AP STAT REVIEW: Sampling/Experiments

Observational Study

Observing with no treatment

Experiment

Imposing treatment

What does a larger sample size do?

Reduces variability when the sampling design is NOT flawed

Nonresponse bias

Selected people do not respond

Undercoverage Bias

Systematically excluding people from being able to be selected

Response Bias

Providing inaccurate responses (on purpose or by accident)

Wording Bias

Confusing wording or question is slanted towards a particular response

Stratified Random Sample

Division of a population into separate groups (strata) based on shared attributes (homogeneous grouping), SRS is done to select people within the strata, and those units are combined to form the sample.

Experimental unit

Used when an experiment is not performed on humans, but on animals or objects

Subjects

Used when an experiment is done on humans

Confounding Variable

Variable that is related to the explanatory variable and influences the response variable, makes it challenging to determine cause and effect

What should a well-designed experiment include?

1. Comparisons of at least 2 treatment groups (one could be a control)

2. Random assignment of treatments to experimental units

3. Replication: enough experimental units in each treatment group to be able to detect a difference

4. Control of a potential confounding variables where appropriate

Bias

Systematic tendency to overestimate or underestimate the true population parameter

Randomized Block Design

Treatments are assigned completely at random within each block. For each block individuals are similar to one another with respect to at least one blocking variable

Benefits of Block Design

reduce variability of results within each treatment group and eliminate possibility of the blocking variable as the confounding variable

Can results be generalized to a larger population

Yes, but only to the population that the subjects/units were randomly selected from

What are two poor sampling methods?

Convenience sampling and voluntary response sampling (neither uses chance to select individuals)

Simple Random Sample (SRS)

A sample in which every group of a given size has an equal chance of being chosen

How to obtain an SRS

number individuals, use a random # generator/table/draw from hat to select which to include in study (excluding repeats)

Control Group

collection of experimental units that are either not given a treatment of interest or given a treatment with an inactive substance (placebo).

Purpose of control group

Provide a baseline for comparison

Explanatory Variable

a variable whose levels are intentionally maipulated

Response Variable

an outcome from the experimental units that is measured after the treatments have been administered

Completely Randomized Design

Treatments are randomly assigned to experimental units, tends to create roughly equivalent groups so that differences in responses can be attributed to the treatments

Matched Pairs Design

Special randomized block design, each pair receives both treatments by randomly assigning one treatment to one member and the remaining to the other, or both subjects get both treatment

Systematic Random Sample

members from a population are selected according to a random starting point and a fixed, periodic interval, (like every 5th person)