For the love of God, if someone knows how to do logitics regression on SPSS PM ME NOW

Bozo · 09-24-2013

Anyone who doesn't know stats, this is just going to seem weird to you.

This "summer" research has bled into my school year, is fucking with my studies, and now I have to have a poster ready by Friday and STILL need to do this regression analysis. I was taught in my stats class that logistic regression could be used as a way of "holding other variables constant" in order to expose the effect of a true independent predictor. In this way, we can sort of control for confounding right?? So far all the tutorials I'm seeing aren't explaining how this keeping constant thing works. It just seems like a way to look at multiple effects at once and see if they are significant, as opposed to doing chi squares for each one. There's GOT to be more to it than that. THat's what I was told. ffs

Look, I want to see if patients who come from nursing homes have worse outcomes after surgeries than those that come from private homes. Let's say my one question is if they end up dying more often 30 days after surgery. I did a chi square. The answer is YES. Yes, patients who come from nursing homes end up dying 30 days after surgery more frequently than those who don't. Now I want to see if things like age or functional status are confounding that finding. HOW FOR TO DO

CuddleBeam · 09-24-2013

You, me, Skype tomorrow.

10 mins.

Bozo · 09-24-2013

thankyouthankyouthankyouthankyouthankyou

OhZone · 09-24-2013

regression, you say? haha

danielsangeo · 09-24-2013

Think it's a different kind of "regression" there. Though that would be a funny, if esoteric, picture.

"I said I wanted to do a LOGISTICS regression! WAAAAAHHHH!!!"

Rachel Bronwyn · 09-25-2013

Quote:

Originally Posted by CuddleBeam

You, me, Skype tomorrow.

10 mins.

HEY!

Don't you two have too much fun, OK? There needs to be something left for me to come home to.

Bozo · 09-25-2013

Rachel's p value is always significant to me.

Rachel Bronwyn · 09-25-2013

Dude, you own my heart.

OhZone · 09-25-2013

rach/boz
bozo/rach
bozrach
rachel/boz

Karma666 · 09-25-2013

Bronzo.

Bozo · 09-27-2013

Okay, here's the money shot as a way to say thanks! And thanks specifically to CuddleBeam for offering his help as well as lurker Miles who kindly reached out to provide crucial 13th hour advice.

Take care of your elderly, people.

Vincent Van Grow · 09-28-2013

So, what is the reason for the greater mortality, though?

09-24-2013	#1
Bozo Banned Join Date: Jun 2005 Location: NYC, NJ Posts: 2,266	For the love of God, if someone knows how to do logitics regression on SPSS PM ME NOW Anyone who doesn't know stats, this is just going to seem weird to you. This "summer" research has bled into my school year, is fucking with my studies, and now I have to have a poster ready by Friday and STILL need to do this regression analysis. I was taught in my stats class that logistic regression could be used as a way of "holding other variables constant" in order to expose the effect of a true independent predictor. In this way, we can sort of control for confounding right?? So far all the tutorials I'm seeing aren't explaining how this keeping constant thing works. It just seems like a way to look at multiple effects at once and see if they are significant, as opposed to doing chi squares for each one. There's GOT to be more to it than that. THat's what I was told. ffs Look, I want to see if patients who come from nursing homes have worse outcomes after surgeries than those that come from private homes. Let's say my one question is if they end up dying more often 30 days after surgery. I did a chi square. The answer is YES. Yes, patients who come from nursing homes end up dying 30 days after surgery more frequently than those who don't. Now I want to see if things like age or functional status are confounding that finding. HOW FOR TO DO __________________ Real niggas do real thangs.

09-24-2013	#2
CuddleBeam DO THE IMPOSSIBLE Join Date: Dec 2012 Location: Spain Posts: 872	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME You, me, Skype tomorrow. 10 mins. __________________ Thank you everyone! My Deviantart: http://cuddle-beam.deviantart.com/

09-24-2013	#3
Bozo Banned Join Date: Jun 2005 Location: NYC, NJ Posts: 2,266	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME thankyouthankyouthankyouthankyouthankyou __________________ Real niggas do real thangs.

09-24-2013	#4
OhZone OhYeah! Join Date: Jun 2007 Location: Jersey Shore Posts: 38,933	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME regression, you say? haha __________________ -

09-24-2013	#5
danielsangeo Process Moderator Join Date: Mar 2006 Posts: 2,908	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME Think it's a different kind of "regression" there. Though that would be a funny, if esoteric, picture. "I said I wanted to do a LOGISTICS regression! WAAAAAHHHH!!!" __________________ Tell me, where in the world is Daniel Sandiego?

09-25-2013	#7
Bozo Banned Join Date: Jun 2005 Location: NYC, NJ Posts: 2,266	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME Rachel's p value is always significant to me. __________________ Real niggas do real thangs.

09-25-2013	#8
Rachel Bronwyn Banned Join Date: Oct 2008 Posts: 4,002	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME Dude, you own my heart.

09-25-2013	#10
Karma666 Frequent Poster Join Date: Mar 2008 Location: The wrong side of Heaven and the righteous side of Hell. Posts: 238	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME Bronzo.

09-28-2013	#12
Vincent Van Grow Artistic beast! Join Date: Aug 2008 Posts: 225	Re: For the love of God, if someone knows how to do logitics regression on SPSS PM ME So, what is the reason for the greater mortality, though? __________________ GTS is the BEST!