This lesson explores the basic definition of why there is the label of quasi-experimental design in addition to what types of designs are quasi-experimental.
I’m going to do a psychological experiment where I make people different ages and then see how they react to loud noises.
Well, I’d like to. Unfortunately, even with our advanced quantum physics and computers, we cannot reverse or control age like that. Psychological researchers are forced to work around the issue.Because we can’t reverse someone’s age, we have to work with people who are already that age. But, we miss some things in the process. But, I’m getting ahead of myself.
A true experiment has one main component – randomly assigned groups. This translates to every participant having an equal chance of being in the experimental group, where they are subject to a manipulation, or the control group, where they are not manipulated.A quasi-experiment is simply defined as not a true experiment. Since the main component of a true experiment is randomly assigned groups, this means a quasi-experiment does not have randomly assigned groups.
Why are randomly assigned groups so important since they are the only difference between quasi-experimental and true experimental?When performing an experiment, a researcher is attempting to demonstrate that variable A influences or causes variable B to do something. They want to demonstrate cause and effect. Random assignment helps ensure that there is no pre-existing condition that will influence the variables and mess up the results.A silly example would be something like, ‘Does chemical X1 cause blindness?’ If you accidentally put all of the people wearing glasses in the condition where you spray X1 in someone’s face, then your results are going to be skewed. This is an extreme and overly simplistic example, but it is demonstrating why normally an experimenter wants to randomly assign people into different groups. Let’s look at some more realistic and typical quasi-experiments in psychology.
Sometimes a researcher needs a particular type of participant or they only have access to a certain group of participants.
This means that the researcher collects participants in a group that cannot or should not be divided up, or more simply, the researcher cannot randomly assign the participants. This non-equivalent group is defined as an experiment where existing groups are not divided.An experiment using non-equivalent groups might take place at a mental health institution. You cannot randomly assign people to therapy and others to not have therapy. That would be unethical. So, you’re forced to assign the entire group to therapy, which means no random assignment.
It is possible to have multiple groups. In our mental health institution example, let’s say that the staff had divided up everyone into three groups. Furthermore, let’s say you have a new type of therapy and an old type of therapy, so nobody is going without.
Randomly assigning the groups, to try and make your study a true experiment, is not sufficient. This is because there is no telling why an individual was assigned to any of the three groups. The reason an individual might be in group B and not in group A could skew your results. You need to be able to assign individuals to the treatment or alternate treatment groups to claim it as a true experiment.
A researcher finds a group of people to test.
Then the researcher introduces a manipulation that should change the people and test to see if there were any changes. For example, you test a group of people on their knowledge of U.S.
history. Then you assign them a study packet and test them again to see if their knowledge has increased. This is known as a pretest-posttest design, which is when participants are studied before and after the experimental manipulation.A researcher can use pretest-posttest in an almost unlimited number of ways, as long as they follow the steps:
- Test the participants prior to the experimental manipulation.
- Perform the experimental manipulation, which is a fancy way of saying that you would do something to the group, like give them homework or give them therapy or deafen them with noise.
- Test the participants after the manipulation to see what changes occurred.
The reason pretest-posttest is considered a quasi-experimental design is because the majority of researchers will manipulate their entire group.
This gives them a larger sample size to see if their manipulation actually changed the group. It is possible to randomly assign people to the experimental or control condition to make it a true experiment, but you’re reducing your sample size, and this could put a strain on your statistics.Another example of a pretest-posttest design might be examining the effects of not sleeping. You take participants and test them to see how good their judgment is, their knowledge and their hand-eye coordination. Then you keep them up all night with cola, games and bright lights. Keeping them up is your experimental manipulation. Lastly, you test them in the morning to see what effect the lack of sleep had on their judgment, knowledge and hand-eye coordination.
Cross Sectional and Longitudinal
Other times you want to study things like age, but age is such a pain to study because you can’t control it. So, that leaves you with two options:
- Cross-sectional designs, which is when participants are representatives of age groups along a developmental path to determine how development at different ages influences a dependent variable.
- Longitudinal designs are when a sample of the population is studied at intervals to examine the effects of development.
Cross-sectional designs take samples from across the age continuum as representatives. So, if you were interested in people’s ability to solve complex math across a lifetime, you would select people who were 5, 10, 15, 20 and so on and then have them solve complex math. The idea here is that a large group of people who are 15 will allow you to create a generalizable statement about 15-year-olds. You would then compare your results of 15-year-olds to 10- and 5-year-olds.
A longitudinal design involves taking a group of people who are all the same age and then checking in with them every couple of years. So, in our math example, you get a bunch of 5-year-olds to do math. Then every five years, you would check back in with them and see how they solve the math problems. These are quasi-experimental because you cannot control age and randomly assign it.
There is no tricky way to get around that problem.
Ex Post Facto
The last type of quasi-experimental study focuses on other things you can’t control. For example, I can’t assign you obesity. I mean I could, but I don’t have the money to feed you. This is known as an ex post facto design, which is defined as how an independent variable, present in the participants prior to the study, affects a dependent variable.
Besides things like obesity, a researcher sometimes needs to work with people who have schizophrenia, traumatic brain injuries or diseases. The researcher cannot randomly assign these things, so they have to work with what they can find.An experiment using an ex post facto design might look at the emotional stability of tall people as compared to short people. You can’t assign height, so you find a bunch of tall people and short people and test their emotional stability. I’m a tall person, so I have my money on them being fairly emotionally stable.
Quasi-experiment is simply defined as a not true experiment. A true experiment has one main component – randomly assigned groups.The main types of quasi-experiments are:
- Non-equivalent groups, defined as existing groups that are not divided
- Pretest-posttest design, defined as participants who are studied before and after the experimental manipulation
- Cross-sectional designs, which is when participants are representatives of age groups along a developmental path to determine how development at different ages influences a dependent variable
- Longitudinal designs, defined as a sample of the population that is studied at intervals to examine the effects of development
- Ex post facto designs, defined as how independent variables, present in the participants prior to the study, affect a dependent variable
After you have finished with this lesson, you should be able to:
- Differentiate between quasi-experiment and true experiment
- Describe the main types of quasi-experiments
- Identify the steps that must be followed in the pretest-posttest design
- Explain the differences in cross-sectional and longitudinal designs
- Define ex-post facto designs