Rebecca Willén

There must be something I don't understand about statistics. Doesn't the sample mean tend toward the true mean with increasing sample size? Isn't adding replicates a form of replications, and if so, how is this a bad thing?

I love this slide and the discussion around it but fundamentally this is a researcher issue not a problem with statistics. A well defined sampling strategy and design of experiment pre collection stage alleviates such situations. Okay to borrow your slide with credit?

Well specified plans will prevent these problems if they are made explicit in a public repository prior to data collection. Well written preregistrations (with follow-ups by reviewers/editors) and Registered Reports (https://cos.io/rr/ ) can prevent the problems.

Forget about the content; it's an awfully presented slide!

I agree! I'm trying out Visme but it's full of bugs; the colours have changed by themselves several times and in the end I just left it due to exhaustion...

It's difficult to get away from Death by Powerpoint! Have you tried Prezi? It's reliable and much more dynamic in the way it brings information forward - more like a brainstorming diagram than a set of notes.

only thing that makes the slide awful is the color scheme. Either reverse the red and the blue, or make the red in the circles a light gray or something nondescript. Having the number reflect the bluish arc drives me nuts

Can I use this slide (with proper credit)? It belongs in every talk I’ll ever give. It belongs right after my financial disclosure slide.

Selectively reporting studies that "worked". Try submitting negative data to a high profile journal and tell us how that works out. It would be nice if there were a publishing outlet for this, but Editors don't want this.

That's how it has been for decades, yes, but this has changed because of the transparency revolution. It doesn't really make sense any longer to blame the issue of publication bias on the journals: there are so many options for authors, e.g. preprints.

Preprints have a role, but career advancement is not possible without peer-reviewed publications. The status quo is still very much unchanged. However, preprint servers are a good outlet for unpublished thesis chapters.

I respectfully disagree. However, it depends on which (sub)discipline you're in. In psychology, a lot has changed during the last 2-3 years: preregister & you'll be able to publish whatever you find. Other disciplines are still quite unaffected, e.g. sociology & health science.

publishing in an open access journal that just checks science and not content should work. e.g. if all authors have a subscription for a few $100, you can publish many papers a year with little cost. or you can go for a $1000/paper plan. there are options, u just have to look

Again, is the goal to be impactful, or just generate manuscripts?