PhD Requirement - 3 ISI Journals
Was chatting with a friend last week. Both of us are in the the academic business and we talked about the requirements set by some local (Malaysian) universities for a candidate to obtain his/her PhD. With all the fuss about publishing... some unis apparently are setting up some, if I may say, CRAZY rules in order for one to obtain his/her doctoral certificate. The general rule was that a student has to produce at least 3 ISI Journals (I believe with impact factor)... if not, the uni will not grant him/her the PhD cert. This is absolutely absurd as 3 journals (ISI pulak tu!!!) within the duration of 3-4 years is quite difficult to produce, especially for a PhD candidate who is just learning the ropes of reading papers!
For some disciplines, yes... this rule can be applied. For instance, it seems that Chemical Engineering students can publish 10 papers (ISI and with impact factor) during their PhDs... similarly, Maths students can also do almost the same. I know this because most of the time, when the School of Computer Sciences USM decides to talk about publishing in impact factor journals, they would invite speakers from the Engineering campus... whose publication records are sky high! While this is not wrong (we do get motivated by these talks), some things have to be put into correct perspective. Once I asked one of my engineering friends (an academic of course)... he told me it was common that each chapter of the thesis to be an ISI paper. The main reason was that, provided the work is good, most of the journals respond within 2-4 months. Comparing with my field (computer science)... it normally takes 6-months just to wait for the answer of either being rejected, or accepted with corrections. Most of the journals that I see (the great ISIs if you may...) can go up to one year! And finally, ultimate acceptance and online publication is only achieved after 1.5 to 2 years.
The thing is this... different disciplines have different timelines for publications. Physics, Chemistry, Maths and some other fields don't take that long for papers to be processed (from what I've learnt). Comp. Science on the other hand takes almost forever. So, to get 3 ISIs might span up to 6-years, and even then only if you're smart and lucky enuff to get acceptance. I really2 do wish that some unis use their brains (as that is what unis are supposedly filled with) when setting such standards. I hear some overseas unis don't even require publications for passing a PhD, and this is from some of the most renowned unis in the world! Malaysia ni terlebih power kot.... We are trying to catch the boat... but thru weird and most of the times absurd policy making. This is my thought :D hehehe... Okes... Assalaam aleykom WBT.
PS:- don't get me started with supervision. Hahaha!!!
Example: This paper took approx. 2-years (Machine Learning paper). Click to view full-size
10 comments:
xpe2..keja swasta kalo derang xbagi phd..sapalah nk ajar bdak2 tu nant yek?
ISI stands for Institute of Scientific Information right? I'm not sure. I read or heard it before.
But even as a layman, I agree with your thoughts. It is not easy to get published in a reputable journal.
All this while I always thought a Phd undergrad only needs to publish once only.
Take care bro.
Anon: Swasta cantek gak. tapi kerja gomen dulu... kutip ilmu banyak2 dan tengok camne things go. kalau makin tak best, baru la paksa diri cabut ke swasta :D
Azwad: I think u mean PhD grads. hehehehe.... mana ada phd undergrad. anywayz.... anywayz... one publication is acceptable (even tho sometimes can ask for more if conference). Kalau they say any typs of reputable publcations (conf or journals) then ok. tapi kaalu wajibkan ISI Journals... gilo apo? I know this very2 smar PhD from Univ College London... and his supervisor is one of the GURUs in Machine learning.... dia nye PhD dia publish 1 conferenec paper, 1 journal paper, and 1 SUBMITTED journl paper. tak accepted lagi tapi. And this from one of the world's most renowned experts (the supervisor I mean).
So with some Msian unis setting so high standards seems very ridiculous.... another classic case of beramal without ilmu? :P
paper ML tu paper ko ke pyan?
aku so far x publish apa2 lg :(
risau pon ada, cuak teramat lah. and the thrid year is looming. pengsan la gini!
Salam Pian,
Here in the UK, there's no such thing as a paper requirement for obtaining a PhD, as far as I know.
Yes, certain fields of studies can generate more papers than other. Honestly for me, a one good paper which is published in a reputable journal (ISI standard etc.) is far better than a tonnes of papers for conference proceedings. I did also heard of cases where students were awarded PhD eventhough they haven't/not yet publish any paper during their course of studies.
So, I pray that Allah may ease each and everyone who is doing PhD at the moment, may success be with them in the future. All the best and good luck, mate.
Salam Ramadhan too...
Noris: Nope. aku nye area bukan ML. as a matter of fact, aera aku ni ntah apa2 ntah. I will need to really2 focus on something a bit more theoretical when (and IF!)I manage to get this PhD done. :P ML is a very2 theoretical area, plus maths knowledge kena strong... things that I currently dont have in my research. Haha!
Nabil PUZ: Yeps... of course 1 good journal paper is better than 10 or 123 conference papers put together (i mean the chikai2 confs that is :P). The power2 conferences mcm ICIP, ICPR etc., kadang2 susah gak nak dapat entry... but anywayz... i agree with UKs (and also other countries' systems yg takde la expect a publication) during PhD. publications are actually results of good research ye dok? Putting publishing first not only gives the wrong idea to newbies (PhD studs that is :P), but also adds unwanted stress to an already stressful postgraduate life! Haha!
And thanks for the prayer n good luck with. I wish the same to you also bro, insyaAllaah :D
Salam. I am also a PhD student in the US. In my field (social sciences) we are required to publish at least 2-3 journals (and cannot cikai2 journals). While it is not mandatory, it is expected. apparently if you are planning on working in the US, you must have 2-3 journal articles by the time you graduate to get employed. Yes, it is disheartening as it is very difficult to get published, and it takes an inordinately long time to get published, with low acceptance rate, particularly in good journals. Good luck to you!
whoa... didn't know that this 'policy' is over there too in the US. But at least its just 'expected' instead of 'required'... but nonetheless, it is quite demanding because u'll be needing it to get a job :( I wish you all the best in getting published...
my mom and dad are in social sciences also and they acknowledge that it takes longer to publish than the sciences...
Which US university you are talking about bro? I guess it's a useless journal. Even MIT PhDs publish one good journal to graduate, and most don't require it at all.
Anon: I don't quite know which US uni the previous commentor is from. As for 'obligatory' publications... well, I see it as a game of catching up with the big guns. At the moment, I see the urge for local unis to become WORLD CLASS. And to do this, good publications is one of the yard sticks used. The thing is, the big guns (e.g. Harvard, MIT, Stanford, and even the Ozi's big 8) actually have research going on. I mean, from what ive learnt from frens, some groups have established problems to solve.... Here, the problems are either made up, or some of the research don't even start from a genuine problem! :P .... And then, most of the supers aren't even publisihing themselves. All pressure is on the postgrads, who are in actuality, newbies. So, it's something like the blind leading the blind (more often than not, the students are less blinded than the supers)
Post a Comment