I was going to write about the log of odds ratios and explain logarithms. This is the very, very odd fact I have noticed in most of social science – people in doctoral programs are often thrust into statistics course for which they really don’t have the basic mathematical foundation. This is because if they were really into mathematics they probably would have gone into that in the first place, but they didn’t. They majored in history or liberal studies or something else. They became teachers or social workers or counselors. In doing so, they took the one (count ’em, ONE) required mathematics course to get a college degree. Further dismaying news, the one mathematics course has been changed dramatically since I was in college – now, you can get a degree with a C- in College Algebra – whatever that is. It definitely does not involve logarithms. When I went to college, Algebra was something you were supposed to have had in high school. But I digress. This whole post is a digression so now I have digressed squared.
So, what did I do all day if not write about logarithms?
Downloaded a dataset from the Interuniversity Consortium for Political and Social Research, which is a site I truly love. What a great idea! Finished with your data? Upload it to the Internet and let anyone else use it for whatever they can find.
Spent an hour (I am embarrassed to confess this) trying to find the error in my SPSS syntax and could not figure out why the HELL it kept saying “file not found” when I could clearly see it there. Finally, as a last resort, went to the c:\ prompt, listed the files and realized that Windows, designed by Machiavelli, hides part of the file name so that my file was actually named college.txt.txt . AAAGH !
Monir, the travel lady, stopped by and took care of my reservations and registration for SAS Global Forum.
Worked on the PowerPoint for my Enterprise Guide class next week. For once, did not waste time rotating the charts in space to see how my bar chart would look sideways (oh, don’t pretend you never did it!)
Enterprise Guide runs pretty slow on my old computer with only 1 G RAM, and EVERYTHING runs really slow with the size of some of the datasets I have been using. My Mac desktop only has 512M RAM (I know, I am deprived) and I was kind of tired of using my laptop and the Unix server for everything.
Today, Justin a.k.a., our hardware guy, came by and told me he had a new computer for me. Like everywhere else, we are watching the budget but somehow he came up with one. So far, I have installed SPSS 17, run a factor analysis with 191,000 subjects and it ran in less time than it took me to type this sentence. I am very happy. Installing the applications I need took a good bit of time, but it will be well worth it in the end in saved time and aggravation. I decided to try something different, so I installed Seamonkey as my browser and downloaded Open Office and Gimp instead of Microsoft Office and Photoshop.
Speaking of sea monkeys, I thought this would be a good example of how statistics can be applied to everything. Even though I work about a block from the museums and Exposition Park, I have only been there once in the last year. So, Tuesday, I walked over to the science museum shop and bought a sea monkey kit. My idea was that I would have it on my desk, collect data and use it in different statistical analyses.
This could also be an example of creativity in analysis in trying to come up with different variables. My original thought was perhaps I could begin with the number of sea monkeys hatched. Unfortunately, statistics for the day are :
Specks floating around that could possibly be sea monkeys – somewhat less than a zillion. I would give a close approximation as 1,000.
Matter which can be definitely distinguished as sea monkeys- 0.
I probably should re-read that code on break point analysis before the meeting tomorrow. I should finish editing the on-line ethics course. Instead, though, I am sitting here with Jenn watching episodes of Numbers on DVD.
She wanted to know, “Is that really true? Can you tell that someone cheated by random numbers? That doesn’t make sense.”
I told her that it was exactly true. It’s like the old Sherlock Holmes story, the Curious Case of the Dog in the Night Time – what was significant was what you DIDN’T see. Sometimes, the telling evidence in an evaluation is that relationships don’t exist where they should, because the numbers were just made up and entered in the database, because, after all, who would ever know?