Updated package: Version 10.0

[Katie Gates]

Hey all, 

There's another big package update.

Part of the update is due to non-reproducible findings people had with the 9.## versions. The default for stopping the model search was altered, and I apologize for the unintended issues introduced here. We didn't think it would have a broad impact. 

Version 10.0 has the following changes: 

- Revert to old stopping criteria: The default stopping criteria matches other recent versions (except 9.##) in that it adds paths until model fit is acceptable even if they are not significant at the individual level. 

- Adding False Discovery Rate (FDR): Two arguments, "group_correct" and "indiv_correct", now have the option to correct via 'fdr' rather than the default Bonferroni correction for multiple comparisons. FDR is less strict and commonly used in fMRI studies. 

Please let me know of any issues, questions, or bugs that you find.

All the best, 

Katie