| Even if I'm super happy with someone's performance, a bullet list of the achievements they want to highlight is extremely useful. Heck, I have a hard time remembering exactly what I accomplished in the last year, it's super risky to assume your recommender will remember for you or happen to mention what you think is important. It's like... I keep a bullet list of things I've accomplished at work and add to it when I feel like I did something worth highlighting as especially helpful. Then during my performance review I've got more in active memory than to discuss what I did last week. Plus your recommender won't know the context of how you're pitching your application. They don't know the other recommenders you picked or what story you want to tell, even if they have the luxury of enough free time to write something personal. As a recommender, the one that I find most obnoxious is that they usually ask you to "rate people relative to others you've worked with" which is just a ridiculous question. I'm at MIT, if I find someone seems to me like they'd fit in as a first or second year grad student at MIT but maybe in the lower half, do you actually want me to write down that they're below average? It's pretty clear that it is not the question you intended to ask when you don't bother to even ask me what kinds of people I've worked with before. You can't ask people to numerically rank performance when there's never been an agreement on the scale! And knowing that you're likely using it as a first pass filter makes it even worse; what, are you going to somehow correct for the difference between the people I've worked with versus everyone else according to some magic objective scale? Even if you know that I'm comparing to MIT students, do you realize MIT has 10000+ students with skill levels all over the map? Do you know that I value practical skills more than most because I've seen what makes someone successful in lab? It's just a part of the recommendation letters I wish they'd just leave out. Filter based on GREs or something else that while equally pointless is at least uniform and then just read the letter and think what you want. Don't ask me to do half the thinking so that you can do none. |