overview of Sampling Design Types

Maps the landscape of probability and non-probability designs at a high level. Prepares students to choose appropriate methods in later chapters.

Lesson 2 • Populations and Samples Defined

Clarifies the distinction between a target population and a drawn sample. Establishes vocabulary used throughout all subsequent sampling decisions.

Lesson 3 • Sampling Error and Bias

Differentiates random sampling error from systematic bias and explains their impact on conclusions. Provides a basis for evaluating sample quality throughout the course.

Lesson 1 • Estimating Parameters from SRS

Derives mean, proportion, and total estimators under SRS and computes their standard errors. Links estimation formulas to the concept of sampling distributions.

Lesson 2 • Constructing a Sampling Frame

Explains what a sampling frame is, how to build one, and common frame errors. A sound frame is the prerequisite for all probability sampling methods.

Lesson 3 • Comparing SRS and Systematic Efficiency

Evaluates relative efficiency, cost, and practical constraints of each method. Students select between SRS and systematic sampling for given scenarios.

Lesson 4 • Simple Random Sampling Mechanics

Covers lottery, random-number table, and software-based selection methods. Students execute SRS procedures and verify equal-probability selection.

Lesson 1 • Post-Stratification and Weighting

Addresses adjusting weights after data collection when pre-stratification was incomplete. Introduces raking and calibration as practical correction tools.

Lesson 2 • Principles of Stratification

Explains why dividing a population into homogeneous strata reduces overall variance. Connects stratification logic to precision gains over SRS.

Lesson 3 • Proportional and Optimal Allocation

Contrasts proportional allocation with Neyman optimal allocation and cost-constrained variants. Students calculate stratum sample sizes under each rule.

Lesson 1 • Design Effect and Sample Size Adjustment

Quantifies the design effect (DEFF) and uses it to inflate SRS sample size requirements. Students recalculate required sample sizes for clustered designs.

Lesson 2 • Probability Proportional to Size Sampling

Details PPS selection methods including systematic PPS and Lahiri's method. Demonstrates how PPS achieves self-weighting samples in unequal-cluster designs.

Lesson 3 • Two-Stage and Multistage Designs

Extends cluster sampling to two or more stages with subsampling within selected clusters. Addresses estimation complexity and variance decomposition across stages.

Lesson 4 • Cluster Sampling Logic and Trade-offs

Explains why clustering reduces travel and listing costs while increasing variance. Contrasts cluster sampling with stratified sampling conceptually.

Lesson 1 • Snowball and Network Sampling

Covers respondent-driven referral chains for hard-to-reach populations. Introduces respondent-driven sampling as a probabilistic extension of snowball methods.

Lesson 2 • Theoretical and Purposeful Sampling in Qualitative Research

Addresses maximum variation, deviant case, and theoretical saturation strategies. Connects non-probability logic to qualitative research validity standards.

Lesson 3 • Evaluating Non-Probability Sample Quality

Provides criteria for assessing rigour when probability sampling is absent. Students apply transparency and fitness-for-purpose standards to non-probability designs.

Lesson 4 • Convenience and Purposive Sampling

Describes availability-based and judgment-based selection and their inherent bias risks. Identifies research contexts where these methods are defensible.

Lesson 1 • Sample Size for Proportions

Adapts the size formula for binary outcomes and addresses the conservative p=0.5 assumption. Students compute sizes for single and multiple proportions.

Lesson 2 • Adjustments for Complex Designs

Applies design effect multipliers and non-response inflation to base sample size calculations. Integrates stratification and clustering adjustments from prior chapters.

Lesson 3 • Power Analysis for Hypothesis Testing

Links sample size to statistical power, effect size, and Type I and Type II error rates. Students use power curves to justify sample sizes for comparative studies.

Lesson 4 • Precision-Based Sample Size for Means

Derives the sample size formula for estimating a population mean within a specified margin of error. Requires knowledge of variance estimation from earlier chapters.

Lesson 1 • Modes of Data Collection

Compares in-person, telephone, mail, and web-based modes on cost, coverage, and response quality. Students select modes appropriate to their sampling frame and population.

Lesson 2 • Non-Response and Follow-Up Strategies

Identifies unit and item non-response sources and designs follow-up protocols to reduce them. Links non-response rates to bias risk established in Chapter 1.

Lesson 3 • Questionnaire Design Principles

Covers question wording, response scale construction, and order effects. Connects instrument quality to the accuracy of sampled data.

Lesson 4 • Interviewer Training and Quality Control

Establishes protocols for training interviewers and monitoring data collection quality. Addresses interviewer effects as a source of measurement error.

Lesson 1 • Handling Missing Data and Imputation

Distinguishes missing-at-random from not-missing-at-random mechanisms and applies hot-deck and multiple imputation methods. Evaluates imputation impact on estimates.

Lesson 2 • Constructing Survey Weights

Derives base weights from selection probabilities and applies non-response and calibration adjustments. Builds on stratification and clustering concepts from earlier chapters.

Lesson 3 • Variance Estimation for Complex Samples

Covers Taylor linearisation, balanced repeated replication, and jackknife methods for variance estimation. Students select the appropriate method for their design.

Lesson 4 • Reporting and Communicating Results

Structures survey reports with appropriate precision statements, limitations, and design disclosures. Translates technical estimates into accessible findings for diverse audiences.