Relationships between variables

In the world of social sciences, research is usually intended to examine the relationship between two variables. For example, does having strict parents cause children to be more rebellious when they're older? Or does having a father leave the child when it's young cause it to suffer from self esteem issues? These are the type of answers behavioral scientists seek to answer when they conduct research. There are two distinct data structures that are used when conducting research: the correlational method and the experimental method.

The correlational method

Using the correlational method involves simply observing two variables as they exist naturally and comparing them to each other. For example, If I wanted to figure out wether kids that come from single parent homes are more likely to have good grades than kids with both their parents. Using this method, I would have a sample of kids take a survey which includes questions about their grades and their parents. I would then observe the data and look for a relationship between the two variables. It's often easier to see patterns in the data when it's presented in a graph, like the one below.

This chart shows a correlation between children being born into poverty and the likelihood they will be top earners by age 30

The results from a correlational study can only show you a correlational relationship between two variables and never a causal one. That is, you can never know that the true reason for why kids from single parent homes get lower grades is that they come from single parent homes. We can only know that there is a correlation between the two. There are many possible explanations for why kids from single parent homes on average have worse grades than kids from dual parent homes. It could be that kids from single parent homes tend to come from impoverished communities. It could also be that these kids tend to come from families who don't place much value on education. The point is, we don't know for sure and can never truly know when using the correlational method.

The experimental method

The experimental method uses two groups with unique variables assigned to them. For example, if you wanted to figure out whether classical music helps people study you would take one group and have them listen to classical music while they study a given topic and another group not listen to anything while they study the exact same topic. The only thing that should be different, or variable, between these two groups is that one group is listening to classical music and the other isn't. After both groups are done studying, you could then give both groups a test on the topic (again, both groups would have to be treated exactly the same). We would then grade their tests and compare average tests scores between the two groups. This is the experimental method.

To be able to say that it's the classical music causing this difference in how much information is being consumed, the researcher must control all other variables that aren't classical music. Not only must the researcher make sure that environmental variables are controlled but also participant variables. This means that researchers must make sure that one out of the two groups doesn't contain too much of one kind of person. For example, if group A consists of mostly college graduates whereas group B consists mostly of high school drop outs, the study would be confounded because it would be impossible to come to a clear conclusion.

Independent vs dependent variables

Using the same example, the independent variable in this study would be the classical music playing and the dependent variable would be the scores on their tests. An easy way to remember the difference between the independent variable and the dependent variable is by their definitions. Something that's independent stands on its own, whereas something that's dependent relies on something or someone else. A dependent variable relies on an independent variable before it can be measures. We cannot the test scores, or the dependent variable, without first introducing a group of people to classical music while they study for the test and comparing their scores to the scores of a group that wasn't introduced to classical music while they study. The independent variable is the variable that's manipulated and usually consists of two treatment conditions. The dependent variable is what's observed as the effect of that manipulation.

Control groups vs experimental groups

When conducting research using the experimental method, controlling variables is key. In the example we've been using, the control condition would be the group of people studying without any music in the background. The control condition is basically when individuals do not receive the experimental treatment and instead receive no treatment, a neutral treatment or a placebo treatment. This specific group is called the control group. The purpose of the control group is to provide a comparison to the experimental condition.

The experimental condition is essentially the opposite of the control condition. The experimental group does receive the experimental condition. The purpose of this group is to observe how the experimental condition affects individuals vs those in the control condition. For example, a group that's given a certain experimental drug is labeled as the experimental group whereas the group that's given the placebo is labeled as the control group.