AP Stats - Experimental Design

Types of Random Sampling, Experimental Principles, and Design

SRS (Simple Random Sample)

Use of Random Number Generator (RNG) to generate a random sample

Stratified Random Sample

Split population into strata (group) based on characteristics they share, then RNG

Cluster Random Sample

Split the population into convenient clusters, then use RNG to choose which group in sample

Systematic Random Sample

Randomly choose a starting position in a sample, and choose with an equal interval from the chosen

Advantages of Stratified Random

-gets a more representative sample, equal chance of being chosen

Disadvantages of Stratified Random

Potential for overlap and issues with classifying a population member into more than one group

Advantages of Cluster Sampling

-Convenient and cost-effective, especially for sparse populations

Disadvantages of Cluster Sampling

More room for higher bias

Advantages of Systematic Random

Evenly distributed sampling, simple, larger variety

Disadvantages of Systematic Random

Possible data manipulation

Comparison (with two or more treatments), Must have random assignments to treatments, Must have control in avoiding confounding variables, Replication uses enough experimental units

Experimental Principles

Completely Randomized Design

Use of RNG to equally split people into two or more treatment groups

Randomized Blocking Design

Spliting volunteers into blocks based on similar characteristics, then given randomized treatmentment

Matched Pairs Design

Subjects are paired based on similar characteristics and randomly assigned treatments

OR

Each subject recieves both treatments, the order is randomized

Double Blind Experiment

Neither the participants nor the researchers know who is receiving the treatment or placebo, minimizing bias and ensuring more objective results

Confounding Variables

Variables not part of the experiment that can influence the response or dependent variable

Why isn’t observational studies not used for causation?

They only give correlation! Only controlled experiments prove causation

Experimental Units

The number of experimental material in which a treatment is assigned

Response Variable

the variable that is measured or observed to determine the effect of the explanatory (independent) variable

Treatments

conditions applied to experimental units to test a specific response variable