Design of Experiments > Random Selection and Assignment
What is Random Selection?
Random selection means to create your study sample randomly, by chance. Random selection results in a representative sample; you can make generalizations and predictions about a population’s behavior based on your sample as long as you have used a probability sampling method.
The word “random” has a precise meaning in statistics. Random selection doesn’t just mean you can just randomly pick a few items to make up a sample. That method is actually something called haphazard sampling, where you try to create a random sample by haphazardly choosing items in order to try and recreate true randomness. That doesn’t usually work (because of something called selection bias). In order to create a true random selection, you need to use one of the tried and testing random selection methods, like simple random sampling.
Example of random selection: You are studying test taking behaviors at a college of 5,000 students. You choose every 50th student from a list (a random selection method called systematic sampling) to create a sample of 50 students to study.
Example of non random selection:
From the same list of 5,000 students, you randomly circle 50 names. This isn’t truly random as your biases (known or unknown) could affect who you circle. For example, you might unknowingly circle boys names over girls, or American-sounding names over foreign-sounding names.
What is Random Assignment?
Random assignment is where study participants are randomly assigned to a study group (i.e. an experimental group or a control group). In a single blind study, the participant does not know whether they are in the experimental group or the control group. In a double blind study, neither the participant nor the researcher knows.
Example of random assignment: you have a study group of 50 people and you write their names on equal size balls. You then place the balls into an urn and mix them well (this is a classic ball and urn experiment). The first 25 balls you draw go into the experimental group. The rest go into the control group.
Example of non-random assignment: you have a list of 50 people to assign to control groups and experimental groups. You use your knowledge and experience to choose 25 people who you think would be better suited to the experimental group (a method called purposive sampling).