DEAR VIEWERS: I HAVE UPLOADED A NEW VIDEO (at ua-cam.com/video/bDwNkHGBixM/v-deo.htmlfbclid=IwAR3lhZTOC74QjCVqZAVub8kin6SULH1B5b8FkzziUmE-rd7zIOIR1F1Byng ) ON SURVIVAL ANALYSIS IN SPSS THAT INCLUDES LINKS TO AN ADDITIONAL POWERPOINT AND DATASET UNDERNEATH THE VIDEO DESCRIPTION. I DEMONSTRATE A BASIC MODEL AS WELL AS HOW TO TEST THE ASSUMPTION OF PROPORTIONAL HAZARDS. MOREOVER, I ADDRESS THE TREATMENT OF FACTOR VARIABLES IN ANALYSES USING SPSS. PLEASE CHECK IT OUT. THANK YOU FOR VISITING MY SITE.
Great video, thank you so much. Please could you explain how to check the PH assumptions using SPSS? also can Schoenfeld’s test of residuals be done with SPSS?
Thank you for the video - So if I do have one category with more than 2 groups, how do I go about specifying this in the "categorical" option you mentioned?
Hi Kitty, thanks for visiting. I address this issue in a new video (see my Dear Viewers comment above) at: ua-cam.com/video/bDwNkHGBixM/v-deo.htmlfbclid=IwAR3lhZTOC74QjCVqZAVub8kin6SULH1B5b8FkzziUmE-rd7zIOIR1F1Byng . Thanks for your question.
Hi Natassha, I just saw your question. If everyone in the study ended up getting divorced by the end of the observation period, then the status would be 1 (if that is the code that designates individuals who experienced divorce as the terminal event) for those individuals. If no cases were censored, then you simply would have no 0's. [see my comment above about a revised version of this presentation with more info available]
Thanks very much this. In this video you explained the output for abstinence from drinking alcohol in a group of people who underwent a detox programme. At 10:48, into the video you explained that the co-variate “Group” was not significant using a two-tailed test because the p-value was 0.081, but if using a one-tailed test you could reject the null hypothesis (i.e. significant). Could you explain this a bit further? Is there a way of getting a P value for co-variates using the cox regression in SPSS using both one-tailed and two tailed tests?
Hi rhizobium, what I was driving at is that since the chi-square distribution (assuming df=1) can be thought of as the square of the values from the unit normal (z-) distribution, then it is possible to carry out a test of a directional hypothesis regarding the relationship between the covariate and hazard for the terminating event by splitting the printed p-value in half. This is pretty much the strategy endorsed, for example, by Heck, Thomas, and Tabata (2014) when it comes to testing variance components in the context of HLM, where the test can be underpowered with a two-tailed test (not to mention that it makes no sense to adopt a two-tailed criteria anyways, because that would mean you are not only testing whether the variance component is greater than zero, but also less than 0 - which is impossible). You'll also see that if you run the same analysis using Stata, the program actually prints out Wald z-test results and associated p-values - with those p-values being exactly the same as the p-values from the chi-square tests you see in SPSS. Given that, splitting the p-value in half in SPSS or Stata would theoretically allow you to test a directional hypothesis. All that said, people seem to rarely do that in practice. I suppose when I was putting the video together I was just noting the possibility. Thanks for your question and for visiting my site!
Hi, I have gained alot from your youtube videos. You are doing a fanastic work. Anyway, I have a couple of questions I would like to ask on statistics and would be grateful if you could answer them for me. I read an article they had some statistics values and I would like to know how the author got that. 1. "Five-year freedom from reoperation and recurrence of mitral regurgitation 3+/4+ was significantly higher in the prosthetic ring group (90.1% - CL90%: 81.9-98.3%) compared with the pericardial ring group (62.6% - CL90%: 43.1-82.1%; P= 0:027) ". How did he get these values and with which program? 2. I watched you videos on cox analysis with SPSS. How can one perform univariate and multivariate cox analysis on SPSS?
OMG, thank you for this. I have been completely lost with understanding this analysis. You made it clear!
DEAR VIEWERS: I HAVE UPLOADED A NEW VIDEO (at ua-cam.com/video/bDwNkHGBixM/v-deo.htmlfbclid=IwAR3lhZTOC74QjCVqZAVub8kin6SULH1B5b8FkzziUmE-rd7zIOIR1F1Byng ) ON SURVIVAL ANALYSIS IN SPSS THAT INCLUDES LINKS TO AN ADDITIONAL POWERPOINT AND DATASET UNDERNEATH THE VIDEO DESCRIPTION. I DEMONSTRATE A BASIC MODEL AS WELL AS HOW TO TEST THE ASSUMPTION OF PROPORTIONAL HAZARDS. MOREOVER, I ADDRESS THE TREATMENT OF FACTOR VARIABLES IN ANALYSES USING SPSS. PLEASE CHECK IT OUT. THANK YOU FOR VISITING MY SITE.
Excellent explanation. Thank you so much for adding the references and data set. Kind regards
We appreciate your efforts Mike
Thank you, Ali, for your kind remark. Best wishes to you!
Thanks! Great guide
great video, very clear and easy to follow, thank you
Great video, thank you so much. Please could you explain how to check the PH assumptions using SPSS? also can Schoenfeld’s test of residuals be done with SPSS?
why no one is talking about Strata? I need to know what this is
Thank you for the video - So if I do have one category with more than 2 groups, how do I go about specifying this in the "categorical" option you mentioned?
Hi Kitty, thanks for visiting. I address this issue in a new video (see my Dear Viewers comment above) at: ua-cam.com/video/bDwNkHGBixM/v-deo.htmlfbclid=IwAR3lhZTOC74QjCVqZAVub8kin6SULH1B5b8FkzziUmE-rd7zIOIR1F1Byng . Thanks for your question.
If i were to study divorce rates, my variable would be all 1. then what do i put into the 'status' category in spss?
Hi Natassha, I just saw your question. If everyone in the study ended up getting divorced by the end of the observation period, then the status would be 1 (if that is the code that designates individuals who experienced divorce as the terminal event) for those individuals. If no cases were censored, then you simply would have no 0's. [see my comment above about a revised version of this presentation with more info available]
thank you for this video
You are very welcome, Rahem. best wishes
Thanks very much this. In this video you explained the output for abstinence from drinking alcohol in a group of people who underwent a detox programme. At 10:48, into the video you explained that the co-variate “Group” was not significant using a two-tailed test because the p-value was 0.081, but if using a one-tailed test you could reject the null hypothesis (i.e. significant). Could you explain this a bit further? Is there a way of getting a P value for co-variates using the cox regression in SPSS using both one-tailed and two tailed tests?
Hi rhizobium, what I was driving at is that since the chi-square distribution (assuming df=1) can be thought of as the square of the values from the unit normal (z-) distribution, then it is possible to carry out a test of a directional hypothesis regarding the relationship between the covariate and hazard for the terminating event by splitting the printed p-value in half. This is pretty much the strategy endorsed, for example, by Heck, Thomas, and Tabata (2014) when it comes to testing variance components in the context of HLM, where the test can be underpowered with a two-tailed test (not to mention that it makes no sense to adopt a two-tailed criteria anyways, because that would mean you are not only testing whether the variance component is greater than zero, but also less than 0 - which is impossible). You'll also see that if you run the same analysis using Stata, the program actually prints out Wald z-test results and associated p-values - with those p-values being exactly the same as the p-values from the chi-square tests you see in SPSS. Given that, splitting the p-value in half in SPSS or Stata would theoretically allow you to test a directional hypothesis. All that said, people seem to rarely do that in practice. I suppose when I was putting the video together I was just noting the possibility. Thanks for your question and for visiting my site!
Hi,
I have gained alot from your youtube videos. You are doing a fanastic work.
Anyway, I have a couple of questions I would like to ask on statistics and would be grateful if you could answer them
for me. I read an article they had some statistics values and I would like to know how the author got that.
1. "Five-year freedom from reoperation and recurrence of mitral regurgitation 3+/4+ was significantly higher in the prosthetic ring group (90.1% - CL90%: 81.9-98.3%) compared with the pericardial ring group (62.6% - CL90%: 43.1-82.1%; P= 0:027) ".
How did he get these values and with which program?
2. I watched you videos on cox analysis with SPSS.
How can one perform univariate and multivariate cox analysis on SPSS?