Section 1.3: Data Collection and Experimental Design

Overview of Statistical Study Design

Designing a statistical study is fundamental to obtaining reliable and valid results. The process involves identifying the variable(s) of interest, selecting the population, planning data collection, describing the data, interpreting results, and checking for errors.

• Variable of Interest: The characteristic or property being studied.

• Population: The entire group about which information is desired.

• Sample: A subset of the population used to represent the whole.

• Descriptive Statistics: Methods for summarizing and describing data.

• Inferential Statistics: Methods for making predictions or inferences about a population based on sample data.

• Error Checking: Ensuring data accuracy and validity.

Types of Data Collection

Data can be collected through various methods, each suited to different research goals and practical constraints.

• Observational Study: The researcher observes subjects without interference. No treatment is applied.

• Survey: Data is collected by asking questions. Surveys can be conducted via interviews, forms, phone calls, or mail.

• Experiment: A treatment is applied to a group, and the effect is measured. A control group is often used for comparison.

• Simulation: Mathematical or computer models are used to imitate real-world processes, especially when direct study is impractical or dangerous.

Examples:

• Observational Study: Recording study hours of students without influencing their behavior.

• Survey: Asking students how many hours they study per week.

• Experiment: Testing a medication by giving one group the drug and another a placebo.

• Simulation: Car manufacturers use crash-test simulations to study collision effects.

Experimental Design

Proper experimental design ensures that results are valid and unbiased. Three key elements are control, randomization, and replication.

• Control: Keeping all variables constant except the treatment.

• Randomization: Assigning subjects to groups by chance to eliminate bias.

• Replication: Applying the treatment to many subjects to improve reliability.

Placebo and Placebo Effect

• Placebo: An inactive treatment used to control for psychological effects.

• Placebo Effect: Subjects respond favorably to a placebo due to belief in the treatment.

Confounding Variables

Confounding occurs when the effects of multiple factors cannot be distinguished, making it unclear which factor caused the observed change.

• Example: Starting a diet supplement and exercise simultaneously makes it unclear which caused weight loss.

Blinding Techniques

• Single Blinding: Subjects do not know if they receive the treatment or placebo.

• Double Blinding: Neither subjects nor experimenters know who receives the treatment.

Randomization Methods

• Completely Randomized Design: Subjects are randomly assigned to treatment groups.

• Randomized Block Design: Subjects are grouped by similar traits, then randomly assigned treatments within each group.

• Matched-Pairs Design: Subjects are paired by similarity; one receives the treatment, the other the control.

Sample Size and Replication

• Sample Size (n): The number of subjects in a study. Larger sample sizes improve reliability.

• Replication: Repeating the experiment on multiple subjects to confirm results.

Identifying Problems in Experimental Design

• Small Sample Size: Results may not be valid; replication is needed.

• Non-comparable Groups: Groups must be similar; random assignment within blocks is necessary.

Sampling Techniques

Sampling is used to study a portion of the population when a census is impractical. Different techniques affect the reliability and bias of results.

• Census: Measures the entire population.

• Sampling: Measures a subset of the population.

• Sampling Error: The difference between sample results and population results.

Types of Sampling Techniques

• Random Sample: Every member has an equal chance of selection.

• Simple Random Sample (SRS): Every individual and every group of the same size has an equal chance of selection.

• Stratified Sample: Population divided into strata; random samples taken from each stratum.

• Cluster Sample: Population divided into clusters; all members of selected clusters are surveyed.

• Systematic Sample: Every kth member is selected after a random start.

• Convenience Sample: Members are chosen based on ease of access; often leads to bias.

Examples of Sampling Techniques

• Stratified Sampling: Dividing students by major and randomly selecting from each major.

• Simple Random Sample: Assigning numbers to students and randomly selecting numbers.

• Convenience Sample: Selecting students from your own class.

Summary Table: Sampling Techniques

Key Formulas

• Sampling Error:

Important Definitions

• Population: The entire group being studied.

• Sample: A subset of the population.

• Variable: A characteristic measured in the study.

• Bias: Systematic error that skews results.