[R-sig-ME] Mixed effects for Discrete Time or Grouped Time Survival Analysis

[Phillip Alday]

Dear Shahla,

I'm guessing that you did not receive an answer on r-sig-mixed-models because your message was rejected for containing HTML. The mail server should have notified you of this. Next time, fix that issue so that your message is actually sent out to the list -- I checked the list archives and my own mail archive and your message directly to me is the only one I received. From now on, please keep the list in CC and send plain-text formatted messages without attachments. This will help us help you. :)

I know essentially nothing about survival analysis, but your mixed model structure is fine in and of itself with two small comments:

1. Usually people list all the fixed effects before the random effects, but although lmer's formula parser can handle both.
2. I see that you also asked this question on CrossValidated, where you received several comments asking about autocorrelation and collinearity of your predictors. These are issues that you need to think about.

Regarding your question about heterogeneity: do you mean heterogeneity of the residuals (heteroskedacity)? Or maybe a lack of sphericity? The latter isn't problematic; the former can lead to biased estimates, which may or may not be problematic depending on your particular application.

Good luck.

Best,
Phillip



> On 11 Apr 2017, at 14:46, shahla ebrahimi <shebrahimi_3622 at yahoo.com> wrote:
> 
> Dear Dr. Alday
> 
> 
> Greetings
> 
> I would greatly appreciate if you could let me know how to do discrete time survival analysis or grouped time survival analysis. In fact, I am studying accounting my data set is related to companies' bankruptcy. My covariates are some financial ratios which are computed at the end of each year. Besides, the issue that a company is gone bankrupt or not, is also determined at the end of each year after preparing financial statements.
> 
> Could you please let me know if :
> 1- It is right to do?
> 
> require(lme4)
> model <- glmer(EVENT ~ TIME + (1+TIME|ID)+x1+x2+x3+x4+x5, data=df, family=binomial (link="cloglog"))
> 
> 2- How to do LR test in order to decide about heterogeneity?
> 
> Some part of my data set is as follows (152 firm during a 12 years period or 1554 firm-year observations of which 50 firm-year observations are bankrupt):
>> 
>> ID	TIME	EVENT	x1	x2	x3	x4	x5
>> 1	1	0	1.28	0.02	0.87	1.22	0.06
>> 1	2	0	1.27	0.01	0.82	1.00	-0.01
>> 1	3	0	1.05	-0.06	0.92	0.73	0.02
>> 1	4	0	1.11	-0.02	0.86	0.81	0.08
>> 1	5	1	1.22	-0.06	0.89	0.48	0.01
>> 2	1	0	1.06	0.11	0.81	0.84	0.20
>> 2	2	0	1.06	0.08	0.88	0.69	0.14
>> 2	3	0	0.97	0.08	0.91	0.81	0.17
>> 2	4	0	1.06	0.13	0.82	0.88	0.23
>> 2	5	0	1.12	0.15	0.76	1.08	0.28
>> 2	6	0	1.60	0.26	0.55	1.31	0.37
>> 2	7	0	1.58	0.26	0.56	1.16	0.35
>> 2	8	0	1.54	0.24	0.59	1.08	0.33
>> 2	9	0	1.72	0.22	0.55	0.84	0.29
>> 2	10	0	1.72	0.21	0.53	0.79	0.29
>> 2	11	0	1.63	0.19	0.55	0.73	0.27
>> 2	12	0	2.17	0.32	0.44	0.95	0.43
>> 3	1	0	0.87	-0.03	0.79	0.61	0.00
>> 3	2	1	0.83	-0.14	0.95	0.57	-0.02
>> 
>> I am so sorry but I sent my question to this group: r-sig-mixed-models at r-project.org . However, I have not  received any answer.
>> Thanks in advance.
>> Best regards,