about eecs

admissions

academics

research

people

special programs

announcements

administration

recent entries

"My take on Thanksgiving"
[posted by jfischer on 2009-11-24 11:22:51]

"Belichick used results from dynamic programming!"
[posted by lespindle on 2009-11-17 23:45:14]

"I know he's not to blame, but..."
[posted by ndilello on 2009-11-17 14:14:41]

"In sickness and in health..."
[posted by tgolfinopoulos on 2009-10-30 21:46:57]

"Fall Beautiful"
[posted by ttulabandhula on 2009-10-26 02:13:57]

May 19, 2008

A typical experiment

Sometimes doing research really cracks me up. Well, it cracks me up in a, “The other option is crying, so I guess I better laugh,” sort of way.

Designing an experiment usually begins with a question. Some previous bit of work has confused you and you want to conduct some more tests to see what’s going on. So you figure out what to do, get the proper training, and get ready to do your experiment.

Before the results come back, you generally think something like, “Either X or Y will happen. If X happens, that implies alpha. If Y happens, that means beta. Awesome. Two possibilities and, when it’s done, I’ll know which one.”

So you run the experiment. More often than not, it shows Z, which seems to imply something really weird like omichron. Or it shows some combination of X and Y which were two things that you previously thought were mutually exclusive. At this point, you’re not even sure that your results are physically possible. This is not unlike the time in my undergraduate circuits class when my lab partner and I were convinced that we had created charge.

After running this experiment and getting Z, naturally, you have to design another experiment to even verify that Z was indeed possible. And, of course, you think that the results will be A or B, but they come out to be C, so you design an experiment to test C, and, eventually, you’ve run out of letters in the alphabet.

I’m not sure, but I *think* that once you’ve run out of letters and you’re at the height of confusion, you’re about ready to graduate.

(This was all based on a recent experiment of mine… though I’m pretty sure that’s transparent.)

The author has filed this entry under the "Uncategorized" category.

10 Responses To This Entry:

    Can you give any details regarding this experiment of yours? (would love to know what sort of stuff an MIT grad student researches)

    Posted by: ninja on May 19th, 2008 at 2:44 pm

    there’s always the greek letters still!

    or you could just state the probability of a given conclusion, rather than actually stating that that is what’s going on.
    or make a “choose your own adventure” thesis…

    Posted by: Laura on May 19th, 2008 at 8:58 pm

    Ninja, the nontechnical explanation is that I’m working with two different materials - germanium and silicon. Fundamentally, they heat and cool differently and we can actually use that difference to engineer some differences in the germanium devices that I make. We wanted to quantify that, so we sent some devices to our collaborators at Freescale. They did some analysis where they shine a laser on the film and can tell from that how much strain is in the germanium. The change in the strain due to heating is a key factor in having these devices perform better.

    The technical explanation is that germanium and silicon have different thermal expansion coefficients. So we expect that annealing (heating) one sample will induce a greater strain in the germanium and modify the bandgap. This change in the bandgap will make it possible to absorb light at longer wavelengths (closer to 1600 nm). Absorbing light in the 1550 - 1600 nm range is key for this particular application. So we sent some samples to Freescale and they did Raman analysis on them.

    I was studying two effects. The results that we got back were the exact opposite of what I expected for one and some combination of what I expected for the other. So now I’ll spend a week scratching my head.

    Posted by: ndilello on May 20th, 2008 at 11:05 am

    Laura, sometimes I think we should all learn Mandarin. There’s no way we’ll run out of characters then. :-)

    I do like the idea of a “Choose your own adventure” thesis.

    “If the results from this experiment demonstrate A, then please turn to page 56. If they demonstrate B, please turn to page 78. If they demonstrate C, please start again at page 1. You’ve broken physics.”

    Posted by: ndilello on May 20th, 2008 at 11:08 am

    Thanks for the explanation, Nicole!

    That research does sound interesting, though I think you lost me after I hit the part about “bandgap”. LOL

    I hope you find your solution soon. Is there a deadline for your research? Or, is it open-ended?

    One last thing…. Ted seems to like Wikipedia better than me, his Blog love-child!

    -Blog

    Posted by: Blog on May 20th, 2008 at 3:54 pm

    Nicolle,

    It seems so cool and exciting to workhat have no solution till date. I u are doing a wonderful job.

    The more you get confused, the more knowledge you gain about such problems.

    carry on the good work.All the best..

    I hope i get it some day.

    Posted by: jitesh on May 23rd, 2008 at 12:46 pm

    actually i’m not gonna ask you about your A typical experiment.
    i will send you a few Questions in your email,
    and i would really appreciate your response.
    hope you the best luck.
    and keep on it

    Posted by: Ayoub on May 28th, 2008 at 5:34 pm

    Haha, in your “choose your own adventure” thesis, you could have every story engineered to set you back to page 1 hehe…

    And you can probably blame the confusing results on the optics people (wink - i am an optics person).

    Posted by: Laura on May 28th, 2008 at 11:40 pm

    Thanks for the encouraging comments, Jitesh. A friend of mine likes to quote Einstein: “If we knew what we were doing, it wouldn’t be called research.” :-)

    Laura, I actually collaborate with a bunch of optics people. It never occurred to me to blame them, but now that you mention it, there’s CLEARLY a problem with the laser.

    Posted by: ndilello on May 29th, 2008 at 11:07 am

    Ninja, whenever yu see alphabet soup like the above, its best not to ask. The response was far more troublesome

    Posted by: BigHead on October 5th, 2008 at 3:59 am

Post A Response