On the new SAT there are questions about statistics that I have not seen on previous tests (nor on the ACT). The questions cover distribution, standard deviation, and sampling. This is all fun for me, for it is a big part of what I studied at Penn State to get my BS in Industrial Engineering.
To get things going on these questions, I am going to start with sampling. Or how to pick what will be surveyed or inspected. Sampling assumes that not everyone or everything can be questioned or inspected in a population. Political polls are a great example of sampling. All manufacturing uses sampling for quality control.
The major methods are:
Random – just as it sounds, the sample is totally random and picked across the whole of the population being tested. A random number generator is handy.
Convenience – the easiest method, but also biased. The classic example is to interview the first 50 people that come by a location.
Cluster – pick a group and hope it is an unbiased representative.
Systematic – set up a procedure to pick every nth sample across the population. Useful on an assembly line.
Stratified – break the population into layers and then chose random samples (or one of the other techniques). Bernie Sanders loves this method for it shows that young people prefer him over Hillary. She likes it too because ethnic voters prefer her. Must have good data on the population.
The SAT does not use these words, they ask questions that incorporate the methods and want to know if there is a bias when using them. More to follow.
Here is a nice youtube video that explains these methods well (time 4:53)