Observational Studies
When completing an observational study, would you want a smaller sample number or a larger one? why?
Larger! So the sample is closest to the population!
blah
blah
What is a placebo?
It is a "fake" treatment. Think of a sugar pill instead of a regular medication
For the Scope of Inference to be able to actually describe a full population, would you want to have:
-complete random sampling? Or
-partial Random Sampling?
Complete Random Sampling
A sample's mean foot length is 10.5 inches with a Standard Deviation of .75 inches. What is the expected largest foot length in the entire population based off this data?
12 inches
(10.5 + 2(.75) = 12)
How would you describe the difference between a double blind study vs single blind study?
Single blind study means the experimenters are the only ones who know which group receives a specific treatment and who received the other treatments. Double blind study means the experimenters and subjects are unaware.
What type of Bias occurs if you want to survey a group of former prison inmates, which then causes you to call a large number of houses and ask "Are you available to discuss any of your past imprisonments and federal charges?" And you subsequently get very little (almost no) answers to your survey?
Nonresponse Bias
The standard deviation of a sample of hair length is 6 inches from a mean length of 3.5 inches. What is the longest expected hair length in the population? What is the shortest expected length? What does this mean?
Longest = 13 inches
Shortest = -1 inches
The expected lengths of hair in the population goes up to 13 inches, and the shortest possible length is not possible. This value is based off an estimation of the sampling.
Why is a random assignment of treatments needed to be able to make a true inference about cause and effect?
Because this is the only way to eradicate any amount of unforeseen bias that could occur if the subjects were able to influence the treatment they received.