Wednesday, January 16, 2008

n and n-1

The SD is the average 'distance' of the observations in a particular variable from the mean of that variable. Usually, for finding the mean, or average, we divide by n, the number of observations. However, it happens that when the number of observations is less than 30, dividing by n slightly distorts the result when using the normal distribution. The distortion can be corrected by dividing by n-1.

Here is a good worked example which you can alter.

Wednesday, January 9, 2008

Skewness

In Lectures we discussed skewness. There are a number of definitions relating the mean, median and the mode to skew. If you're interested, here is a link to a journal article:

http://www.amstat.org/publications/jse/v13n2/vonhippel.html


Notice that when in Stata you can find out the skew of a variable by using the command

summarize (varname), detail

If it's positive, the values are skewed to the right and vice-versa. Zero skew means a normal distribution.

Monday, January 7, 2008

WebCT going

The WebCT for AGSC252 is running now, and I have posted some files. Check you can access it and let me know of any problems.

Saturday, January 5, 2008

Labs 01, 04, 05

This message applies just to those students registered for these three labs.

I just checked the seating for the labs and found (I am sorry to say) that the three labs 01, 04, 05, are slightly overbooked. There are only 20 terminals in the lab, so the overbooking means that some students will need to share. This isn't ideal, so you might wish to change your lab. Some labs are almost empty: for example L09 has only 2 students registered so far. If you switch, you'll get more attention from the instructor.

Thursday, January 3, 2008

WebCT and Stata pick-up

Two quick points:

The WebCT for the course isn't running yet. I hope that we can have that going by next week. I will usually use the WebCT for sending out lecture notes etc.

If you ordered a copy of Stata, but didn't collect it: I will be passing out your copy at either a lab or the first lecture. Whichever I see you at first. Please bring along the $1 for the customs fee. And if you were one of those who owe Penny Brasher that $1 from December, please pass it to me.

Wednesday, January 2, 2008

Getting Started

Lab Passwords: as I've said before, it's important everyone can use the computers in MacMillan 192 right from the first lab on Monday. So, make sure you get all set up well before your lab. It takes a day or so for the main computer to allocate a set-up to you. A helpful student kindly provided the following notes on how to do it. Notice that you can't do this off-campus.


Step #1 - GO TO ROOM MCML 194 to create the password!

Step #2 - Choose a computer. The screen will ask whether you already have an account, not have any account, or if you are not a LFS student (check the one that applies)

Step #3 - type in your student number on the next page and it will direct you to CWL log in

Step #4 - log in as you would to CWL

Step #5 - fill in the information asked (this is where you will choose a password!)...and then you are done (make sure you log out)

After, I headed to room 192 to try logging in. The screen should ask you for:

USERNAME: (type in your STUDENT NUMBER! NOT CWL username!)
PASSWORD: (the password you just created)
LOG ON TO: LANDFOOD


AND YOU'RE DONE! :)

It shouldn't be a big problem for most people because as I've said, it is step-by-step. The only thing that might confuse some people is that they try typing in their CWL username as the "username" for these computers. Therefore, emphasize the fact that it is their STUDENT NUMBER the computers are asking for.

Materials for labs, worked example sessions, and lectures: In the e-mail I attached these documents:

1. Lab 1 Print out and bring with you to your lab. if you have a copy of Stata, you can work through it on your own if you wish.

2. Lecture 1: Print out and bring with you to the first lecture on Tuesday.

3. Module Notes 1: As above. We may not start this on Tuesday, but certainly we will on Thursday. Good to have with you as you can then pick up the way we'll be working.

4. Worked Example 1: if you're attending a WE session, bring this along with you. Whoever is taking your session will be working through these questions, which are typical exam questions. Notice that we won't go through these questions outside of the scheduled WE sessions. Your choice!

5. The Stata dataset for the Cheese example, which appears in Lecture 1. This is just an interesting example of the way that we'll be using Stata and statistics in relation to food and nutrition type topics.