How can you get something quantitative

送交者: cornbug 于 2012-01-12, 23:03:10:

回答: Cornbug: how is film compared to digital? 由猎人于 2012-01-12, 22:34:50:

out of Kaplan-Meier Survival curve. The only thing you can do is to visualize it. If there are two groups for comparison, how can you get some meaningful number out of it so that you can say one is significantly different from the other?

I am asking because somebody suggested me to do the survival analysis that I have never heard before. Also the quality of life analysis. I simply don't understand the statistics.

所有跟贴:

Missed your Quality of Life question. here is my 2 cents: - 猎人 (855 bytes) 2012-01-13, 08:07:57 (641189)
- I have "change scores" as primary. Not sure Survival analysis - cornbug (105 bytes) 2012-01-13, 09:06:23 (641197)
  - there is some confusion here. - 猎人 (199 bytes) 2012-01-13, 11:00:38 (641231)
    - RCT design to test the effectiveness of an intervention - cornbug (647 bytes) 2012-01-13, 11:31:10 (641240)
      - I am now getting your point - blackbox (461 bytes) 2012-01-13, 13:02:42 (641251)
        
        if I remember correctly, these are all called "quality (control) study" - 猎人 (334 bytes) 2012-01-13, 13:41:08 (641256)
        
        Questionare could collect rating or scale data, - blackbox (43 bytes) 2012-01-13, 16:11:09 (641283)
        
        Questionare is OK, for the endpoint score - cornbug (68 bytes) 2012-01-13, 16:41:34 (641294)
        
        I don't understand this"...development of symptoms....." (无内容) - 猎人 (0 bytes) 2012-01-13, 19:05:55 (641310)
        
        The severity is a function of exposure time. (无内容) - cornbug (0 bytes) 2012-01-13, 23:25:16 (641333)
        
        it may more sense by reading this and the one below - blackbox (176 bytes) 2012-01-13, 18:18:30 (641306)
        
        I have never been involved in a clinical trial - cornbug (929 bytes) 2012-01-13, 13:32:55 (641254)
        
        without much details re: type of intervention and outcome, - 猎人 (318 bytes) 2012-01-13, 14:43:13 (641261)
        
        ha, that is what I can say - cornbug (736 bytes) 2012-01-13, 15:05:14 (641264)
        
        currently, is there a measure to manage this condition? - 猎人 (86 bytes) 2012-01-13, 15:20:59 (641267)
        
        then your procedure and/or med is to abolish the attack - 猎人 (2058 bytes) 2012-01-13, 17:48:57 (641305)
        
        suvival analysis and QoL are additional measures - cornbug (38 bytes) 2012-01-13, 16:38:39 (641292)
        
        Every thine has a psych impact, but not serious. - cornbug (302 bytes) 2012-01-13, 16:34:05 (641290)
        
        Yes. She will do and get 5-10% of her salary. (无内容) - cornbug (0 bytes) 2012-01-13, 16:23:38 (641286)
        
        triggered externally. (无内容) - cornbug (0 bytes) 2012-01-13, 16:21:36 (641284)
        
        is the condition persistent or episodic? (无内容) - 猎人 (0 bytes) 2012-01-13, 16:07:19 (641282)
        
        still don't feel - blackbox (386 bytes) 2012-01-13, 16:04:08 (641281)
        
        what do u claim about your procedure compared with the med? - 猎人 (124 bytes) 2012-01-13, 15:54:43 (641278)
        
        Yes. Medications with temporary relief and side effects. - cornbug (20 bytes) 2012-01-13, 15:28:48 (641270)
      - few comments - 猎人 (817 bytes) 2012-01-13, 12:35:47 (641245)
        
        Thanks. I have never had a cell. (无内容) - cornbug (0 bytes) 2012-01-13, 12:57:53 (641250)
        
        Oh man! you may have a loud voice and pair of sensitive ears. (无内容) - 猎人 (0 bytes) 2012-01-13, 13:12:35 (641252)
        
        I am a bug with the sixth sense. (无内容) - cornbug (0 bytes) 2012-01-13, 13:39:30 (641255)
  - there is some confusion here. - 猎人 (199 bytes) 2012-01-13, 10:59:28 (641229)
- good points, thanks! (无内容) - 通知 (0 bytes) 2012-01-13, 08:38:07 (641191)
it is good opportunity to learn new stuff - blackbox (325 bytes) 2012-01-13, 07:58:49 (641188)
- it is pointless to manually calculate. - 猎人 (46 bytes) 2012-01-13, 12:45:32 (641246)
  - true. Stats software willl do all the calculation (无内容) - blackbox (0 bytes) 2012-01-13, 13:15:14 (641253)
- SPSS has K-M analysis too (无内容) - 通知 (0 bytes) 2012-01-13, 08:39:50 (641192)
  - I leave these to a statistician but I have to know about them. (无内容) - cornbug (0 bytes) 2012-01-13, 09:00:18 (641194)
    - good practice. for statistician, they prefer SAS. SPSS has lots of limitations (无内容) - 通知 (0 bytes) 2012-01-13, 10:02:11 (641218)
1, to quantify resolutionn enlarge them until u see a difference. - 猎人 (87 bytes) 2012-01-13, 05:59:46 (641177)

加跟贴