Select a focus. Pick a topic that is of interest to you to increase the likelihood that you enjoy the experimentation process. Try engaging options like testing a potential recall enhancement method, measuring the impact of vivid images of natural disasters on human emotion or testing the degree to which individuals relate to stimuli, such as the sight of the dining hall, to a task, like eating dinner. When you select your topic, pause and consider whether you will be able to create an experiment on this topic with the means at your disposal. If not, reselect.
Contact the Human Subjects Review Board. If your experiment will involve human subjects, you will likely have to register it with the Human Subjects Review Board (HSRB) at the institution you attend or work at. This board is generally under the Office of Research and Compliance, but at smaller colleges or universities, it may be part of the science or psychology department itself. Check with your psychology professor or administrators within the psychology department for specifics as to the HSRB application process at your school. Take care of this early in the process to ensure that your experiment is not held up by lack of HSRB approval. If you are using human subjects, select a facility that is sizable enough to accommodate all of your participants if you plan to have more than one subject testing at the same time.
Compose a hypothesis. Write your hypothesis down, as you will need to reference it later.
Develop a method. Plan out a step-by-step method for completing your experiment. After drafting this method, review it carefully in an attempt to identify potential problems. Correct any you find.
Craft an experiment schedule. If human subjects are necessary for your experiment, consider these individual's schedules to ensure that they are available when need be. Also, take into consideration the impact that the time of day could have on your experiment. For example, if your experiment involves a complex task for which subjects must be wide awake, an early morning experiment schedule may not prove the best choice. If your experiment is for a class, or you are trying to meet a publication submission deadline, keep this deadline in mind when crafting your schedule to ensure that you complete the experiment during the allotted time.
Include a control. For your experiment to be scientific, you must have a control group. For example, if you are going to test the impact of listening to classical music while studying on exam grades, you can't have all of your subjects listen to music while they study, so you must instead create a control group that does not listen to the selected tunes.
Select a data collection method. If your experiment will yield numerical results, data collection will likely prove simple. If not, you may have to depend at least, in part, upon observation. If your experiment will involve observation, create a rubric to guide this observation.