Please see
www.visionfromfeeling.com/study.html with links to plenty of drafts of the study procedure. But here it is:
Claimant: me
Participant: persons who assist me in the practical assignments of the study
Volunteer: person from the public who fills in the questionnaire and is seen by the claimant and one or two participants
The study will require two separate locations that are out of sight and sound of each other. Volunteers will begin at the first location, where one participant is stationed and hands them the health questionnaire to be filled in. The questionnaire also comes with an information page that describes the study, what is involved in volunteering, and how to fill in the questionnaire. The participant is also available to discuss the study with the volunteer and also to answer any questions that the volunteers might have. The participant holds on to the filled in volunteer's health questionnaires.
Once the volunteer has finished with the questionnaire and handed it in, a second participant takes the volunteer to the second location. Once the volunteer is seated and settled in, the claimant and one or two additional participants will spend up to ten or fifteen minutes to look at the volunteer and fill in our own health questionnaires based on our impressions of the health of the person. The volunteer may of course leave early if they wish. The reason I propose that the volunteers be taken to a second location is because many volunteers might gather at the first location and should not be seen early by the claimant or the one or two participants.
There is no speaking or other interaction between the claimant and the one or two participants and with the volunteer. The volunteer will be seen from the back and no eyecontact is possible. The health questionnaires of the claimant and one or two participants who also do the viewing are then handed in preferrably to an additional participant who is stationed at this second location.
So the study would hope to involve five participants but can do with fewer. At the end of the study, health questionnaires pertaining to one particular volunteer are matched based on the identification numbers printed on their top margins and the correlation is determined by the participants though the claimant has the right to be present. The volunteers remain entirely anonymous with regard to their identity. I am even considering having the identification numbers concealed during the study with respect to the privacy of the volunteers so that "no one can remember the ID number and what person it pertained to".
The one or two participants who also see the volunteers act as controls and are encouraged to try any trick that they can think of to acchieve a higher correlation. They can try any method of guessing or any method of cheating that they can get away with, because, we want to know if I am doing such things and whether my correlation would be higher.
The study is not a test. If it were intended as a test there would be plenty of issues with its quality. But it serves the purpose of a study very well. I will gain additional experience with the claimed medical perceptions, get to find out what ailments I am better at identifying than others and hopefully good ones could be selected for a test. How many times when a particular ailment occurs do I detect it and at what percentage do I not detect it, so that we know at least how many subjects with such a particular ailment would be required for a real test. And we can get something of an estimate as to whether my correlation seems to be higher than that acchieved by the one or two participants. Is my correlation (ie. accuracy) high enough to be interesting for further investigation? These are just some of the valuable things I'd hope to learn from taking the study.
The study is not very quantitative, there is no specific point-scale system. It will produce rough estimates, but from that I will then design an improved 2nd study which builds on what was learned from the 1st one and is closer to what an actual test would be like. Provided that the claimed ability is not falsified under this first study. There is no specific point-scale system that would determine a falsified claim and a non-ability might pass through to a second study, but would be caught eventually.
I do believe that this study is an appropriate step toward a real test and toward a real conclusion in this investigation. I know that some of you won't think so.

But I'm working on it.
...
The volunteers fill in a health questionnaire on their own and their answers should not be known by any of the participants or the claimant as their health questionnaire is handed in and put aside until after the study is completed. Their answers should remain anonymous.
I will be nearby a second location. The volunteer fills in the questionnaire at the first location and is then taken to the second location and I do not see the volunteer until they are seated and settled which is when I and the one or two participants who will attempt to do what I do arrive at the second location. So, we are physically separated by being in separate locations.
I need a minimum of two participants, one to assist the volunteers in filling out the questionnaires and one to fill in health questionnaires alongside me so that we have a control to compare my correlation with.
I am working on finding participants for the study. I might engage university students. (Ones who I do not know prior.) If they are university students I might not need to mention anything about a paranormal investigation, since after all at this point nothing paranormal has been established. For all we know it is a study in detection of health problems based on external symptoms. And even if it were the case of extrasensory perception, such does not need to be described as paranormal.