Of posters, practicals and people in the gyp room

[enismirdal]

I have bacon in the freezer. I want that bacon. The trouble is, I can't do a great deal until the guy down the Hall and his numerous friends finish cooking and give me the kitchen. *tummy rumbles ominously*

I am doing the Plant Sciences poster competition. Today me and the girl I'm working with were meant to meet up and go over stuff. I swear to god she said my room. Well, I do have the obvious advantages of a computer, scanner, colour printer and Internet access. Apparently not. So the meeting sort of failed as we were both in our own rooms. So we're meeting more quickly later. Oh well, such is life. But it kind of splits up my evening a bit. And means video will come late (we're watching But I'm a Cheerleader!)

I have reaffirmed my hatred of Microsoft Software by participating in a Maths practical involving statistics and Excel. My supervisor seems to hold out faint hope for this lecturer and has advised me to get a book on Stats and read it. Wonderful :-\

Drosophila still doing fine.

Oh yeah, I was meant to be going shopping, owing to the fact that I have precisely NO crunchy nut cornflakes left.

Which Harry Potter guy falls in love with you? by Chili
Name/UserName:
House:
Who:	Lord Voldemort
When...	...he meets you in Hogsmeade.
How he tells you:	He sends a howler.
What you do then:	Marry him.
Created with quill18's MemeGen 2.0!

Comments

[girlofthemirror.livejournal.com]

Hope meeting is ok in the end. Have fun with smoking Julia. The film isn't too long, justlet me know when you are ready. We have chocolate!!! Unless I get hungry and eat itall... actually, Imight be good and put it away so that I can't do that.

See you later sweetie

*kisses*

[enismirdal.livejournal.com]

Should be all OK now :) Should be fun, although I do not personally intend to participate in the tobacco-related activities, hehehe!

Choccy sounds good...can't wait!

*kisses and free bonus smooch*

[lifes-a-dream.livejournal.com]

Eek, the maths prac sounds evil :S I skipped it and went to bed because I was so tired! I hate stats.

[emperor]

I could probably help with stats - the stuff so far seems pretty basic to an evil grad like me :)

Oh, and we have Crunchy Nut Cornflakes here in Girton :-)

[enismirdal.livejournal.com]

Oh, and we have Crunchy Nut Cornflakes here in Girton :-)
Grrrrrrrrrrr.....

Oh well, sonce you're here:
1. What is the power of a test (in layman's terms)?
2. Do you really need to add up all the cumulative probabilities just to disregard the null hypothesis with the binomial?
3. What was this Dr T was on about in the supervision today after today's lecture...something about one of the equations being wrong. The one about s squared in a two-sample t-test - and what's all that about anyway?

Please?
Eni

[emperor]

Hey, it's only a gentle cycle ride to Girton :)

1. the power is basically the chance your test has of correctly accepting the null hypothesis. To give a more concrete example: Suppose we're testing a new treatment for cancer. Suppose that it gives a increase in survival time. The power of our trial is the chance of our trial detecting this fact.

2. When n is small enough, it's the easiest way to get an answer, yes. Think back to those graphs I drew in cantabrigiensis - that's the probability distribution. What you're doing by adding up cumulative probabilities is saying "What's the chance of getting a result this extreme or more extreme?" - analogous to measuring the area under that probability distribution curve, If that probability is low enough, you can reject the null hypothesis.

I suppose this is just an example of hypothesis testing: you generate a null hypothesis and alternative, sample the population to get some data. You then measure how extreme that data is - how likely it is that it is explained by the null hypothesis. If that's unlikely enough, you can reject the null hypothesis.

3. I didn't make yesterday's lecture. If you can copy me the handout, I'll have a look.