TITLE: Beginner's questions on the post-doc application
QUESTION [30 upvotes]: I am going on the (research oriented) post-doc job market (mainly in the US) this winter and would like to ask some questions. I am aware of the site http://academia.stackexchange.com, but I am writing here because want more math-specific answers. 
Two beginner's questions that were not previously asked in MO are:

Can I send more than what employer institutions ask for? For example, if your teaching evaluation of your previous class looks great, can I include it even if they won't require? Does extra reference letters hurt my application? A reason to ask this is that I am afraid of my application to be overlooked among hundreds of applications. I want to make a difference, but shouldn't I take any risk?
What is the difference between "publication list" and "publication list section in CV"? Should the publication list be more than a mere list? Include abstracts? A reason to ask this  is that I have only a few preprints and the list looks a bit miserable. 

These days one post-doc position often receives a very large number applications, and it is not likely that the hiring committee read all the applications. What filter is there in the first round? Names of your school, advisor and letter writers? 
Thank you very much for your advice.

REPLY [32 votes]: I've served on the postdoc hiring committee at Rice for several years and am chair of it this year.  My answers are geared to research-oriented postdoc applications.
For question (1), it is fine to include extra letters of recommendation, but don't go nuts.  We ask for 3 including a teaching letter, but most of our applicants include 3-4 research letters plus a teaching letter.  A letter from someone that doesn't know you very well is damaging to your application (so don't sacrifice quality for quantity), and if you get more than 4 research letters and 1 teaching letter then it starts to look weird.
However, don't include things like teaching evaluations we don't ask for.  We don't read them and make an effort not to let them influence our decisions, but it does look a little weird.
For question (2), just include a list of publications excerpted from your cv (to emphasize, this list should also be in your cv).  To be honest, I'm not really sure why we ask for a separate list of publications and we don't really look at it; probably it is there because of some decision made by a hiring committee long ago, and inertia is a powerful force.
As far as your unnumbered question, a member of our committee does actually read every application.  The first read (to form a long list of people who are read more carefully) is quick, but it does happen.  The most important things we look at in the first round are research area (to make sure it is at least vaguely related to someone in the department) and the research letters (both who they are from and what they say).

REPLY [11 votes]: In addition to other useful remarks: 
The first filter is your "pedigree", which is PhD school and/or post-doc institution, advisor, and letter-writers.
The length of your research (and teaching) statement(s) can be whatever you want... however, in those statements you are "selling yourself" in the face of competition from many other roughly comparable applicants. The point is that a boring or tedious or too-technical research statement will "put off" hiring-committee people in other fields, so they won't read to the end, or even past the first page. Thus, the research statement has to "sell" you both to specialists and to non-specialists, since, after all, there's probably just one or two people on the hiring committee who could appreciate the nuances of your work. Even though it typically happens only marginally, thinking in terms of "broader appeal" can be very helpful. Thus, in effect, the _first_page_ of your research statement should have optimized visual impact, and be an artful combination of info-for-experts, and of overview-and-broader-significance for non-experts.
On similar principles, there should be an immediately-visible version of publication list that lets both specialists and disinterested-non-specialists see "numbers" and "journals" at a glance. Including abstracts and further information is fine, on subsequent pages, but bear in mind that many people will not look past the first page, so that's your chance to make an impact.
Also, grants, fellowships, honors have an impact on "disinterested non-specialists", since these are intelligible. Get that on the first page of CV.
And, disagreeing with some other remarks, and agreeing with some: avoid lukewarm letters at all costs. Hiring committees do not "average" the letters. Rather, even if you have two glowing letters, a single negative or even ambivalent letter can be fatal. The reason is that, in effect, there are so many applicants that there is some incentive to find a reason to eliminate as many candidates as possible, before looking at the pluses of the remaining ones. Thus, a "blemish" can be a target for hiring committee members in other fields, who might be happier to hire someone with interests closer to theirs anyway.

REPLY [6 votes]: Let me start with (2) since fewer people have addressed this already.  I certainly have wondered about it the few times I've been on the job market.  However, I've also seen many applications at my institution.  I think most people:
Include the publication list in the CV and in the separate document.
I don't think it is necessary to include abstracts in the publication list (some people do, most people don't).  The main part of this data presumably should also be in your research statement.
For (1), having extra letters of reference do not hurt.  Especially for tenure track jobs at big schools (like Penn State) having the minimum is actually very unusual.  
On the other hand, it seems unusual to include things that are not asked for.