I was talking with a friend today about why research on prospective graduate students is so difficult and costly. Clearly there is a great need among marketers of graduate programs to understand what motivates their primary target audience. And, Deans often want to know what NEW programs should be offered to fill a niche. Gathering this type of information is complicated because there is no dominant and high quality sample source; it's very difficult to secure a representative sample of people who want to earn a graduate degree in psychology, for example. Where would you buy that list? 

Unlike the undergraduate level, where the names and contact information of high school seniors are readily available through a variety of sources, records of adults who are considering graduate school are difficult to come by. If you want to survey the general population of adults interested in a particular program, you have to buy a list and screen to find the people you want, e.g. adults who are seriously considering going back to school to earn a DNP in the next three years. You can often buy lists that zero in on your target audience but there is still typically a great deal of screening involved, and here's where the costly part comes in.

Let's say you find a reasonable list source for adults who could possibly be interested in your MS in Civil Engineering. Before you can survey them, you'll have to screen to make sure they are *really* a prospective student of your program. If 80% of the folks you screen meet your criteria and qualify to participate in your study, you are doing great. But, if only 20% do, you are going to pay a LOT more for your study and it may even be too costly for you to conduct.

Sorry, I am not offering any great and wise solutions here. Just pointing out the problem and venting a bit with those of you out there who share my pain. Research with prospective graduate students is not impossible, and some programs are easier than others. But, generally speaking it's rather costly, time consuming, and a lot of error is involved. Not my favorite combination.