Data Classification
We want to study the effects of the supervolcano under Yellowstone national park erupting.
Simulation!
We choose to survey every 15th person on the Berlin Registrar's list of voters.
Systematic Sampling
Sometimes, a subject of an experiment will change their behaviors because they know they are receiving treatment. What do we call the process of giving some subjects a "placebo" and not telling subjects whether they are getting the "placebo" or "real" treatment?
Blinding
Why do we often use a sample for our study instead of getting information from the entire population?
- It may not be possible or may require too many resources to collect info from the entire population.
- If done right, a sample can give us a good idea for the population.
The Student ID numbers of the members of the BHS Marching Band.
Qualitative (not all #s are quantitative!), Nominal
We want to study the favorite colors of BHS freshmen.
Survey
We use a random number generator to choose 55 random people from the list of registered voters to survey.
Random Sampling.
Sometimes, a subject of an experiment will change their behaviors because they know they are receiving treatment. What do we call the process of giving some subjects a "placebo" and not telling subjects OR the researchers interacting with them whether they are getting the "placebo" or "real" treatment?
Double-Blinding
We conduct a study where we observe 50 chimpanzees in zoos. Can this study be used to generalize to chimp behavior in nature?
No...
The high temperature from each day in August.
Quantitative, Interval
We want to study the effects of increased exercise on resting heart rate.
Experiment
We group registered voters by levels of household income, then choose 15 people from each level to survey.
Stratified Sampling
We take the subjects in the subject pool, split them into male and female groups and then assign half of each group to receive medication and half of each group to get a placebo.
Randomized Block Design
If a study is conducted on 150,000 male college students in the United States, what is the population that this study can be generalized to?
college aged, college enrolled males in the United States
The letter grades of each BHS Math Teacher in their college Differential Equations course.
Qualitative, Ordinal
We want to study the interactions between chimpanzees and elephants in nature.
Observational Study
We ask people as they leave Stop and Shop
Convenience Sampling :(
Matched Pair Design
Why is non-response bias the hardest bias to remove from your study?
It is the one bias that you have no control over...
The time it takes Mr. T to drive to work each morning in the month of November.
Quantitative, Ratio
We want to study how likely people are to shoplift in a given store.
Observational Study
We find that Town Hall has a pre-made study group that is a mix of Berlin residents and choose to survey every member of that group.
Cluster Sampling.
We use a random number generator to assign members of the subject pool to get either a full dose, half dose, or no dose of a medication in our study.
Completely Randomized Design
You choose a random sample to survey, but over half of the people you reach out to don't respond! What type of bias has been introduced to your study?
Non-response Bias
The elevation of each town in the Netherlands.
Quantitative, Interval
We want to study the impact that a new community center would have on traffic on Patterson Way.
Simulation.
The MTA chooses to search every 5th person that boards a train.
Systematic Sampling
We split the subject pool up into people with low blood pressure, moderate blood pressure, and high blood pressure. We then assign one third of each group to receive a placebo, one third to receive medication A, and one third to receive medication B.
Randomized Block Design
You ask someone a question about senior citizen rights while wearing a tee-shirt with your Grandma's face on it. What type of bias have you introduced to your study?
Response Bias
The ranks of each member of the Berlin Police Department
Qualitative, Ordinal
We want to study the effects of 15 minutes of sunlight immediately after waking up on a person's cognitive performance an hour later.
Experiment.
We put each BHS student's name in a hat and pull 30 names to survey without looking.
Random Sampling.
Each member of the subject pool is assigned to four additional members with similar characteristics. From here, one member is given Supplement A, one member is given Supplement B, one is given C, D, E...
Matched Pair Design
We decide to gather information about townspeople's opinions on a new town project by holding a Town Hall at 8:30 pm on a Wednesday night. How has selection bias been introduced to your study?
...