About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Experimental Update, For Those Who Care | Main | Between Friends. Sort of. »

April 28, 2005

Speak for Themselves, They Do

Email This Entry

Posted by Derek

I thought about writing a whole post welcoming back my fellow Arkansan, the Ivory-Billed Woodpecker, but decided that not many folks would sit still for that. It does strike me, though, that the bird was rediscovered not that far from where I went to college.

I didn't have much time to hang out in the Cache River Wildlife Refuge, anyway, even if I'd been so inclined. I was living the life of the virtuous chemistry major, which for one semester led me to have all-afternoon lab sections every single day of the week. There's no doubt that I learned something in them, but not all of it was in the official curriculum.

Take one time in a physical chemistry lab, for example. We were doing something that no one in their right mind does in the real world, a determination of molecular weight by boiling point elevation. That's a relic of the sealing-wax era of chemistry, but it's not quite as much of a waste of time as those qualitative organic tests I was railing about the other week. I still wouldn't keep it in a lab course these days, but the way this one was set up, it did have some value.

We were figuring out the molecular weight of benzoic acid by adding increasing amounts of it to a solution (toluene, I seem to recall) and seeing how much the boiling point increased. We then plotted this out, running it through Raoult's Law to get the answer.

Well, of course, we already knew the molecular weight of benzoic acid. But as we took boiling point after boiling point, in a finicky apparatus that splashed boiling solvent over the lowered bulb of our thermometer, we could tell that something was going wrong. We were getting something way over 200, and benzoic acid weighs 122. Hrm. We checked everything again, but the data all looked pretty good. Way over 200. . .about 244, actually.

Then it hit one of us. "Dimer!" he said suddenly. "Benzoic acid forms a dimer, and that weighs twice as much!" We all slapped our foreheads and grinned. But the guys next door to us had put their minds at rest even before we did.

Not that they had the right answer. In fact, their plot showed the molecular weight of the dissolved benzoic acid at 122, right on the nose. Pretty good curve, too, pointed right at the seemingly-correct-but-impossible answer. How did they get there? By taking so many data points that the freaks and throwaways could be assembled, Frankenstein-style, into a zombie plot that gave the answer that they just knew it had to give. Never mind that the other 90% of the data pointed to twice that number - that can't be right. What do the data points know about right answers, anyway?

Comments (6) + TrackBacks (0) | Category: How Not to Do It


1. Lou Wainwright on April 29, 2005 3:11 PM writes...

One of the things I always tell my characterization group to do before using Excel to fit curves to data is to take out the outliers (which, in data with non-normal distibutions can make very misleading curves). However, I insist that the outlying points not just 'disappear', but that they get plotted as a different series on the same graph. That way other eyes have a chance to judge if good choices were made on outlier removal, but the visual impact of the curve fitting isn't impacted. I'm always shocked by how little training engineering graduates have had in interpreting and displaying data.

Permalink to Comment

2. steve on April 29, 2005 4:10 PM writes...

In high school physics, we had a lab where we whirled weights on cords with tension spring scales attached. I don't remember what exact relationship we were looking for, only that the teacher accidentally got the wrong thing in the denominator. When our group went to him to say that our results were getting the "wrong" sign (we thought we might be doing something wrong), he figured out his error but told us not to say anything to our classmates. Sure enough, every other group found the "right" relationship even though it was physically impossible. It was a good lesson in integrity and/or avoiding self-deception.

Permalink to Comment

3. Jason on April 29, 2005 4:42 PM writes...

I had a professor once who told us that doctoring undergrad lab data should be its own course.

Permalink to Comment

4. jsinger on April 29, 2005 5:22 PM writes...

Reminds me of the old Bill Maher bit: "Chemistry is important because it teaches you how to do your taxes. I think most of us do our taxes the way we did things in chem lab -- start with the answer and work backwards."

I disagree on the woodpecker, though. I couldn't tell it from any other woodpecker, but still think it's an amazing story.

Permalink to Comment

5. Drew on May 1, 2005 2:02 AM writes...

Teaching undergrad o-chem lab sections at at a major research university, students would often have a hard time getting the wet, wet crystals to drop to the bottom of the mp tube, and then report a dead-on mp, with a 1-2 degree range....!

Permalink to Comment

6. muhamed on May 6, 2005 11:15 AM writes...

The most memorable moment in my short research career is consistently provided by a pharmacology graduate now doing a PhD in another lab. This young 25 yrs od lady calls at least 3 times a month to ask her friends to work out concentration of solutions that she needs to make up. The story gets even funnier when the postdoc she called for help asks for advice from other people in the office while her friend is still on the phone. The irony of this story is that being good in memorising factual information in school doesnt make a good scientist but rather critical thinking is where more emphasis should be based.

Permalink to Comment


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry