| Attachment | Size |
|---|---|
 warning.png | 23.78 KB |
Dear developer:
When I use ACE model to calculate heritability，I got a warning messages during the analysis（See Attachment）. Is there a problem with my data and will it affect the results?
when there was a warning message, I found that the estimated results will change under different start values.
Could you please provide me more details?
Thank you very much.
Best regards,
Zhenzhen Luo
Printer-friendly version
Are you analyzing threshold variables?
Rob is right that IFAIL=6 is a common result with ordinal data, due to the limited precision of numerical integration necessary to evaluate the likelihood.
A few attempts to recover the same solution, but using different starting values, will usually help. If the same solution is repeatedly found starting at different points in multivariate space, then the hypothesis that the solution is correct gains some weak support by not being rejected.
HTH