This is a great video. It is the only one about conditional logistic regression I think. The important thing is that you show how to do it in SPSS. I use it for air pollution data and associated acute coronary syndromes and it is the preferred method in this field. Thank you very much for your effort!
Hi Robin. Thank you for this comprehensive and well explained analysis. Can you please suggest if I can perform this analysis where my case control matched sets are not of same size? This would mean that the cases in my sample will have at least 1 and at max 8 controls. The size if each strata is not same. Can I apply the same analysis as shown above?
This is great! Thank you! In APA you need to report % variance explained by the model. Is there a way to generate this using the Cox regression methods?
Thanks for the video! I have question. The one for monitoring is 4.4, so that means that inadequate monitoring increased chance of death by 4.4 times. And for bmi the chance is actually slightly decreased (0.02) but not significative? Is R automatically giving the result for "death" because it is coding "1"?
Hi can i use conditional logistic if i have data for sample (before and after) application specific treatment on the sample. the research question is: Does utilizing of refill medication reminder feature of DHA have a positive statistically significant impact on medication adherence levels of diabetic patients in developing countries? before (1=adherence , 0=non-adherence ), after (1=adherence , 0=non-adherence ) addition covariate variables like (gender, age, weight and length....) Thank you
Great video! And the first explanation of this analysis that I found that was easy to understand!!! Thank you! One point I am confused on is the first example in SPSS the dummy variable time is used (and I understand why), but then in the R example the dummy variable is not used and from the code it is unclear to me where the specification of case/control is. Any clarification you could provide would be most appreciated!
Hello Robin, This video is most helpful! I created interaction terms and I get zero values for regression coefficient and SE but I do get a value for Wald Statistic, p-value, Odds Ratio, and CI. However, I am not confident any of the interaction output is correct. I am using conditional logistic regression because my research design in a population-based case -control individual matched (1:3). Any insight would be much appreciated.
Thank you so much for the video!!! It helps a lot!!! One question, if the case control ratios are different among pairs, that is some are 1:0, 1:1, and 1:2, and others are 1:3 and 1:4, could SPSS analyze them in the say way you showed here as long as the stratum makes the pairs clear?
Excellent video Robin. One question - isn't the conclusion of the asthma study that more deaths were found with adequate monitoring ? The number of deaths ie. cases with adequate monitoring was 13+9 = 22, whereas the number of deaths with inadequate monitoring was 10+3 = 13.
hi Jeremy not quite of the 35 who died in hospital 23 had inadequate monitoring. of the 35 historical controls 13 had inadequate monitoring - hence the odds ratio of dying in hospital from inadequate monitoring is 13/3 =4.33 I must admit I find this us of historical controls as worrying - and it is not to be encouraged - often used as the preliminary approach to a more robust study design latter
Great video, I do have one question: Is it possible with spss to handle multinomial conditional logistic regression? I have been all over youtube and most only touch binary dependent variable. Thank you
I am having difficulty in the IBM SPSS Statistics system I'm thinking my particular version of SPSS doesn't have full functionality. The only options I have under Analyse>Regression are: Linear/Curve Estimation/Partial Least Squares/Ordinal. Does my version or software not have Binary Logistic? Is any other students experiencing this problem or have in the past? Please help
This is a great video. It is the only one about conditional logistic regression I think. The important thing is that you show how to do it in SPSS. I use it for air pollution data and associated acute coronary syndromes and it is the preferred method in this field. Thank you very much for your effort!
Thank you so much please any video on unconditional logistic regression
Hi Robin. Thank you for this comprehensive and well explained analysis.
Can you please suggest if I can perform this analysis where my case control matched sets are not of same size? This would mean that the cases in my sample will have at least 1 and at max 8 controls.
The size if each strata is not same. Can I apply the same analysis as shown above?
This is great! Thank you! In APA you need to report % variance explained by the model. Is there a way to generate this using the Cox regression methods?
Thanks for the video! I have question. The one for monitoring is 4.4, so that means that inadequate monitoring increased chance of death by 4.4 times. And for bmi the chance is actually slightly decreased (0.02) but not significative? Is R automatically giving the result for "death" because it is coding "1"?
Hi
can i use conditional logistic if i have data for sample (before and after) application specific treatment on the sample.
the research question is: Does utilizing of refill medication reminder feature of DHA have a positive statistically significant impact on medication adherence levels of diabetic patients in developing countries? before (1=adherence , 0=non-adherence ), after (1=adherence , 0=non-adherence ) addition covariate variables like (gender, age, weight and length....)
Thank you
Great video! And the first explanation of this analysis that I found that was easy to understand!!! Thank you! One point I am confused on is the first example in SPSS the dummy variable time is used (and I understand why), but then in the R example the dummy variable is not used and from the code it is unclear to me where the specification of case/control is. Any clarification you could provide would be most appreciated!
Intersting video Rubin....
Hello Robin,
This video is most helpful! I created interaction terms and I get zero values for regression coefficient and SE but I do get a value for Wald Statistic, p-value, Odds Ratio, and CI. However, I am not confident any of the interaction output is correct. I am using conditional logistic regression because my research design in a population-based case -control individual matched (1:3). Any insight would be much appreciated.
Thank you so much for the video!!! It helps a lot!!! One question, if the case control ratios are different among pairs, that is some are 1:0, 1:1, and 1:2, and others are 1:3 and 1:4, could SPSS analyze them in the say way you showed here as long as the stratum makes the pairs clear?
Excellent video Robin. One question - isn't the conclusion of the asthma study that more deaths were found with adequate monitoring ? The number of deaths ie. cases with adequate monitoring was 13+9 = 22, whereas the number of deaths with inadequate monitoring was 10+3 = 13.
hi Jeremy not quite of the 35 who died in hospital 23 had inadequate monitoring. of the 35 historical controls 13 had inadequate monitoring - hence the odds ratio of dying in hospital from inadequate monitoring is 13/3 =4.33 I must admit I find this us of historical controls as worrying - and it is not to be encouraged - often used as the preliminary approach to a more robust study design latter
Great video, I do have one question: Is it possible with spss to handle multinomial conditional logistic regression? I have been all over youtube and most only touch binary dependent variable. Thank you
I am having difficulty in the IBM SPSS Statistics system I'm thinking my particular version of SPSS doesn't have full functionality. The only options I have under Analyse>Regression are: Linear/Curve Estimation/Partial Least Squares/Ordinal.
Does my version or software not have Binary Logistic? Is any other students experiencing this problem or have in the past?
Please help
Look under Analyze -> Survival
Thank you very much. This was really helpful!
Hello Robin,
can we use this test to compare between 2 treatments (placebo=0, treatment=1) in a case matched group?
yes
@@RobinBeaumont Why no intercept calculated for each strata then?
Nicely done.
Awesome video, thank you!
clogit function missing error = probably means you dont have the relevant library loaded
Ty