BIOL 100 Experimental Design Basics
Experimental Design: Hypothesis, Variables, and Groups
When using the scientific method, we get to a point where we formulate a hypothesis and then design an experiment to test that hypothesis. Here are some pointers on this process.
The hypothesis states a special relationship between an independent variable (X) and a dependent variable (Y). A generic hypothesis is a statement that looks like "If I do X, I will see a measurable change in Y." Basically, the hypothesis looks like a statement of “cause” and “effect”. Your hypothesis is a direct statement and is not written in the form. of a question.
Remember that the independent variable is "what you do", the dependent variable is "what you measure".
Let’s use a drug study to create a specific example of what a hypothesis should look like, and to see good examples of independent and dependent variables. Drug studies are usually straightforward.
A pharmaceutical company designed a drug (drug X) to treat migraine headaches (MH).
For this experiment, the hypothesis is “If I give drug X, patients will experience a reduction in migraine headaches (Y= # of migraine headaches/month)."
Drug X is the independent variable (what you do or give in the experiment), and the number of MH’s/month is the dependent variable (what you measure, where you see the result in the experiment).
Subjects (participants) in experiments are organized into either a “control” or an “experimental” group. The subjects should be randomly assigned to one or the other.
First, we need a large pool of people who suffer with MH, and we randomly assign these people into two groups. These groups should be IDENTICAL in every way- average age, same number of male and female, same average number of MH's per month, same exercise levels, etc. All these variables that are equal – age, health, diet, are called “controlled variables”.
One group will remain the same throughout the experiment - the control group. They get placebo pills. The other group will get drug X, and they become the experimental group.
Drug X is the ONLY variable that now makes the 2 groups different. This is important, for when we measure the average number of MH's per month in both groups, and we see a difference in the experimental group, only one variable, (X), can be associated with that difference.
That leads us back to the hypothesis, "If we introduce X, we will see a change in Y."