How much does "having publications" mean to a undergraduate direct-PhD applicant?

I am an EE student specializing in Communications and Networks (actually a bit more on sensor networks/signal processing) who plans to apply for a direct PhD program in the top US universities.

I have seen some undergraduates who have been directly offered PhD programs by the top electrical engineering (EE) US universities, e.g. MIT, Stanford, and UCB.

Needless to say, they all have a super high GPA and some have very good GRE scores. (I say some instead of all here, because MIT does not require GRE at all!)

Besides these two characteristics, they have also published several papers whilst undergraduates. Maybe just by coincidence, they are all from Microelectronics. As a layman in Microelectronics, I heard that it would be relatively easy for an undergraduate to publish a paper in this field, because a lot of publications are experiment-based.

But it seems quite difficult for me and other peers who also specialize in Communications and Networks to publish some work during as undergraduates. It seems that publishing something in this field requires more knowledge that is out of the scope of a undergraduate.

1. Do most of the EE undergraduate successful applicants have publications? What about in the Communications and Networks field?


2. If an applicant has no final results (papers), but a 9-month research experience, will the experience help? By how much?


3. How can a undergraduate applicant outperform a master applicant who holds some publications during his/her master study?

Remarks: Any generic answers are warmly welcome!

Answer

While I can't comment about microelectronics or EE specifically, I should point out that many US graduate schools admit external candidates only to a Ph.D. program; there is no option available for admission to a master's program en route to a doctoral program. Consequently, the students met the requirements for admission to the school, period.

Research experience is always helpful, particularly because it gives you a potential letter of reference from someone who can directly comment on your skills as a researcher, which is the biggest thing that graduate departments are looking for. They want candidates who are capable of doing high-quality research, and having confirmation of that—whether or not there are papers already written—is a vital component of that process.

I would also suggest it's probably not worth it to try to compete with faceless peers whose credentials you can't guess. That's a hopeless endeavor that will just make you frustrated and dispirited. Just do the best job that you can with the resources available to you, rather than worrying about "keeping up with the Joneses."