The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
The Guiding Principles and New Rules document is now in effect.

[Solved]

dexterdexter Registered User regular
edited May 2014 in Help / Advice Forum
I have a data set with 7 variables and I'm developing a basic Cox proportional hazard model (this is an assignment for a masters course in survival analysis). We've been given a pretty neat outline of what is required for the report, but I just wanted some help with "perform preliminary inspection of data using tables or graphs".

What exactly should this entail? My undergraduate was not in statistics, so this is my first time writing up a report. So far I've sketched a couple of Kaplan-Meier survival estimate graphs, I wondered what else I should include. Things like mean age of the subject, mean time of survival?

Some of the variables have 2-4 levels.

Specifically, my question is as a statistician (not necessarily in survival analysis) what sort of things should I include in my preliminary inspection of data? Any resources you could recommend within this area would be pretty good too.

Thanks in advance, guys!

3694663-1.png
dexter on

Posts

  • LilnoobsLilnoobs Alpha Queue Registered User regular
    I don't mean to sound rude, but why can't you ask the professor these questions?

  • dexterdexter Registered User regular
    That's a fair question. Admittedly I have left this a little bit late - I only have a few days until it's due. I have actually spoken with my professor, but I found it difficult to follow her. It was in person so I don't have an email which I can look back on either.

    3694663-1.png
  • MrTLiciousMrTLicious Registered User regular
    To me, that sounds like summary statistics (that said, I'm not in survival analysis so you could be in a whole different ballpark).

    What you include will be dependent on what is important later, but you'll want at a minimum means and standard deviations of all your variables. It sounds like some of your variables are categorical so you may also want to include that info by category if it's relevant to the rest of your analysis.

  • dexterdexter Registered User regular
    Cool, thanks a lot MrT. I've completed the paper but this is helpful. I have a lot to learn!

    3694663-1.png
Sign In or Register to comment.