The first thing that strikes me is that the graduate admission process was expensive. GRE subject test fees ($130 per), GRE general test fees ($140), an extra $20 per score report, $12 to get your scores by phone. Then you have fees to send official transcripts, and then actual application fees - usually about $50 for your application. For us poor applicants, getting into a graduate school is worth well over the one to three thousand dollars we can expect to spend on the process, but it still feels like we're getting gouged.
To say nothing of the time invested in essays, forms, letters of recommendation, et cetera, that we would rather be spending on research or coursework. I have a short list of pet peeves coming out of this that many graduate program directors - or graduate school administrators, in some cases - might be well advised to read.
- Ask for one official score report and one official transcript. Photocopies are much cheaper than official reports (pennies vs dollars, even accounting for labor). What is the application fee supposed to cover, anyway?
- Exorbitant applications fees will shrink the size of your applications pool, but won't improve its quality. Or socioeconomic diversity, come to that.
- Have one application form that can be filled out completely online. Embark and Applyweb are good, although they could be improved. Having your own online application isn't bad. Having multiple independent online forms, or requiring some materials online and some via snail-mail? Please.
The second thing that comes to my mind is the game-theoretic disaster that the admissions process is. Look at - say - physics. Many top-half programs' rejection rates closing in on 90% (or, in a few cases, higher) rejection rates for many reputable doctoral programs, you have to apply to multiple schools if you want to go on with your career.
So let me just say that in physics, the subject GRE is required by almost all top-half schools - which are, naturally, more popular places to apply than bottom-half schools. Yet less than 2500 people take the physics GRE per year (ref), while around 3000 students start graduate work every year in physics (ref).
Those of you doing math on your napkins may have noticed something: If the top-half programs reject an overwhelming majority of all applicants (around 90% in many cases), and the overwhelming majority of applicants get in somewhere, then most applicants are sending out a great number of applications (somewhere around ... ten, you might guess).
Students have an incentive to apply to more schools in order to increase their odds of getting in somewhere. With individual rejection rates like those, your application is a crapshoot even if you have impresive qualifications. End result? Hundreds of millions of dollars and millions of pages of forms wasted simply in the process of trying to allocate graduate students to schools, in a giant orgy of statistical noise. There has got to be a better way to do this.