I can understand your frustration because I till this day don't know the various types of studies (it can take various forms). First and foremost, I suggest you find an epidemiology class and take it because it will help you answer a lot of these questions.
Clinical trial and retrospective analysis fall into two separate categories. The first is experimental and the second is observational, respectively.
Clinical trials are where the experimenter is involved with the trial. For example, the experimenter will give a drug X to a randomized group of people (the people who get drug X are the subject group and the people who get the placebo are the control group). So because he or she is giving drug X, they are involved in the trial.
Retrospective studies (I hope this is what you mean by analysis) are under the broad category of observational studies. This means the experimenter is not involved with the study. For example, a study that looks at the risk factors for Parkinson's disease is where the observer just notes Parkinson's and the risk factors. The experimenter does not give any drugs or intervention in the study (just data collection).
Now this is the part that will be difficult for you to understand. There are two types of retrospective studies. The first are case-control studies. This is where the outcome is known, but the risk factors is unknown. For example, Parkinson's disease is the outcome, but you do a case control study where look for the risk factors that cause the disease. You find that smoking might be a risk factor. This is an example of case-control study.
The second are retrospective cohort studies. This is where the risk factor is known, but the outcome is unknown. Let's say you want to look at alcoholism and Parkinson's. You look for data with these two aspects (remember we are look at . Having your subject group with known alcoholism and the control group that doesn't drink much. Then you map out the percentages of both who develop Parkinson's. This is an example of a retrospective cohort study.
It is hard to understand this, so I suggest you look up journal articles or educational websites explaining the differences. Good luck.
