Since I sometimes get up to 10 recommendation letter requests for a given week (seriously, this is the maximum; the average is more like 2 but if I get 5 or more it just kills me), and since the general practice in the group is to draft your own that I then edit, I thought I'd write up some tips on how you should write your own (or others') recommendation letters. This is how I do it and it's not always the same but it's a general structure.
- Provide a comprehensive draft to your recommender. The reason to write your own letter is that eventually in your career you will (hopefully) reach a place where the person who writes a letter on your behalf (chair, dean, president, boss, whatever) doesn't really know you or has not kept up with your research. At this stage, you have to write this letter pretty much pretending you're them otherwise you most likely risk a poor recommendation. And please do not be shy! The worst thing that can happen is that I'll say "no". But in general, I want to support you as much as I can. I understand that in most cases, when these requests are made, you're deep in the process of doing whatever it is that required you to have recommendation letters. I'm sorry about the additional stress, but these are the only opportunities you get to write you own letters which will be absolutely necessary in the future and an extremely important skill to develop. (And believe me, it's a lot more work to edit your letters than for me to write them.)
- The better the initial draft, the better the recommendation will be in the end. In general, assume the recommender has little time and will do their best to edit and modify (in my case at least) but the better the starting point, the better the output.
- When you ask for a letter, provide as much advance notice as possible, and make sure you inform and request a due by date (usually a day before it's actually due). Feel free to bug the recommender if they're not delivering since they might be busy (and you know what happens with busy people). At the same time, in my experience, we all provide some leeway for recommendations being late.
- A comprehensive draft in general would start out with exactly who the letter should be addressed to and the salutation. Asking for a letter and being vague or just pointing to a URL is inadequate.
- The first paragraphs are usually introducing the recommender to the committee that is reviewing for admission/award/whatever. It usually sets the context for the evaluation. For example, "I've mentored 50 students and this person ranks 1/50." This is usually something you don't need to provide in a draft to me, but you generally would need to do this if you're writing one.
- The next paragraphs are setting up how you know each other, and in what context, and for how long. This is where the recommendee can do a good job in assisting the recommender.
- The next paragraphs are setting up the specifics of the association. Here I personally feel details are important. For example, describing the research you performed with references in some detail (even if it means obfuscating the readers) is probably not a bad idea. This is where the recommendee can do a good job in assisting the recommender.
- Then it's usually important but not science related stuff as to how well you work with others, content about extra curricular activities, honours and awards, and so on. This is where the recommendee can do a good job in assisting the recommender especially with regards to accomplishments the recommender is not familiar about.
- And then a final concluding paragraph that again justifies your value for whatever it is that you're asking and how it'll help you advance your career.
Specifics to our group
- In general, if I send you LaTeX source in ASCII, I want it edited back in ASCII without removing the LaTeX formatting commands and in a Unix text file format. If this means learning how to use a text editor on our Unix systems, then all the better. This is going to ensure the fastest turnaround.
- I will put up some good examples if people agree to have theirs publicised, like Juni's.
- Aaron Goldman from Dec 2007 for a NSF Graduate Research Fellowship. Aaron is also the recipient of an NSF IGERT award. Aaron won the Whiteley Award which is given to the best graduate student for his year is now a postdoctoral fellow with another NSF fellowship at Princeton. Aaron graduated in four years (I believe this applies to all my graduate students as of 2012.)
- Jeremy Horst from Aug 2005 for a Ruth L. Kirschstein National Research Service Award (NRSA). Jeremy obtained the award on his first try which funds ALL his DDS and PhD studies (the first UW DDS/PhD student). Jeremy also won the Magnusson Scholar Award which is one of the highest awards given annually by the University to graduate students. Jeremy also graduated in four years and is now a pediatric resident and a postdoctoral fellow at UCSF.
- The inconsistency of the date is because I obviously must've rerun LaTeX on it later and the text actually represents a resubmission which at the last minute was unnecessary (so the original letter was somewhat weaker, but this example illustrates not only how to write a letter but also how to write in response to critiques of an application, which indicates an active mentor/mentee relationship).
Academia vs. industry
This is my personal biased view, but it reflects a large amount of reality based on observations of many many people who've gone through our group. A fundamental assumption is that you wish to contribute to the body of humanity's knowledge and wisdom (or lack thereof), rather than just finding a "job" in which case, specifically to our group's skill sets, there are much better ways.
- The best science graduate school educations consist of teaching you be an independent thinker, which generally contradicts what industry at this level expects from you: to be a cog in a machine. The bigger the machine (Pfizer, Merck, etc.), the more of a cog you'll be. Even the best and brightest academics at all levels (including full professors) with the best reputations are in the end cogs since industry is about a business and not science. But the cog factor declines as you advance more in your career, mainly through gaining experience. In contrast, however, even though the tradition of being a postdoc is relatively new and sometimes not necessary in non life science fields, being a faculty member or a funded postdoc gives you a lot of autonomy.
- A corollary to this is that in general it is good to be self funded no matter what stage of your career you're in, or how independently wealthy you are, since a point is to achieve progress in science.
- Graduate education in general has ill prepared you for industry positions and interviewing. We have had enough people here in our group associated with us with industry positions to talk about their experiences and I'd encourage you to talk with them about it.
- A corollary to this is to use all the resources you have and network as much as you can throughout your graduate education. Your best resource is your mentor, who depending on their own career stage, may have different motivations and will in general become busier over time. Personally speaking, there was a time when my peers in industry were recruiting when I started off as a faculty member and it was easy to recommend people. Now the inertial barrier is higher and generally speaking, there's no reason to be shy to ask your mentor to do something. The worst case scenario is that they do nothing (which is possible) so having a backup plan is always a good option. In general, the likelihood of someone doing something for you (in life in general) increases the more you make the task easier for them (see section on writing recommendation letters above :).
- There is a growing relationship between academia and industry among the more progressive companies, and the best industry positions are those that encourage the creative and independent thinking you've developed in graduate school. An example of this would be Rosetta (which is a fully owned subsidiary of Merck). Merck has decided in this case that Rosetta can act as an autonomous unit, and people at Rosetta generally have a lot of freedoms. In general, the UW life sciences departments have had good relationships with Rosetta and there's a lot of cross talk which helps with positions.
- Faculty positions by definition are hard to obtain and maintain since there are far more qualified candidates than positions and this will almost always be the case. In industry, there are "hot" and "cold" areas at any given moment, and even in bad economic climates, I've seen people go from startup to startup their entire life (sometimes changing jobs within months) since people (read venture capitalists) are always looking to make money.