Rounding and significant figures carry information about precision. A figure quoted to four significant figures from a method with two per cent variability is overstating what is known.
13 likes 23mo
Research Methods · Statistics · continued
Correlation in a self-tracked dataset: what it can support posts 151–167
This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1.
Post #151 answers the question as asked. The question underneath it is different.
A definition problem is doing most of the work in this correlation discussion. Once the term is pinned down I suspect the disagreement mostly goes away and what is left is small.
4 likes 23mo
ra.mensa, post #91: Since correlation keeps coming up, it should probably be a maintained page rather than a recurring thread. I am happy to draft it if someone with more direct experience will review it. Go to post
Time-to-event analysis handles differing follow-up in a way a simple proportion cannot, which is why event rates and Kaplan-Meier estimates can differ.
I would call that likely rather than established.
Narrowing post #154, because the general version has more than one answer.
I would be cautious about generalising from the correlation example above. It is a good example. It is one example.
8 likes 23mo
n.lehtinen, post #109: A p-value is the probability of data at least this extreme given the null hypothesis. It is not the probability the hypothesis is false, and almost every plain-language gloss gets that backwards. I have separated what I observed from what I concluded, which does not always happen. Go to post
Correlation between two derived quantities that share a component is partly artefactual. It is a common trap in analyses of ratios.
Collapsed as off-topic by two members at trust level 3 or above
What I would check first on correlation is whether the thing being measured moved or whether the way of measuring it moved. Those look identical in a graph.
I had written a reply contradicting post #156 and deleted it. Here is what survived.
Working an example through by hand once makes any of these concepts stick better than reading about them, and the arithmetic is usually a single line.
Written quickly, so the reasoning may be tighter than the wording.
26 likes 23mo
Correlation is well covered in the tag pages, and the older discussions are better than the recent ones because they were argued out properly. Worth twenty minutes before adding to this one.
12 likes 23mo
OFalkenberg, post #151: Rounding and significant figures carry information about precision. A figure quoted to four significant figures from a method with two per cent variability is overstating what is known. Go to post
On correlation the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that.
Where an analysis was changed after seeing the data, the honest thing is to report both and say which was pre-specified.
Worth checking against a second source before it gets quoted onward.
0 likes 23mo
Collapsed as off-topic by two members at trust level 3 or above
Effect sizes: the magnitude of a difference, not just whether it is statistically significant. A difference that is significant (p<0.05) might be too small to matter. A large effect might not be significant if sample size is small.
Happy to expand any of that if it is the useful part.
Everything in post #162 holds. The case it does not cover is the one I have.
The practical version of correlation is three sentences long. The rigorous version is three pages and reaches the same conclusion with the conditions attached.
14 likes 23mo
n.kuusela, post #8: Effect sizes: the magnitude of a difference, not just whether it is statistically significant. A difference that is significant (p Go to post
On correlation, I would rather understate and be corrected upward than overstate and be quoted. That is a house style here and it is a good one.
j.hartmann, post #45: Correlation would be much easier to settle if anyone reported the denominator. Almost nobody reports the denominator. Go to post
Number needed to treat: how many people need to be treated to prevent one bad outcome or achieve one good outcome. More intuitive than relative risk reduction.
Posting it because the silence on this was starting to look like agreement.
This topic was closed 45 days after the last reply. Closing is automatic for quiet topics so that a settled answer does not collect new questions underneath it. If you have a follow-up, open a new topic and link back to this one — that keeps both readable and gives your question its own title.
Moved from N-of-1 designs by t.vasquez. Category placement is not obvious from outside and getting it wrong is expected. This topic will get better answers here. The move is recorded in the public log citing R7.
Suggested topics
| Topic | Participants | Replies | Views | Activity |
|---|---|---|---|---|
|
About the Statistics category
Effect sizes, intervals, multiplicity, and the difference between absent and undetected. This post is a community wiki: any member at trust level 3 or above can edit it, and every edit is recorded with its…
|
+4 | 8 | 2.3k | 7mo |
|
Measurement error in home scales, with a worked standard deviation
Measurement error in home scales, with a worked standard deviation Writing it up because I had to work it out twice and would rather nobody else did. What changes if the standard account of measurement error…
|
+26 | 31 | 4.3k | 4mo |
|
Regression to the mean in progress reports
On the subject in the title: Regression to the mean in progress reports Working notes rather than a conclusion. Something about regression does not reconcile and I would like a second pair of eyes before I…
|
+30 | 34 | 49k | 12mo |
|
What a confidence interval means, from scratch
What a confidence interval means, from scratch — that is the question, and I have not found it answered plainly anywhere I have looked. Asking about confidence interval directly, because I have read four…
|
2 | 2.6k | 2mo | |
|
[2026 update] Correlation in a self-tracked dataset: what it can support
Posting this under the heading it deserves: Correlation in a self-tracked dataset: what it can support Everything below is what sits behind that. Posting a small dataset on Correlation. It is mine, it is…
|
2 | 31k | 6h |
Related topics — sharing the tags effect size, heterogeneity, worked example
| Topic | Participants | Replies | Views | Activity |
|---|---|---|---|---|
|
The most common factual error about semaglutide on the internet — what changed since
The most common factual error about semaglutide on the internet — what changed since — setting out what I have, and where I think it stops being reliable. Putting the numbers in the first post, because a…
|
+42 | 46 | 1.7k | 5mo |
|
Immortal time bias in a claims-database study — one year on
Posting this under the heading it deserves: Immortal time bias in a claims-database study — one year on Everything below is what sits behind that. I would like to disagree carefully with the settled view on…
|
+41 | 46 | 3.7k | 22h |
|
Retatrutide's phase 2 heart-rate signal and how to think about it
On the subject in the title: Retatrutide's phase 2 heart-rate signal and how to think about it Working notes rather than a conclusion. Putting the numbers in the first post, because a question about a…
|
+19 | 23 | 20k | 7mo |
|
Second pass at: Journal club: LEADER as the historical anchor
Posting this under the heading it deserves: Second pass at: Journal club: LEADER as the historical anchor Everything below is what sits behind that. Two things I would like separated before anyone answers on…
|
+75 | 81 | 2k | 2y |
|
Peak-to-trough ratio at steady state for a weekly agent
On the subject in the title: Peak-to-trough ratio at steady state for a weekly agent Working notes rather than a conclusion. Working through the kinetics rather than the pharmacology, because I think the…
|
+58 | 64 | 9.7k | 2mo |