Upload
shana-malone
View
216
Download
0
Embed Size (px)
Citation preview
Warm-upWarm-up A newspaper article about an opinion poll A newspaper article about an opinion poll
says that “43% of Americans approve of says that “43% of Americans approve of the president’s overall job performance.” the president’s overall job performance.” Toward the end of the article, you read : Toward the end of the article, you read : “The poll is based on telephone interviews “The poll is based on telephone interviews with 1210 adults from around the United with 1210 adults from around the United States, excluding Alaska and Hawaii.”States, excluding Alaska and Hawaii.”– What variable did this poll measure?What variable did this poll measure?– What population do you think the newspaper What population do you think the newspaper
wants information about?wants information about?– What was the sample?What was the sample?– Are there any sources of bias in the sampling Are there any sources of bias in the sampling
method used?method used?
Section 4.2Section 4.2Designing Designing
ExperimentsExperiments
Experiment vs. Observational Experiment vs. Observational StudyStudy
Recall: in an experiment, we actually Recall: in an experiment, we actually do something to people, animals, or do something to people, animals, or objects in order to observe the objects in order to observe the response.response.
Observational studies observe Observational studies observe individuals and measure outcomes individuals and measure outcomes but do not impose treatment on the but do not impose treatment on the individuals.individuals.
TerminologyTerminology
Experimental units: the individuals on Experimental units: the individuals on which the experiment is conducted.which the experiment is conducted.– When units are people, we call them When units are people, we call them
subjects or participants.subjects or participants. Treatment: the specific experimental Treatment: the specific experimental
condition applied to the units.condition applied to the units.– Example: a treatment might be 500 mg Example: a treatment might be 500 mg
of naproxen, or 30 minutes of exercise of naproxen, or 30 minutes of exercise per day…per day…
TerminologyTerminology
It is important to note the difference It is important to note the difference between response and explanatory between response and explanatory variables in an experiment.variables in an experiment.
Factor – another name for Factor – another name for explanatory variableexplanatory variable
Level – a specific value of a factor Level – a specific value of a factor (ex. dosage of medication)(ex. dosage of medication)
Placebo – a “dummy” treatmentPlacebo – a “dummy” treatment
Example – Physician’s Health Example – Physician’s Health StudyStudy
Does regularly taking Does regularly taking aspirin or beta carotene aspirin or beta carotene help reduce the risk of help reduce the risk of a heart attack?a heart attack?
Subjects: 21,996 male Subjects: 21,996 male physiciansphysicians
Factors: Aspirin and Factors: Aspirin and Beta CaroteneBeta Carotene
Levels for each: Yes/NoLevels for each: Yes/No Treatments: Four Treatments: Four
optionsoptions
Aspirin Aspirin and Beta and Beta CaroteneCarotene
Aspirin Aspirin and and placebo placebo Beta Beta CaroteneCarotene
Placebo Placebo Aspirin Aspirin and Beta and Beta CaroteneCarotene
Placebo Placebo Aspirin Aspirin and and Placebo Placebo Beta Beta CaroteneCarotene
Beta Carotene
Yes No
Asp
irin
No
Yes
Some ways to make Some ways to make experiments betterexperiments better
You need to have a control group. You need to have a control group. This is a group in which no treatment This is a group in which no treatment is given but all other aspects of the is given but all other aspects of the situation are the same.situation are the same.
The placebo effect… Some patients The placebo effect… Some patients respond favorably even if they are respond favorably even if they are not receiving treatment but think not receiving treatment but think they are.they are.
Comparative ExperimentsComparative ExperimentsExample of Single TreatmentExample of Single Treatment
Units Treatments Observe/Measure Response
Single treatment designs are not optimal. There is no control group, and the design doesn’t control for placebo effect (when patients expect relief or improvement).
Comparative ExperimentsComparative ExperimentsUsing a control groupUsing a control group
Units
Treatment 1
Treatment 2:Placebo
Compare Response
Group 1
Group 2
The question is HOW DO WE DECIDE WHICH UNITS RECEIVE
WHICH TREATMENT???
The answer is RANDOMIZATION. You
MUST label this if using a diagram and make a
statement as to how you are going to randomize!
Principles of Experimental Design Principles of Experimental Design – Things you should consider and – Things you should consider and
discuss in your answers.discuss in your answers. Control – comparison of several Control – comparison of several
treatments in to a group without treatments in to a group without treatment is the simplest form.treatment is the simplest form.
Randomization – randomly assign subjects Randomization – randomly assign subjects to treatment groups in order to reduce to treatment groups in order to reduce systematic differences among the groups.systematic differences among the groups.
Replication – Replicate each treatment on Replication – Replicate each treatment on many subjects to reduce chance variation many subjects to reduce chance variation in the results. You are more likely to find in the results. You are more likely to find statistical significance if you have more statistical significance if you have more people.people.
Illustration of Completely Illustration of Completely Randomized DesignRandomized Design
Random Assignment
Treatment 1
Treatment 2:Placebo
Compare Response
Group 1
Group 2
Random Assignment illustrates the principle of RANDOMIZATION.
Having a control group illustrates the principle of
CONTROL.
Choosing an adequately
large sample ensures
REPLICATION.
CautionCaution
There is a difference between There is a difference between random random selectionselection of participants and of participants and random random assignmentassignment of subjects to of subjects to treatment groups.treatment groups.– Choosing an SRS is random selection. Choosing an SRS is random selection.
All of these participants are in the All of these participants are in the experiment.experiment.
– Then, randomly assign those subjects to Then, randomly assign those subjects to treatment groups. treatment groups.
Example 2Example 2 A pharmaceutical engineer is studying the effects A pharmaceutical engineer is studying the effects
of a new medicine for pain relief. She wants to of a new medicine for pain relief. She wants to try two different dosages (500 mg and 1000 mg) try two different dosages (500 mg and 1000 mg) and 3 different daily intakes (1/day; 2/day; and and 3 different daily intakes (1/day; 2/day; and 3/day). Five people will be tested at each 3/day). Five people will be tested at each combination of dosage and daily intake.combination of dosage and daily intake.
Identify all the explanatory variables.Identify all the explanatory variables. How many subjects are needed?How many subjects are needed? Outline in diagram form an appropriate design for Outline in diagram form an appropriate design for
this experiment. Indicate how many people are this experiment. Indicate how many people are assigned to each treatment groupassigned to each treatment group
Use Table B starting at line 108 to select the Use Table B starting at line 108 to select the people assigned to the 1people assigned to the 1stst treatment group. How treatment group. How did you label the people?did you label the people?
Cautions about Cautions about ExperimentationExperimentation
We need to treat all the experimental units We need to treat all the experimental units in the exact same way except for the in the exact same way except for the difference in treatment.difference in treatment.
One way to accomplish this is by using a One way to accomplish this is by using a “double-blind” experiment.“double-blind” experiment.– Neither the subjects nor the personnel Neither the subjects nor the personnel
who have contact with them (especially who have contact with them (especially the ones collecting the data) are aware the ones collecting the data) are aware of which treatment each subject of which treatment each subject receives.receives.
Other Weaknesses in Other Weaknesses in Experimental DesignsExperimental Designs
Lack of realism is a serious potential Lack of realism is a serious potential weakness.weakness.– In 1986, before a third brake light was In 1986, before a third brake light was
required on cars, an experiment found required on cars, an experiment found that adding the third brake light to cars that adding the third brake light to cars would reduce rear-end collisions by 50%would reduce rear-end collisions by 50%
– In actuality, the reduction was only 5%. In actuality, the reduction was only 5%. What happened???What happened???
Another Favorite of the AP Another Favorite of the AP Gang – Matched PairsGang – Matched Pairs
Gives individuals both treatments – Gives individuals both treatments – very favorable because you can see very favorable because you can see differences for each person.differences for each person.
Example: You want to know what Example: You want to know what type of music makes students type of music makes students perform better on tests. You could perform better on tests. You could give each student the same test give each student the same test twice but with different music each twice but with different music each time. The order of test would be time. The order of test would be random.random.
Block DesignsBlock Designs
Grouping by another factor first.Grouping by another factor first. Example: You might “block” a cancer Example: You might “block” a cancer
treatment study into gender.treatment study into gender. Block designs can have blocks of any size Block designs can have blocks of any size
(ex. 20 females and 50 males in the (ex. 20 females and 50 males in the experiment).experiment).– In a block design, the random assignment of In a block design, the random assignment of
units to treatments is carried out separately units to treatments is carried out separately within each block.within each block.
General Format of a Blocked General Format of a Blocked DesignDesign
Subjects
Block 1
Block 2
Treatment 1
Treatment 2
Treatment 3
Treatment 1
Treatment 2
Treatment 3
Observe/Compare Response Variable
Observe/Compare Response Variable
Random
Assignmen
t
Random
Assignment
Putting subjects into blocks is NOT random.
2002 #22002 #2 A manufacturer of boots plans to conduct an A manufacturer of boots plans to conduct an
experiment to compare a new method of experiment to compare a new method of waterproofing to the current method. The waterproofing to the current method. The appearance of the boots is not changed by either appearance of the boots is not changed by either method. The company recruits 100 volunteers in method. The company recruits 100 volunteers in Seattle, where it rains frequently, to wear the boots Seattle, where it rains frequently, to wear the boots as they normally would for 6 months. At the end of as they normally would for 6 months. At the end of the 6 months, the boots will be returned to the the 6 months, the boots will be returned to the company to be evaluated for water damage.company to be evaluated for water damage.
A) Describe a design for this experiment that uses A) Describe a design for this experiment that uses the 100 volunteers. Include a few sentences on how the 100 volunteers. Include a few sentences on how it would be implemented.it would be implemented.
B) Could your experiment be double-blind? Explain.B) Could your experiment be double-blind? Explain.
HomeworkHomework
Chapter 4 Chapter 4
# 46, 50, 54, 56, 62# 46, 50, 54, 56, 62