Xiuhua Wei
-
How to standardise error reporting across different experimental assays in o...
Method Help
A single manuscript often combines microscopy, sequencing and behavioural tests. Each community reports uncertainty differently. Inconsistent error notation confuses readers and reviewers. What standardisation strategy works well?
Donald Thomas, Dennis Campbell and 23 others4 CommentsView more comments-
Define whether error bars mean SD, SEM or confidence interval in one central place.
2 -
Prefer confidence intervals where inference is the main goal.
2 - View more comments
-
-
What steps do you take when a key experiment cannot be reproduced by your ow...
Method Help
Encountering non‑reproducible results inside your own group is both frustrating and scientifically important. It forces a careful re‑examination of methods and assumptions. How do you systematically investigate a failed reproduction?
View more comments-
Compare the exact experimental conditions and parameters between the original and reproduction runs.
2 -
Check for unrecorded variations such as reagent batches or environmental differences.
2 - View more comments
-
-
Key considerations for setting up shared code repositories for multi‑party r...
Collaboration
Shared code repositories are central for joint computational projects. Beyond basic git usage, what collaboration conventions reduce merge conflicts and keep shared code maintainable for multiple distributed contributors?
View more comments-
Define branch‑naming rules and code review requirements before multiple people push code.
2 -
Require concise commit messages explaining what each change accomplishes.
2 - View more comments
-
-
Step‑by‑step approach to create a reasonable ablation study plan
Method Help
Poorly‑designed ablation studies waste computing resources and deliver limited insight. Many new researchers struggle to decide which components should be ablated. What systematic thinking guides your ablation experiment planning?
View more comments-
First define your core claim; every ablation should directly test one part of that claim.
3 -
Ablate one variable at a time whenever feasible.
2 - View more comments
-
-
How to respond when reviewers raise overly harsh comments
Receiving tough reviewer feedback is normal in academic publishing. Stay objective, respond point‑by‑point and avoid emotional replies, even if you disagree with certain viewpoints.
View more comments-
I once got very critical reviews and almost gave up resubmission.
2 -
Make sure every rebuttal point has solid supporting evidence.
2 - View more comments
-
