-.center[
-![](https://book.the-turing-way.org/_images/reproducible-matrix.jpg)
-.pull-right[(from [The Turing Way](https://book.the-turing-way.org/reproducible-research/overview/overview-definitions#table-of-definitions-for-reproducibility))]
+
+
+
+
+.boxed-content[
+.large.center[
+# Processing and reproducibility
]
+.larger.center.gray-text[How do they interact? Why *fMRIPrep*?
+]]
+
+---
+
+.center[
+
+.small[
+*The Turing Way project* illustration by Scriberia. Used under a CC-BY 4.0 licence. doi:10.5281/zenodo.3332807.]
]
---
# Reproducibility: increasing measurements' reliability
-
.large[
@@ -581,12 +598,15 @@
Standardize the measurement procedure
- .gray-text[Reduce variance (potentially at the cost of bias)]
+ .gray-text[Reduce methodology variance (potentially at the cost of bias)]
]
-.center.larger[.larger[Neither ensures the *validity* of the measurement]]
+.center.large[
+Both strategies *reduce the variance* of the measurement.
+
+Neither ensures the *validity* of the measurement]
---
@@ -599,7 +619,7 @@
]
.larger[
-* NARPS ([Botvinik-Nezer et al., 2020](https://doi.org/10.1038/s41586-020-2314-9)): single dataset, 70 teams, 9 *ex-ante* hypotheses.
+* NARPS ([Botvinik-Nezer et al., 2020](https://doi.org/10.1038/s41586-020-2314-9)): single dataset, 70 teams, 9 ex-ante hypotheses.
Striking analytical variability (even with teams using same preprocessing)
@@ -755,7 +775,7 @@
* Use *Nipype* to abstract from tool details and construct modular workflows
]
-.large[ Standardization and best-practices]
+.large[ Standardization and best-practices — [check (Esteban, 2024)](https://doi.org/10.31219/osf.io/42bsu)]
.larger[
* BIDS (inputs), BIDS Derivatives (outputs), BIDS Apps (architecture).
* Containers, versioning, LTS, open-source, CI/CD, etc.
@@ -789,6 +809,28 @@
]
+---
+
+# Standardization in the research workflow
+
+.boxed-content[
+
+.center[
+
+]
+
+
+.larger[
+Data collection easy to standardize (BIDS, scanners' software)
+
+Statistical modeling keep analytical flexibility not to stifle the development of new ideas
+
+Preprocessing offers an **opportunity** to reduce analytical flexibility and **enable** cross-study comparisons
+]
+
+]
+
+
---
# Standardizing preprocessing: visual reports
@@ -825,6 +867,25 @@
---
+# Building a community was important to *fMRIPrep*
+
+.boxed-content.large[
+.no-bullet[
+* Feedback & support:
+ * error & bug reports help development,
+ * feature requests help driving a roadmap,
+ * questions are the entry point to deliver support
+ * [ NeuroStars.org](https://neurostars.org)
+
+* Engage contributors (see [our guidelines](https://www.nipreps.org/community/CONTRIBUTING/)):
+ * documentation, assistance debugging, code patches, everything counts!
+
+* Reach out, increase user-base.
+]
+]
+
+---
+
# Standardizing preprocessing: software versions
.larger[