Reproducibility
The gold standard of science is independent replication. I've attempted to briefly track replications here - please let me know if you tried to replicate one of our experiments, and it didn't work.
Failures:
- fMRI of working memory didn't replicate in single-subject analyses; reasons might be subject size and our reliance on spatial smoothing / normalization / group average / fixed-effects.
- Striatal D1 dopamine blockade was not replicated by a well-done study from Kamada and Hata. This may have to do with with dorsomedial vs. dorsocentral effects or sample size - we'll try again in mice...
- Let me know of other attempts and failures!
Replication/Triangulation:
- Rodent medial frontal inactivation and temporal control of action: repeated replications by Xu et al 2013, Hardung et al 2017., Emmons et al. 2017, and to some extent by Buhusi et al. 2018.
- "Ramping" or climbing signals in the medial frontal cortex and striatum during interval timing: Wang et al., 2018
- Post-error firing in rodent medial frontal cortex after errors: similar activity in humans in Sheth et al., 2012 and Fu et al., 2018
- Frontal theta and post-error signals / cognitive control: Fu et al., 2018
- Hyperdirect pathways, frontal cortex and STN 4 Hz coherence: Zavala et al., 2018 and Micinovic et al., 2018
- ~4 Hz stimulation improving cognitive tasks in mice, rats, and humans: Scangos et al., 2018
- Emmons et al, 2016, 2017 describing striatal theta activity and interval timing, replicated by Suzuki and Tanaki 2019
- Let me know of other attempts!
Rigor
We are committed to rigorous and reproducible science. This means:
Shared datasets and code. We have been sharing data since 2010 in various forms.
Access to our papers. We also have been publishing in open-access journals such as Frontiers when possible.
Convergent evidence from multiple levels (mice, rats, humans) and multiple experimental systems (single neuron analyses / LFP / EEG, or pharmacology / genetics / virus). Samples sizes are powered by a priori analyses - even though our experiments can be costly and difficult (i.e., intracranial implants in transgenic mice or intraoperative neurophysiology).
Science is a human process, and we make mistakes. Even journals make mistakes. Hopefully, by being as open and transparent as possible we can move closer to making observations that help human disease.