Interview with Dr. Charles Laidi, winner of the Replication Award 2023

Author: Rahul Brito
Editor: Elisa Guma, Kevin Sitek

Dr. Charles Laidi

Reproducibility of research has been a constant area of concern in neuroimaging. As the field seeks to understand how the brain works and varies between people, it is critical that findings can be rigorously replicated to properly inform future questions, therapies, and interventions. Dr. Charles Laidi set out to do this important work in his paper “Cerebellar Atypicalities in Autism?”, which won him the Replication Award at the 2023 OHBM Annual Meeting.

In this paper, Dr. Laidi and team sought to replicate inconsistent previous findings showing anatomical differences in the cerebellum in individuals with autism. Hypothesizing that lack of statistical power was a culprit, Dr. Laidi analyzed a larger dataset, and found no differences in cerebellar anatomy due to autism when using multiple modeling approaches. To learn more about this, find a more in-depth summary of his paper here. 

Dr. Laidi is an Associate Professor in Psychiatry at the University of Paris Est Creteil. Prior to this he completed a postdoctoral fellowship at the Child Mind Institute in the Center for the Developing Brain where he primarily focused on links between changes in the cerebellum to mental health and psychiatry through imaging. Before working at the Child Mind Institute, he was Assistant Professor in Psychiatry at the Paris University Hospitals (AP-HP),  head of the non-invasive brain stimulation unit at Henri-Mondor University Hospitals, and a Research Scientist with NeuroSpin. He completed his medical training with a residency in both adult and child Psychiatry at the Sorbonne University School of Medicine in Paris. He also his PhD in neuroscience at the French Institute for Biomedical Research (INSERM) and at Neurspin Neuroimaging Research Center,w here he studied the role of the cerebellum in autism, schizophrenia, and bipolar disorder.  Read on to hear Dr. Laidi talk more about his views on reproducibility in neuroimaging.

Q1: How close do you think we are to alleviating the reproducibility crisis in neuroscience, and which obstacles do we still need to overcome in order to do this? 

Charles Laidi (CL): I think that we are still quite far from alleviating the reproducibility crisis in neuroscience, mostly because of the scientific/academic system. I am afraid that the “publish or perish” system incentivizes young scientists to produce a large amount of publications in order to get access to permanent positions. Publishing many studies in a short amount of time (which is relatively easy with brain imaging) is incompatible with the idea of replicating results in an independent cohort or performing rigorous and time-consuming quality control. I think that we should focus more on publishing less and replicating more of our results in general. However, this is difficult,  especially if you are seeking a permanent position or if your lab is running out of funding. 

The authorship system is also a major roadblock for replication. There is little incentive for someone to help you replicate your results in an independent cohort as they won’t be a first author. Likewise, two independent research groups could work on the same topic, compare their results, and publish a paper together, which would be very insightful. But here again, authorship is the problem, especially for large projects. Using alphabetic order in the publication (or a random order to avoid favoring the first letters of the alphabet) could be a solution as in other research fields such as high energy physics or math. 

Q2: In your experience, has journals’ receptiveness to reproducibility efforts improved significantly, or is there still considerable push-back in favor of novel, but perhaps less solid results? 

CL: I think that a lot of journals are more receptive to the reproducibility effort. This point could still be improved but I feel that it’s definitely heading in the good direction. Registered reports are a great way to improve reproducibility. The authors are not encouraged to have positive results; their negative results will be published, which is great for the field.

Q3: How can we further incentivize replication studies and papers?

CL: I think that funding is critical to encourage replication in general. I believe that scientists should be funded to independently replicate results that are important for the field. I would love to see federal agencies propose “replication grants.” If a result is important, it needs to be replicated independently. Those could be “top-down” grants, where a funding agency selects a team to replicate an important paper from the field.