Study design
Key concepts:
Types of Studies
- Experiments are studies that involve manipulating a treatment that each participant receives
- typically treatments are randomly assigned(see Randomization) to participants
- we then measure participants’ response to the treatment
- The treatment is the explanatory variable
- establish a cause-and-effect relationship
- We need to control everything(confounding) that could change the response, except for the treatments (randomization of treatments is the best way to accomplish this)
- Observational Studies are studies that do not involve manipulating the explanatory variables for participants.
- we are simply ”observing” what is going on without intervening
- nearly all surveys are observational studies
Sample
We want our sample to be representative of our population
- this means that our sample is nearly the same as our population, only smaller
- a representative sample allows us to generalize our results from the sample to the pop.
- biased: a sample that is not representative
Random Sample
We can choose people at random from our population to reduce the chances of getting a biased sample. Usually it is the best way to get a representative sample