Hey everyone - we need a statistics major to sit in on a presentation practice and make sure we aren't talking out of our asses too much. We're meeting in the Learning Commons between 1:30 and 4, and anticipate needing help between 2 and 3. Will buy you lunch/beverages or give a small cash honorarium for your trouble.

PM if interested.

4/9/2008 6:43:44 PM

why don't you just post your claims--someone will prolly tell you they're bullshit for free if you take the work out of it

4/9/2008 7:14:52 PM

im a stat major. whats the presentation about?

4/9/2008 7:18:47 PM

I graduated in Stats. I have to work Friday, but yeah if you post/pm some of your claims, I'd look at them.

4/10/2008 8:39:21 AM

Thanks - we are looking for someone to sit in on the presentation and ask questions. I'm pretty confident that what we are claiming will stand up, but am not sure if the regressions we're posting will spur questions we can't answer. We really want to avoid that.

4/10/2008 9:06:38 AM

Why don't you post your questions and let TWW question the validity, you have some pretty smart people here.

4/10/2008 9:39:13 AM

Ah, Regression, one of my few A+'s. Don't know what you're doing, but some general questions: What's the significance of your overall model? What's the R-square? What methods did you use to include/exclude predictors from your model? Are the residuals normally distributed with a constant variance? Were there any problems with collinearity?

4/10/2008 6:04:30 PM

ah - that should do it. i have a lot of work to do tonight.

thanks!

4/10/2008 6:58:07 PM

A lot of work? Heh, if you can't answer any of those in 5 seconds, you may not know what you're doing.

4/11/2008 3:47:02 PM

I'm an undergrad business major, of course I don't know what I'm doing.

I still have my notes from the stats we were required to take and the last two items were specifically mentioned at the end as beyond the scope of the course. Thank God for the intarwebs.

4/11/2008 8:57:18 PM

For which course are you making your presentation? How heavily are you relying upon your regression models (by this I mean, is this the core of your presentation, or just an aside to further substantiate your argument)?

4/12/2008 12:04:18 PM

It's for our IT capstone class. Personally, I consider the models to be an aside (the sample is very small and far from representative), but there will be people from SAS at the presentation asking questions and I don't want to look like more of an idiot than my innate lack of intelligence dictates.

4/12/2008 1:02:43 PM

Oh. Well then be prepared for "What's the point of doing analysis on bad data?"

4/12/2008 1:34:32 PM

small sample = bad news. if i were you, id double or triple your data as it is now just to make it from a bigger sample haha, but thats just me. im a corrupt stat major

4/12/2008 5:26:41 PM

lol - the client knows about the small sample and says it's ok
whatever

4/12/2008 6:22:07 PM

i was reading this just to see where it went I saw you say something about a small sample that's not representative. your sample has to be representative to have an effective and good statistics model; otherwise, you just threw away hours of your time making a model that doesn't apply.

4/27/2008 10:45:51 PM