Autocorrelation Fail
by noel on Apr 30 2008 (1835 Views)Sorry to interrupt the intense discussions!
I’ve been sitting in the same position and working on this lab report since 2PM (I just finished a couple minutes ago, at 11PM), it was fun for a while until I realized that I have six spreadsheets open and my poor antique computer refused to run Excel any longer. I had to give up either FireFox or iTunes, so I wrote the rest of my report in silence. I could’ve sworn that the last time I turned around to look out the window, it was still bright outside…
And you know what Noel likes to do when she’s in the middle of working on stuff… she blogs!
After a series of very serious topics, I figured I’d bright something humorous to the blog. Since there hasn’t been any most of those puppehs-that-look-like-giant-Asian-skunks around, I thought I’d share something from my personal stash. I’m not sure how many of our dear readers are familiar with The Fail Blog. Take a swing at it, it’s the latest graduate student time waster. 
Anyway, I present my latest failure…
That was not (supposed to be) noise!!!!!!! Ah chemistry… I’d still like to congratulate myself for finishing the report like a champ.
Noel
P.S. Dear pchem GSI’s, if you’re reading this, my data turned out perfectly. Please give me an A.
Are you sure you can’t see a peak? I think you could tease element 126 out of there. :p
This post inspired me to look back at an old o-chem lab of mine for my own example of fail.
“Evaporation of the solvent by heating gave crude 10 … as a brown oil at a 70.34% yield. Flash chromatography … yielded 10 as a yellow oil in 2.85% yield…”
The TA’s comment makes it even better (worse?): “You lost 68%?!”
Fortunately, I did not fail that assignment.
The real lesson here is to never ever try and use Excel for a statistical analysis program! It is not.
It is a spread-sheet program.
As a follow up, and to test some anti-spaming mojo.
I usually use GraphPad Prism (http://www.graphpad.com/prism/) for my data analysis. It is quite powerful but easy to use. But not everybody can or is willing to buy it; but I do so much data compiling and combing of data it is worth it to me. Quite often I spend ~8 hours straight compiling and analyzing data from a group of people.
Believe it or not, a really good free program was one create in 1993 for Window 3.1. NCSS Jr (http://www.ncss.com/download_ncss_junior.html) it’s graphs and plots and such are not as pretty and modern looking, but they get the job done.
There is also a nice website with a really good complete statistics and analysis package. StatCrunch (http://www.statcrunch.com/) now, it is like $5 for 6 months and $8 for 12. It is geared more towards Students.
And of-course when you want to jump up to the big boy (which post people never need) is R. http://www.r-project.org/
The analysis I was working on here is for an introductory physical chemistry lab course. Unfortunately, the only readily available analytical tool on this level is Microsoft Excel. To allow everyone equal access, the whole manual is written in Excel.
I’ve been told repeatedly that it’s a bad program go use. Thanks for the recommendations. I’ll definitely look into those, since I have another semester of analytical lab to go.
Sometime I can not stop. I actually enjoy statistics and have learned as much as possible.
If you (you = anybody) wants to do themselves a big favor and has an afternoon to kill, read these two books (they are small, and easy to read):
http://www.amazon.com/Cartoon-Guide-Statistics-Larry-Gonick/dp/0062731025/
and
http://www.amazon.com/How-Lie-Statistics-Darrell-Huff/dp/0393310728/
You will be a better scientists if you do.
And actually, the Cartoon Guide to Chemistry (and the one to Physics) are really good too.
I make new web side for students. I hope you like and use it. Thanks.
http://www.periodic-elements.blogspot.com