The topic this article is on is
correlation research. This kind of research is used when determine how or if
two or more variables are related in a single group of people. The difference
between correlation research is that there is no attempt to manipulate the
variables at all and it is done through observation and random variables. In
the article, the difference between correlation and experimental was discussed
in detail.
In experimental
research, the variables are manipulated in some way and then the effects are measured.
An example that was given was showing the relation between cholesterol and
blood pressure readings. In an experimental research study, they would somehow
manipulate one variable to see the outcome of the other. This can show that if
there is a change in one then that could be the reason for the change in the
other. Looking at the correlation research, no manipulation would take place,
just measurements would be made and the outcome would be as is. This kind of data
can only be interpreted but cannot prove anything.
After reading
on this, why would anyone want to use correlation research over experimental research?
What is the point of interpretation versus an actual proven statement? Well,
correlation has a direction that can easily be seen when the results are
plotted. A positive correlation shows if someone scores above or below the
average on one measure then it should be similar when looking at the other. A
correlation can also show the strength of the relationship between the two
things being measured from the range of 1 to -1, where 0 is no relationship, or
uncorrelated, and 1 and -1 having perfect relationship. The relationship can be
seen anywhere between this range and this is quick way to show if there is a
relationship at all without having to go into the depths of performing an experiment.
Correlations
are important because they describe a relationship that things have in common
with each other. The major take away is that correlations do not prove a cause
and effect like an experiment can do. There are different kinds of relationships
such as positive, negative, linear, and nonlinear. These can all be shown when
graphed appropriately and is an easy way to visualize if there is a relationship
between things in a different way. Correlation research can easily be performed
in the class room and can help teach how different things can relate to others
and how much, if any relationship is there.
Hi Anne! After reading your blog post I see you answered your own question but let me add to it. In my opinion, since I'm not a scientist, correlations do indeed show a relationship between what you're testing. We know that. But just like theories we need to test them multiple times in order to create an experiment. When I was getting my undergrad degree in psychology, I never realized the amount of research that one has to do in order to create an experiment and then attempt to have it published. It's a lot of work. There were many people who were research assistants and their job was to do research. I don't know about you but when I research things for classes I usually get tired after a couple hours of finding articles. People do that for a career. Nevertheless, people will do many hours of testing in order to find the correlation between two variables and then repeat that experiment with other variables. This all turns out to be experimental. You know how kids use to mix their pops together? They make predictions and see if their hypothesis is correct and so on.
ReplyDelete