Determine the kind of observational study that you will use. A single-blind study includes patients who are unaware if they are receiving the real treatment, whereas a double-blind study comprises patients and doctors who are unaware of who is receiving the real treatment. Prepare for your specific study. For instance, in a double-blind medical observation study, you will divide your medicines and placebos by random patient numbers; you must ensure that each patient has a sufficient quantity of his assigned treatment.
Decide which method of separation you wish to use to divide your test subjects. The randomization method assigns patients to groups randomly, whereas the match method links patients by specific factors, assigning one randomly to the control group and assigning the other randomly to the test group.
Separate your test subjects into two groups: your test group and your control group. As an example, you decide to use the match method for your medical study and link patients into groups of two by similar cholesterol levels. In each group of two, you randomly assign one to the test group and the other to the control group. For double-blind studies, remember to keep the researchers unaware of which patient receives the real medicine.
Assign each subject a subject number. Use this number to identify patients during the testing process. For double-blind studies, instruct a secondary team to assign the subject numbers to each individual. For instance, in your medical study, nurses may handle the assignment of patient numbers and the distribution of medicine, although your nursing staff should not be made aware of which patients are receiving the real treatment and which patients are receiving the placebo.
Record all information about your test subjects by their subject number --- including notes on patient's needs, medical changes in subjects or important information about subject's complaints or discomforts. Bring your information together after your study and evaluate your results. As an example, after your study, your researchers can look at patient numbers, use the number to determine which treatment the patient received and then make a determination about how the patient reacted, finally knowing whether the patient received the treatment or the placebo.