Types of Quantitative Research

Just in Case you aren’t a Math Nerd, like I am, but still somehow have to understand this to make it through your research class!

Diet Pepsi: You’ve got the right one baby!

Experimental research follows the scientific method, with a question needing to be answered, variables that can be measured, calculated and compared. The most important thing is that it’s done in a controlled environment. You will have a test group and a training group, where one group gets something the other does not.  To me, this is the right way to do things.

Root Beer: Is it soda or beer?

A quasi-experiment is kind of an experiment, but kind of not. It estimates the cause of an intervention on a certain population, but without randomness. Got to think twice about this one, because who knows where they started? Was it at the same point? The training group and test group should have the same starting point, so we can tell what the actual cause was.

Soda Water: What the heck? This isn’t even soda!

Non-experimental research is the researcher trying to predict the outcome without actually doing an experiment. He is unable to manipulate the variable in the non-study. Exactly! What are you really doing??? You’re not even experimenting: hence the name. So the researcher must come to his very informal conclusion by some sort of correlation (not scientific) or case study, or some other way to fabricate a relationship or conclusion.

There are 3 things, or conditions, that must be present in order to determine cause and effect. And yes, you need all three.

1)            The first step in establishing causality is looking at the numerical relationship. Well, to me anyway. You have to check with the correlation coefficient is. So either the variable will be super related to the outcome or it will be super unrelated. So, you’re looking for the number to be between 0 and 1, the closer to 1, the stronger the correlation. And the same for 0 to -1.

• The second thing is the chicken before the egg thing. Did the chicken cause the egg? Or, did the egg cause the chicken? You have to, again, be super sure (preferably with numbers to back it up) that the variable caused the outcome.  

• The third thing is my favorite part; I call it the ‘foolishness factor’. It’s like the first two things are present, but does it actually make sense? Or, simply put, are you talking foolishness?

So, yes there is a relationship between the weight and gestational periods of mammals. Humans: 9 months. Whales 16 months. Elephants: 22 months. Giraffes: 15 months. Correlation coefficient is .2689. But this doesn’t mean that the nice lady I saw last week on My 600 lb Life, is going to be pregnant for a year!

I mean, the velvet worm and the rhino have the same gestational period, but it would be foolish to conclude that the heavier you are, the longer you are pregnant!

Typical weakness for not determining cause and effect might be the timeline factor. Say for instance there are 100 people in jail. And 90 of them are practicing some religion. Don’t care which one. You may be able to conclude that there is a relationship between, say, being religious and being incarcerated. But unless you know when that person converted to said religion, it might not actually show cause and effect.

Jeez, why am I thinking about the Q-doggs, my friend Mike and my college days?

A quasi-experiment is kind of an experiment, but kind of not. It estimates the cause of an intervention on a certain population, but without randomness. Got to think twice about this one, because who knows where they started? Was it at the same point? The training group and test group should have the same starting point, so we can tell what the actual cause was.

Non-experimental research is the researcher trying to predict the outcome without actually doing an experiment. He is unable to manipulate the variable in the non-study. Exactly! What are you really doing??? You’re not even experimenting: hence the name. So the researcher must come to his very informal conclusion by some sort of correlation (not scientific) or case study, or some other way to fabricate a relationship or conclusion.

Experimental research follows the scientific method, with a question needing to be answered, variables that can be measured, calculated and compared. The most important thing is that it’s done in a controlled environment. You will have a test group and a training group, where one group gets something the other does not. To me, this is the right way to do things.