First of all, would just any person recently diagnosed with diabetes be a candidate for this experiment? Of course not. There could be any number of conditions, dependent on the local culture in which you practice; however, for the time being let's keep it simple and just say that we want only adolescents, male or female, who have recently been diagnosed with diabetes.
Theoretically, this creates a very large population that you should draw from randomly in order to select a pool of participants to be assigned, randomly, to experimental and control groups. Again, to keep things simple, let's say that we are only interested in the persons that come to the offices where you work. You may even be able to include all of them, instead of a randomly selected subset, for your experiment.
Regardless of whether you select a random sample or every single patient, and assuming for the moment that your experiment indicates success for the new technique, can you confidently recommend it to a colleague in another state or city?
Yes No