September 14, 2006

Midnight Oil Review

Why do we do analysis of variance (ANOVA)? Because we care. A lot. Unless I take another desperation break sometime after the nightly nap, it's going to be light blogging (by which I mean that your author will probably lurk around dKos a little bit and post frak-all) tomorrow. Today. Whatever, you know? But we have music, oh yes we do. And a book, or rather, a Book.

1. Maggie's Farm - Bob Dylan
2. The Distance - Cake
3. What It Is - Mark Knopfler
4. God Shuffled His Feet - Crash Test Dummies
5. Glass, Concrete & Stone - David Byrne
6. Lion's Mane - Iron & Wine
7. Suzanne - Leonard Cohen
8. Kerosene Hat - Cracker
9. Come Together - Aerosmith
10. Out Of This World - Bush

I would definitely not be listening to and/or reading d r i f t g l a s s' (widely linked) David Bowie fueled 9-11 tribute, because there's grim to spare around here this evening, particularly in light of the sad news that Ann Richards is no longer with us. Nor would I touch with your Nine Inch Nail this very, very wrong Star Trek slash mix; one emotion per day, people, one emotion per day. Instead, especially if my professor is visiting, I'm reading ...

Agricultural Experimentation: Design and Analysis - Little, TM; Hills, FJ.(1978) John Wiley & Sons, New York.

I can really only speak first hand about chapters 3 and 5 at the moment, but if you need to perform ANOVA on the data from a randomized block experiment, Little & Hills are your men. Word.

It's apparently a follow up to Statistical Methods in Agricultural Research and to look through the other chapter headings, they might have wanted to stick a little closer to that name. Because I know what you're thinking. You're guessing from the title of their 1978 tour de force that if I shuffled over a few pages from the ANOVA discussions, that I might find loads of handy tips on designing projects to measure nutrient cycles or calculate feed efficiency for livestock. But you'd be flat wrong. You really have to look for that sort of thing, buried as it is in examples to the larger points. It's all linear correlation and regression this, curvilinear relations that and analysis of counts the other. They're big on blocks and plots, though, have to give them that.

Dry stuff, serious stuff, but very handy when you've come up with some experiment loosely based on Stephen Gliessman's Field and Laboratory Investigations in Agroecology, or just some paper you read last quarter, and your professor looks you straight in the eye and wants to know if your results are significant. In other courses of study, that might be a soul-wrenching and existential question, but I really get enough of that sort of thing just going through life. Don't you? Indeed, it's a great comfort to me to know that verily, Little and Hills can help with the answer, that it isn't about me and that not even our current administration has managed to eradicate all our degrees of freedom.

"We gladly run in circles, but the shape we meant to make is gone." - Lion's Mane by Iron & Wine

Posted by natasha at September 14, 2006 12:03 AM | Random Mumblings | Technorati links |