# Confusion flows downhill- a plea to methods teachers for unified messaging

**URL:** <https://discourse.datamethods.org/t/confusion-flows-downhill-a-plea-to-methods-teachers-for-unified-messaging/24409>\
**Category:** statistical\
**Tags:** rct\
**Created:** [December 22, 2024, 2:19pm UTC](https://discourse.datamethods.org/t/confusion-flows-downhill-a-plea-to-methods-teachers-for-unified-messaging/24409 "2024-12-22T14:19:45Z")\
**Posts on this page:** 1\
**Showing post:** 1

<div class="post-metadata">

**Author:** ![ESMD](https://discourse.datamethods.org/letter_avatar_proxy/v4/letter/e/9e8a1a/32.png) [@ESMD](https://discourse.datamethods.org/u/ESMD)\
**Post date:** [December 22, 2024, 2:19pm UTC](https://discourse.datamethods.org/t/confusion-flows-downhill-a-plea-to-methods-teachers-for-unified-messaging/24409/1 "2024-12-22T14:19:45Z")

</div>

Proposal: Gather the most prominent names in stats and epi in a room together and lock the door from the outside (with meals and bathroom breaks of course), until they can formulate unified messaging on certain foundational concepts.

Recently, on Bluesky:

[https://bsky.app/profile/miguelhernan.bsky.social/post/3lbu5vz4ke322](https://bsky.app/profile/miguelhernan.bsky.social/post/3lbu5vz4ke322)

> **[Miguel Hernan (@miguelhernan.bsky.social)](https://bsky.app/profile/miguelhernan.bsky.social/post/3lbu5vxagjk22)**
>
> Does #randomization ensures balance of risk factors between groups? Consider this:
> 
> In Denmark 860 individuals were randomly allocated to either intervention or control. Individuals were unaware of their allocation. No intervention took place....

One response:

[https://bsky.app/profile/jdwilko.bsky.social/post/3lbukgotmqc2n](https://bsky.app/profile/jdwilko.bsky.social/post/3lbukgotmqc2n)

> **[Jack Wilkinson (@jdwilko.bsky.social)](https://bsky.app/profile/jdwilko.bsky.social/post/3lbukgotmqc2n)**
>
> With respect, this suggests quite a serious misunderstanding about why we randomise. We do this to ensure imbalances are random. Our analyses treat them as such, and ensure valid inference. The problem here is with the idea that point estimates are...

Another response:

[https://bsky.app/profile/stephensenn.bsky.social/post/3lbuaksqnac2w](https://bsky.app/profile/stephensenn.bsky.social/post/3lbuaksqnac2w)

> **[@stephensenn.bsky.social](https://bsky.app/profile/stephensenn.bsky.social/post/3lbuaksqnac2w)**
>
> We should all stop thinking about this in term of point estimates. Once we start thinking about it in terms of inferences (uncertainty distributions) it’s obvious that you should adjust with that which is prognostic, whether or not it’s imbalanced.

It’s hard to believe, as we come to the end of 2024, that these kinds of crazy-making exchanges still occur “at the top.” Let’s ponder how they affect students, researchers, and the medical literature…

---

_[View the full topic](https://discourse.datamethods.org/t/confusion-flows-downhill-a-plea-to-methods-teachers-for-unified-messaging/24409)._