2 replies to this topic

[Teodora Dumitrescu]

Hi,

 

I have a few questions regarding bootstrap:

 

1) Is the Phoenix bootstrap method parametric, or non-parametric?

2) My understanding is that a return code of 1 to 3 is an indicator of the successful convergence. Is this true for bootstrap runs?

3) Is there a way to find out how many of the bootstrap runs have a successful covariance step?

 

Thank you!

Dora

[fang ke]

1.Since the NLME resampling from the origin data with replacement but not from the fitted parameters, i think it's a non parametric bootstrap.

2.I think it's Yes.

3.it seems that use the descriptive stat module could do that or use the split worksheet module.

 

Correct me if i'm wrong!

[Teodora Dumitrescu]

Hi Coco,

 

Thanks for your answer.

A successful covariance step should result in SEs for each individual run. I couldn't find thes anywhere. I assume this is because the whole purpose of bootstrap is to obtain the SE nonparametrically. However, I was just curious if is there a way to retrieve those.

 

Dora