Log in

Previous Entry | Next Entry

What do you think of this? Too harsh?

I've seen a drop in the number of applicant queries lately, which could be due to the above (a good thing, IMO) or due to graduate students telling their friends that I'm not hiring lately. (I haven't been, but it's because there haven't been Ph.D. students in my exact area, not because I have no research assistantships.)

Ben Perry and Julie Thornton, two of my students at the time, told me at Edmonton in 2002 that they read it and would have been afraid to apply for research assistantships if they'd seen it - which isn't the idea, given their caliber.

The problem is that I often get e-mail such as this:

From: [name omitted]_i20
Date: Fri Nov 4, 2005 3:24 am

Hey ppl,
I'm [name omitted], a prospective student 2 fall 2006. My profile goes like this :

Btech. - final yr in Electronics & Instrumentation engg frm a college
affiliated 2 [university omitted], [city omitted].
B.Tech. % - 58% till 3rd yr 2nd sem with 1 uncleared backlog.

GRE - 1220 (q-730, v-490, a-4.0)

TOEFL - 203 to 263

Extra cirriculur -
3 national level certificates 4 paper presentations
1 national level certficate 4 mini project presentation.
few other organising of national level symposia certificates.

Program I intend to join : Computer engineering / Computer science.

IS d school & the department good ??

What do u guys think are my chances of getting admitted in ur univ ?

Can I apply to the univ or is it too high 2 my profile ?

What is the current financial aid situation ?

Waiting 4 ur replies soon,
Thanking u,
[name omitted]

One of my CIS 730 (Intro to AI) students replied:

From: [name omitted]
Date: Mon Oct 31, 2005 12:12 pm
Subject: Re: [ksu-cis730-fall2005] Urgent help needed regarding admission

hey [name omitted],

dude, u should totally apply to this univ. it is gr8. i am a masters
student 74% til 8th quarter of my degree. i write all my assigns in
l337 so u would totally fit right in. i know that d dean of admin
totally scans this web group every day, so i m sure he'l contact u n e
day now.

c u l8 r,
[applicant's name spelled backward]

Names have been withheld to protect the innocent, or the clueless as the case may be. I took out the university and city to minimize undue influence on those of you who administer admissions at your respective universities. As for this particular student, such atrocious queries are all too common.

As you can see here on Dan Andresen's page and here in a Wayback Machine archive from Sanjoy Das's page, Cluebie Grad Applicant spam is very common. As one CIS professor asked, "Do these people have any idea what kind of impressions they are making with such e-mails?" Apparently, they don't know, or they don't care. Sometimes it is that they don't care: I've been told by our grad program director, who heard it from some of our students, that some universities in India (for example) are understaffed and have such high student-to-faculty ratios (low faculty-to-student ratios) that the only way to get attention and letters of recommendation is to bug the living daylights out of faculty until they give in. Apparently one negative ramification of this culture is what we are seeing.




( 31 comments — Leave a comment )
Nov. 10th, 2005 08:32 pm (UTC)
I've been told by our grad program director that some universities in India ... Apparently one negative ramification of this culture is what we are seeing.

Tis true.
Nov. 10th, 2005 08:47 pm (UTC)
Education in India and bioinformatics
Hi, prahlad. I forgot to mention that DAG actually heard it from some Indian students. And you're probably the third Indian student who's confirmed it to me, so I definiitely believe it.

BTW, I talked to Wallentine about Herman's request for a non-programming bioinformatics course, and we agree that while the following are possible:

  • 1. a non-algorithmic 500-level modeling course (packages such as TIGR TM4, RasMol, etc, with a vade mecum on using these; Grid computing and the BLAST-to-PubMed/PDB/GenBank pipeline; web services and SOAP; data models, data transformations; GO, minimal semistructured data: XML, markup languages)
  • 2. a dynamic programming-oriented 500-level course on biological sequence processing and BLAST, CLUSTALW, etc.

we also concluded that the "how to run stuff with zero coding" will not fit in the bioinformatics degree specialization (for CS majors, of course, or for biologiists). IMO, it could be a 1-hour service course or a non-credit short course (during the summer session or intersession) for life scientists interested in bioinformatics but not looking to be bioinformatics majors (which still means "developers" to me, to some nonzero extent). The question then is whether CIS should be teaching that at all.

CIS 111, the intro programming course for non-majors, would be the lead-in for the two 500-level courses. I think they need CIS 300 (Data Structures) as well, but we haven't decided yet whether to let CIS 111 students into CIS 300.

Nov. 10th, 2005 08:56 pm (UTC)
My personal philosophy on hiring/selecting is that I'd rather spend extra time filtering out unqualified applicants rather than turn away a potentially qualified candidate before I ever get the chance to meet them. So in that sense, I think your message may be a bit harsh (I'm particularly looking at the part where you say they have a 1% chance of making it). Of course I'm saying this with no idea of how bad your spam problem really is. Another issue is that, with that message, you're selecting against candidates who are skilled but haven't yet developed confidence in their skills - something that describes many with a freshly minted Bachelor's degree. I'm of the opinion that it's easier to develop confidence in a talented scientist than to install scientific talent in someone who's overconfident.

As for the letter...

I've got a theory that's been rolling around my head that's come from seeing examples like yours as well as my own dealings with persons educated in India. We had a CS grad student from India working for our group. He was very intellient and articulate in English when speaking to him in person. However, e-mailing or IM-ing him drove me nuts since it was all 1337-speak gibberish like you show above. My theory is that there's a cultural disconnect along the following (greatly simplified) lines:

+Really skilled hax0rz write in leet-speak.
+I want these people to think I'm a serious hacker d00d.
+Therefore, I will type in leet-speak to show that I'm in the club and know their jargon.

I've no evidence to support this beyond a collection of antecdotes, but it would help to explain the "brilliant person:horrific e-mails" disconnect.
Nov. 10th, 2005 09:09 pm (UTC)
d00d, u r0xx0r2.
Good grief, Charlie Brown!

I'm not against corresponding with people, but I had better have a question that I felt could not be answered by other avenues of research. (Oh say, the university financial aid website?!) Also, sending emails does not mean that a person should forfeit all claims to good grammar and decency.

This makes me sad.

Nov. 10th, 2005 09:34 pm (UTC)
I don't think it's too harsh at all. As a hiring manager, I would immediately delete any email that was even remotely similar to the one you reference above. In fact, even minor typos don't make it into my "for consideration" pile.

Here's why: An application letter, resume, introductory email, etc. is an applicant's one chance to shine. Theoretically, they will never work harder to impress me than they will while trying to get a job. Therefore, I can assume that all subsequent work and communication will receive a bit less attention than that initial document.

An email like that does not bode well for future performance. To send such a message, without knowing one's audience, demonstrates an appalling lack of judgment and absolutely no sense of appropriate behavior.

I'll concede that I do all of my hiring in a business, not academic, environment and I do recognize that there are differences. However, professionalism is universal. Assuming that you expect even a modicum of professional behavior from your staff, applications of this sort don't merit even a moment of your time.

Yours in curmudgeonliness,
-- Amanda.
Nov. 10th, 2005 10:45 pm (UTC)
Do you know of any empirical results on the efficacy of these sorts of hiring rules? Do you know of any companies that automatically scan incoming resumes and cover letters for common spelling errors and reject based on them? Thank you.
Nov. 10th, 2005 11:24 pm (UTC)
I don't, but I'd be very interested in such analysis. Since I'm not typically part of the corporate screening process, I don't have visibility into the general review that occurs before I get the first batch. I do know, however, that I've never been satisfied with the quality of screening performed by any of the companies for which I've hired staff.

Typically, I receive pre-screened resumes from HR or Recruiting, and then conduct an initial screen of my own in which I reject applicants with significant errors. I then do another quick scan to reject obviously poor matches that weren't caught by HR (spotty work history, poorly matched skillset or education, etc.) and return the rest to HR for a phone screen. Those who pass the phone screen proceed to a phone interview with me or a manager reporting to me, followed by a scheduled interview for the very few who pass that.

My personal hiring process is very focused on only bringing in those I have a reasonable expectation of hiring; I don't believe in putting people through a fairly stringent interview process if their odds of success are low.
Nov. 11th, 2005 01:32 am (UTC)
Are you mostly dealing with applicants who speak English as a primary language?
Nov. 11th, 2005 06:12 am (UTC)
Yep - and that sets my expectations accordingly. I do accommodate for language barriers, but the roles for which I hire all require excellent communication skills, both written and verbal, in English.
Nov. 11th, 2005 06:13 am (UTC)
That was me. Apparently I'm tough on language but tolerant of failure to log in. :)
Nov. 11th, 2005 06:26 am (UTC)
No problem
Anonymous comments are fine; I reserve the cans of Whup-Ass for the anonymous trolls. :-D

Nov. 10th, 2005 10:09 pm (UTC)
I don't think it's too harsh.

Of course teh 10% comment only makes me want to send it 9 more times if it's unanswered ;)
(Deleted comment)
Nov. 11th, 2005 01:20 am (UTC)
Thanks - excellent advice, part 1 of 2
First, let me say that your suggestions are great, and though my internal calibration differs with yours in some respects, I will follow them all.

Lose the annoyed tone in the first section. Get rid of the "I respond to fewer than 10%..." sentence, for example; it does nothing but make you come across as a curmudgeon.
It's true (or it was), but you're right, I don't have to quote stats as a prohibitive.

Lose the "if you believe you fit this description" clause. That's a dead turn-off for students who can cut the mustard but don't yet believe they can.
That's a double-edged sword. I want to promote self-confidence, and you wouldn't believe the ridiculous number of students who come to us asking (literally) "I'm not qualified, and I know it, but will you take me anyway?" It's insulting, and it makes me wish US News and World Report would post more "typical admissions" demographics along with their rankings.

Yes, I realize that a conditional such as "an average admitted student had 700Q/590V/4.6A, as opposed to saying "an average applicant had 600Q/490V/4.1 GRE" is not 100% accurate and theoretically could deter good students. In practice, though, it probably turns away 1 otherwise-qualified student for every 50-100 students who have no business applying. Is that an acceptable loss rate? It depends, but I think in this case, it is, because "otherwise-qualified" is tempered by two things:

1. Do they have the right attitude for graduate school? The students I have seen go farthest are perseverant, enthusiastic, diligent, open-minded, and appropriately self-confident (neither self-deprecating nor self-aggrandizing). We get too many people who come because they have nothing better to do, which in general isn't a good reason.

2. Do they believe they belong here? Those who feel they don't really belong, or whose classmates are sure they don't, will languish. Those who are "angry to be here" because they applied to MIT and didn't get in, and are constantly asking for recommendations to the Top Five so they can "escape", are the single most toxic group of troublemakers I've seen at any university.

Lose the "This is not a well-formed question" and "Yes. Next question" and like that. It's condescending rather than funny, at least in this medium. If you want RAs and TAs who act like adults, treat them as such from the outset.
I do, but before I posted the above FAQ, those two were my number one questions (literally "frequently asked") among phone interviewees and new students asking for enrollment permission. Do I sound annoyed? You're right; I sure as hell was.

"How many people get TAs?" and "How many years does it typically take someone to graduate?" are reasonable questions in response to "Do you have any questions about our program?" I repeat that question until they say they have no more; then I move on to "Do you have any questions about the campus or town environment?" Now, I could understand if occasionally, they say "oh, I forgot to ask", but every time, I got the TA-ship question again. It just tells me that the students are either uniformly very slow on response (which again, I could understand, if it was just that English is not their first language), or not interested in any other aspect of this university. I fully realize that they can't come here without financial support; but if they aren't interested in anything else, then I can't communicate with them for purposes of assessment or recruiting.

It's as irksome to me when students cop an obvious "you're my safety university; I needn't act as if you have selective admissions" as it is when they say "I know my curriculum is completely below par for your program, but you're my last and only hope". All I'm saying is that people need to be serious about admissions, and show us either way (even though it'd be a "step down" or "step up") that they are realistically interested in us.

Nov. 11th, 2005 01:20 am (UTC)
Thanks - excellent advice, part 2 of 2
Let me tell you, the first time I said "4-7 years" to a prospective Ph.D. applicant and he literally yelled "SEVEN YEARS?!", I almost hung up the phone. Conversely, I also got a request for a letter of recommendation from a student who said, "I don't really want to go get a job yet, so I guess I should apply to grad school". He was honest and qualified, so I wrote him the letter, but I didn't feel so good about it. Hence the "if you're overly concerned with minimizing or maximizing that time".

As for "is your course hard?" - depending on how the question is asked, it could be legitimate, inappropriate, or downright insulting. Sometimes I get "elective refugees" (people who need an elective but are fleeing a course they perceive as harder) or people who want to study AI but want to skirt the hard parts. I want to disabuse them of the notion that AI, or any subfield of CS, might not be serious business, as much as I want to disabuse them of the notion that it's scary and forbidding.

Reverse the order of your want-to-know list to highly/moderately/slightly/not. Far more welcoming, far less condescending, far more useful to the kind of applicant you want.

Absolutely. Thanks!

I am impressed with the amount of real information on this page -- more grad schools and advisers should do pages like this.

Thanks. I took my cue from Sanjoy, Dan, and Marina Meila at UW.

All you really need to do is make it an expression of your concern for your students and students-to-be rather than your annoyance at people who will never become students anyway.

Good advice, but originally IRTA "people who will become students anyway", because seriously, that is the attitude of a lot of these applicants. (We can reject them, but once they're in, we are not so careful at vetting people for TA-ships; qualifications then matter more than any assessment of diligence, responsibility, or general student attitude, and it becomes a hazard.)

Nov. 11th, 2005 01:34 am (UTC)

How do I apply to ur sk00l?
Nov. 11th, 2005 01:48 am (UTC)
"I agree with the comment"?
LOL, funny. Now, you might think "WTH? He has to tell them such things?"

Sadly, yes. And some of them even pass our screening process, despite my being on the admissions committee. So, IMO, we need to raise our standards. Perhaps my missive isn't the best way; I will follow cavlec's suggestions in revamping it.

However, I do admit to feeling a teeny tad bit curmudgeonly about all of this. Details to follow in Part 2.

Nov. 11th, 2005 03:04 am (UTC)
Thank you, cavlec...
For making my comment shorter!

William, I agree with what cavlec said.

In addition, I agree with a comment that teremala made, about adding anchors or splitting the page into two or three separate pages (one for prospective students, one for new students).

Since cavlec and teremala addressed my main comments, I'm left with just a few things to say.

1. Why link to the definition of "en masse", yet not to one for "vis-a-vis"? Admittedly, as a French major, my perceptions are a bit skewed, but it seems to me that if you expect people not to understand one, you should expect them not to understand the other. I just checked; both are listed in the main section (that is, not under "Foreign Words and Phrases") in my dictionary. Oh, and there should be a grave accent over the a in "vis-à-vis". (The link ties in with the air of condescension. It's as if you're saying that you expect the readers to be ignorant rubes who don't know the meaning, and can't figure out how to use a dictionary.)

2. Are you directing people to this page at the present time? If you are, it needs to be updated. The fact that the page hasn't been touched in 2 years isn't very reassuring. "Okay, that's what was happening in Fall 2003. What's happening now?" If I were a student and weren't totally put off by your tone, I'd be writing back and asking whether you had any job openings.

3. When you update the text, do a cut-and-paste into a word processor and run spell-check on it. It won't substitute for a proper proofreading, but it should catch things like "Futhermore" and "hourler".

(These two tie in with the idea of modelling the behavior you want to see. If you want to see well-written text, make sure that you avoid obvious mistakes in your text. If you want to see evidence of conscientious work habits, make it an annual routine to go through these pages and update them to the current school year.)

4. Oh, and pick another color for the word "slightly". That yellow (okay, okay - "gold") is hard to read against the white, and it draws attention toward the word itself, and away from the list that follows. May I suggest using gray or brown? Something that's about as dark as the red, at least? (Yes, yes, yes... Cute visual cue there, using the colors of traffic lights. But the blue breaks that pattern anyway, so you might as well surrender now and save time.)

(This ties in with the theme of "Cute is cute, but if the point is to get information across, and cute detracts and distracts... Ditch it." No, it may not be 1337 ha><0r speak, but the effect is similar in terms of the extra effort it takes to read what's written.)

(Can you imagine how long this would have been if I'd had to say everything I was going to say?)
Nov. 10th, 2005 11:39 pm (UTC)
I get two distinct impressions from that page, based on the question of how one first finds it. If it were linked from a relevant page and I came across it on my own, I would smile a bit at the "shut up and leave me alone" sections (that is, until "What To Tell Me About"), feel throughly appreciative to see your list of exactly what you want to hear about and what you don't, then, if I still felt the need, send out a carefully worded e-mail that made it clear I had read your page and was following your instructions. I'm not a grad student, but I am currently looking for a job, and it's always a challenge to decide what the potential employer cares about and what doesn't matter. While I can see the argument for a statement like, "I don't really care about personal references - give me the name of someone who knows you can work well in a lab" being somewhat condescending, its helpfulness far outweighs any hurt.

However, if you replied to an e-mail I had sent - especially if I had used proper English and simply asked an oft-repeated question - with the link, then I'd feel vaguely insulted by the first section (I know this; why does he think he has to tell me?), ponder despairingly which part of the "Information For Applicants" you thought applied to me (Does he think I don't know I have to be admitted first? That I lied about what I know how to do? That I'm too stupid to have read the website before writing him?), and wonder why you couldn't have just copied one of the FAQ answers into an e-mail and sent it back. 100 e-mails per year isn't really all that many, I'd be thinking - he must think I'm an idiot if he just told me to read this, and really hate idiots if he just redirects all of us here. So perhaps you could split the document into several pages, or at least add some anchors so people are directed to the relevant part, to avoid potentially scaring off people who are just ignorant (FAQ/"What To Tell Me About"/"Information For New Students"-worthy) rather than malicious ("Information For Applicants") but think that the whole document is meant for them.

With the 1337 guy, though, the response seems justified - if it's simply a clash of cultures, he should at least pick up the sarcasm about the Dean of Admissions, and know who he should sent his questions to instead. Perhaps while looking through the website for said Dean's e-mail, it might occur to him that no one else is typing in 1337 and that maybe proper English is the way to go.
Nov. 10th, 2005 11:49 pm (UTC)
While a grad student, I was a TA. I would get assignments turned in that weren't as easy to read as your l33t missive. (Back before l337 speak, doncha know...these were just people with the inability to form a complete thought, sentence, paragraph....) I became a demon with a red pen. Seriously, at one point, I think my fail rate for freshman approached 80%. Good lord, how did these people get out of high school?

Nov. 10th, 2005 11:51 pm (UTC)
Q3: Are your courses hard?
A3: Yes. Next question.

My favorite part.
Nov. 11th, 2005 12:26 am (UTC)
I liked that too. If only there was something like the following after it:

"As an example, COMP XYZ requires programming in Brainfuck on the IBM 360"
Nov. 11th, 2005 01:29 am (UTC)
Soup Nazi reference
I'm not Grad School Nazi ("no more Ph.D. for you!"), but that's a reference to Yev Kassem, the Soup Nazi from Seinfeld. ;-)

Nov. 11th, 2005 12:12 am (UTC)

While your webpage is a bit intimidating, I wish all schools were as up front about what they're looking for. I think it's better for everyone in the end.
Nov. 11th, 2005 12:46 am (UTC)
I have to agree with the person above me - and I'm not even a CS person. You make no bones about what you want and what you are looking for. You're firm and to the point. I feel that if I were thinking about applying to your program, I'd know where I stood from reading that. And I've read harsher instructions from grad school departments, to be sure.
Nov. 11th, 2005 01:51 am (UTC)
Re: Your applicant paragraph
Sorry, meant to do this all in one go, but got distracted by quiet coming from the child's room. ;)

I think, just by the way you have it ordered, it may send a more discouraging message than you intend.

Were I attempting to "care bear" it up a bit, I might phrase it thusly:

Dear Prospective KDD Group Graduate Student,

Thank you for your interest in our department. I, like the other faculty in our department and at our university, am personally always looking for talented and motivated students with research ideas and initiative. If you have a specific interest in intelligent systems (please read this document carefully if you aren't sure), please contact me with your academic and research qualifications.

However, it is important to note that due to the sheer volume of inquiries we receive, and the high caliber of our applicants, that not only applicants who may be invited to go further in the application process will get a personal response.


Obviously, I cut a bunch out. Logic behind the cut is this: it's more information than they need, and really only serves to make them feel bad. :)

Nov. 11th, 2005 02:50 am (UTC)
Harsh? Maybe. But great information. That's exactly the kind of stuff I would want to know. As a matter of fact, I went searching for pages just like this when I was looking at schools. I had advice from my professors but not a lot of confidence, and I felt like I was floundering. I may have been too intimidated to apply though. Actually, I probably still would be, even after independent research projects, Hopkins, and guest lecturing at a university. But that would be my problem, not yours. I did like some of the suggestions - i.e. reversing the order of the things that interest you (my favorite part of the page, by the way). But I also liked the "Yes. Next question." It made you seem a little less intimidating, and you did have some serious explanation after that. Besides that, I can't imagine any professor saying "no worries, my class is a blowoff."
Nov. 11th, 2005 04:56 am (UTC)
minor nit to pick
The link to CIS 830 is wrong on the assistantships page. You need to change the 2003 to 2004 in the URL, unless this is some clever test of the motivation level of potential students. :)

Although, lacking a Macintosh with a pseudo-recent version of IE, I have my doubts that I could watch the Tengrity lectures anyway..
Nov. 11th, 2005 05:52 am (UTC)
Actually, CIS 830 should link to 2005
The whole site is rather horrendously out of date, a state I hope to remedy soon, God and Nanowrimo willing.

Actually, there are QuickTime codecs for Tegrity now; they just have to be enabled by the instructor, and average 250Mb a lecture instead of 65Mb. :-P

(MS bailed Tegrity out of bankruptcy around 2000 or 2001, so they are a tad bit beholden. :-P)

Nov. 11th, 2005 07:04 am (UTC)
Re: Actually, CIS 830 should link to 2005
250 Mb instead of 65 Mb? Oi, that's awful. It's worth me digging up a PC to view them at that rate! I'm a little surprised that if they do Quicktime that they don't have a H264 codec (and thus tiny files), given that H264 is apparently the new black. I shall have to check the lectures out on another machine.

Except, not tonight. I've got a train to catch in two hours that's taking me down to Champaign. The biggest bummer about living in bu-fu Minnesota (besides the winters) is that travel to other places tends to be... involved. Like catching the only Amtrak train that goes through town at 4 AM, so that I can avoid a 10 hour drive.
Nov. 13th, 2005 06:38 am (UTC)
It's interesting to see this entry and your page regardings to getting a research position, I am currently in the same situation since I am strongly thinking about applying to a Ph.D program in C.S also.

Below is a few comments, your replies will greatly help me prepare for my plan. Thanks in advance.

Although preferred but I don't think most universities require Subject C.S test. Even with that test, how can you tell if that person is suitable in working with you? It is quite general as it covers many concepts from C.S.

I doubt if people with Bachelor degree have exposed too much into researching yet and hence will not have publications, how 'many' related publications are considered sufficient experienced ?

It's also discouraging that you are not interested at all in activities and hobbies. Unrelating ones are essential but what about related ones ? Those can strongly influence a person' interests and research backgrounds. E.g., hobbies in warfare tatics and strategies can be very beneficial to research works such as agent cooperations & behaviors etc (that's my experience). Even this livejournal hobby can help reveal lots about a person. Is it just a coincident that many of your students also like keeping journals ?

( 31 comments — Leave a comment )

Latest Month

December 2008

KSU Genetic and Evolutionary Computation (GEC) Lab



Science, Technology, Engineering, Math (STEM) Communities

Fresh Pages


Powered by LiveJournal.com
Designed by Naoto Kishi