The only type of selection process in the text book definition states simple random sampling uses selection of elements from the selection frame (Burns, 2011). References Burns, N.. Grove, S. K. , & Gray. J. (2011). Understanding nursing research: Building an evidence-based practice. Maryland Heights, MO: Elsevier/Saunders. Your post brings up the very important fact that the sample size need to be large enough to represent the target population (Ramsey, 2011).
Target population size eyeing too small is a common problem with random sampling as you have described with Autism ratios. If there were more states Involved with the CDC Autism survey, the results may differ. The question or research being done may not need a large population as well. Random sampling need to have a large enough size, must represent the target population and must truly be random (Ramsey, 2011 Ramsey, D. J. (2011). Statistics for dummies. Hoboken, N. J: Wiley.
Cluster sampling Is a productive way to obtain random sampling of a very large population. National restaurant chains test new menu items In this way. When I lived in Ohio, it was a big test market area. Supposedly the population of Ohio was thought to be representative of Midwestern people. Area restaurants such as McDonald's, Taco Bell, and Domino's introduced new menu items in this area to survey their popularity. I do not know what type of demographic data was used to determine that Ohio residents were representative of the Midwestern population.
It was nice to get to try the new menu Items before the rest of the country. The first McDonald's salads and Domino's try at breakfast pizzas were yummy. The Taco Bell French fries didn't make it. This was a good example of being part of a large cluster sampling process without ever knowing it. I think you covered all the different types of sampling very thoroughly. I am not sure the term census Is a type of random sampling. Census Is usually a method of attempting to collect date from an entire population (Ramsey, 2011).