The Medusa Fora

bowelling wrote:

Irishbookish wrote:

bowelling wrote:I failed an interview to work at waterstones. Think it was because when asked what book I would reccomend for someone on a train I said Borstal Boy.

Could have also been that lime green sock hat you wore. You should have suggested, Subway by Bruce Davidson. Hope you get an interesting interview soon.

Green hat ??!!? I dont get the reference at all...

Uhm...it was a joke. As I said, hopefully something else will come up.

I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM wrote:

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM, you clearly missed that this is "multilevel" modeling. I'm affraid to ask but either that means modeling from different heights? Or possibly one piece at a time disrobement?

NewJerseyRich wrote:

DzM wrote:

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM, you clearly missed that this is "multilevel" modeling. I'm affraid to ask but either that means modeling from different heights? Or possibly one piece at a time disrobement?

Ahahahahahaha! If only... multilevel modeling is the modeling of data that is "nested" (ie, children nested within classrooms, children nested within families, etc.). The key here is the assumption that kids in the same family will be more alike than kids in other families, which seems simple enough...yet makes the actual modeling a bitch. Basically what I've been doing is writing equations, tossing them into SAS (stats software), looking at what's wrong with the model, re-writing the equations, and doing it all over again. At least 20 times. I shit you not. I've been talking to other students in the class, and they're pretty much giving up. I'm about there...

girlfromcountyhell wrote:

NewJerseyRich wrote:

DzM wrote:

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM, you clearly missed that this is "multilevel" modeling. I'm affraid to ask but either that means modeling from different heights? Or possibly one piece at a time disrobement?

Ahahahahahaha! If only... multilevel modeling is the modeling of data that is "nested" (ie, children nested within classrooms, children nested within families, etc.). The key here is the assumption that kids in the same family will be more alike than kids in other families, which seems simple enough...yet makes the actual modeling a bitch. Basically what I've been doing is writing equations, tossing them into SAS (stats software), looking at what's wrong with the model, re-writing the equations, and doing it all over again. At least 20 times. I shit you not. I've been talking to other students in the class, and they're pretty much giving up. I'm about there...

Ahhh so some type of early childhood Edu/Psych class?

NewJerseyRich wrote: Ahhh so some type of early childhood Edu/Psych class?

Nah, it's straight-up multilevel. I'm in Quantitative Psych, and it's a class in the Quant department. But yeah, most of the application stuff we've done has focused on developmental data.

girlfromcountyhell wrote:

NewJerseyRich wrote: Ahhh so some type of early childhood Edu/Psych class?

Nah, it's straight-up multilevel. I'm in Quantitative Psych, and it's a class in the Quant department. But yeah, most of the application stuff we've done has focused on developmental data.

Well you sound far ahead of the curve here, Picturing you as Shia LaBeouf in The Transformers : Revenge of the fallen movie......going all wacky with the formula he was writting all over his doorm room.

NewJerseyRich wrote:

DzM wrote:

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM, you clearly missed that this is "multilevel" modeling. I'm affraid to ask but either that means modeling from different heights? Or possibly one piece at a time disrobement?

I figured multilevel just meant something like "you and those two models should disrobe and strike an interesting pose." Sadly multilevel modeling is still illegal in many states.

DzM wrote:

NewJerseyRich wrote:

DzM wrote:

girlfromcountyhell wrote:I've been working on a goddamn homework assignment for my multilevel modeling class for over 35 hours. I still have a ways to go.

Huh. Modeling must be more difficult than one would expect. Is there more to it than "take your robe off and strike an interesting pose"?

DzM, you clearly missed that this is "multilevel" modeling. I'm affraid to ask but either that means modeling from different heights? Or possibly one piece at a time disrobement?

I figured multilevel just meant something like "you and those two models should disrobe and strike an interesting pose." Sadly multilevel modeling is still illegal in many states.

I wish that's all it was. I think life would be a lot simpler if that was the case. Luckily, the assignment was handed in today. Next up: modeling trust over time in people at West point (ie, time points nested within soldiers nested within squads nested within squad leaders...and it goes on for awhile). This is gonna get painful

girlfromcountyhell wrote:Ahahahahahaha! If only... multilevel modeling is the modeling of data that is "nested" (ie, children nested within classrooms, children nested within families, etc.). The key here is the assumption that kids in the same family will be more alike than kids in other families, which seems simple enough...yet makes the actual modeling a bitch. Basically what I've been doing is writing equations, tossing them into SAS (stats software), looking at what's wrong with the model, re-writing the equations, and doing it all over again...

All that head-banging Quantative Analysis and here's me thinking about which child I can ask to bring me a cup of tea.

girlfromcountyhell wrote:I wish that's all it was. I think life would be a lot simpler if that was the case. Luckily, the assignment was handed in today. Next up: modeling trust over time in people at West point (ie, time points nested within soldiers nested within squads nested within squad leaders...and it goes on for awhile). This is gonna get painful

Hi HellGirl. Could you tell me something? In terms of your multi-level modelling what is SAS?
I've been seeing it referred to by a lot of Irish students on Facebook. I quickly worked out that they weren't being attacked by the S.A.S.*, but were, in fact, being highly vexed about some arcane statistical analysis maths type stuff. So, what's it stand for? And good luck with your studies too, by the way.


* Special Air Service

SAS = Statistical Analysis Software. http://www.sas.com/ Their company headquarters is about 3 miles up the road from me and the town I live in was built in large part with SAS dollars.

When I was in school I had to use SPSS.

kmurray105 wrote:SAS = Statistical Analysis Software. http://www.sas.com/ Their company headquarters is about 3 miles up the road from me and the town I live in was built in large part with SAS dollars.

When I was in school I had to use SPSS.

Cheers, KMuzz. SPSS was what the ex-wife was using for her pysch degree/masters. Thought SAS would be something similar. Thanks for the heads-up.

Craig, they're similar in that they can do a lot of the same things. SPSS (now called PASW, pronounced "paws") is more of a point-and-click program, while SAS is syntax-based (ie, you have to tell the program what to do rather than clicking a button) and tends to be less user-friendly. But it's growing on me...the learning curve is incredibly steep though.