Shannon Starke and Kimberly Stinedurf

1
Chapter 5

• Shannon Starke and Kimberly Stinedurf
• Block 4

2
Designing Samples
3

• What is a sample?? A sample is a small
• part of the population used to gather
  information.
• A simple random sample (SRS) guarantees that all
  individuals have an equal chance of being
  selected.

4

• What is meant by design?? The design is the
  method used to choose a sample.
• A) Convenience Sample - when the easiest people
  to reach are chosen.
• B) Voluntary Response - Made up of people who
  choose to respond themselves.
• Population is the entire group we want
  information of.

5
Extras to do with Sampling
6

• Multi-Stage Sampling - When you select
  successively smaller groups within the population
  in stages.
• What is meant by a Block in my sample?? - Group
  of subjects that are similar in ways that can
  affect the response to treatment.

7

• Bias - When a study favors certain outcomes.
• Confounded - When the effects of two variables
  cannot be distinguished from one another.
• How can I have a stratified sample?? - The
  population are divided into similar groups that
  are important to the response.

8

• Non-Response - This occurs when an individual
  chosen for the sample can't be contacted or
  refuses to cooperate.
• Under coverage - This occurs when some groups
  in the population are left out of the process
  when choosing a sample.
• What exactly is meant by Placebo?? - This is
  when a fake treatment is given to the subjects.

9
Key strokes for Random Digits

• MATH gt PRB 5
• In ( )
• (Lower bound, Upper bound, number of trials)

10
Designing Experiments
11

• An observational study gathers data but
  contains no interference, but an experiment
  imposes some type of a treatment.
• In a Double Blind experiment, the experimental
  units and the people giving the experiment (or
  who have contact with them) do not know which
  treatment they are being given.

12
Comparative Data

•  
• Treatment Observation
•  
• Observation 1 .. Treatment .. Observation 2
•  
•  

13
Completely RandomizedExperiments
Group 1 Treatment 1
Measure

• Random
  
  Compare
• Allocation
  
  Results

Group 2 Treatment 2
Measure
14

• Matched Pairs Design is a common form of
  blocking that compares two treatments. Sometimes,
  each subject could end up receiving both
  treatments..
• To remember what to do with an experiment, you
  can remember "Randy Blake Can't Reply". This
  all means Randomize, Block, Control group, and
  Replicate.

15
Simulating Experiments
16

• Here are the steps of Simulation
• a) Describe the problem or describe the
  experiment.
• b) List the assumptions.
• c) Assign digits which will represent outcomes.
• d) Simulate (use many repetitions).
• e) State your conclusions.