Data & Tools · Datasets · continued

Contributing data without breaching anyone's privacy posts 61–72

Data & Tools › Datasets data table confounding observational data Solved

This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1 · go to the accepted answer.

OTeixeiraTL3Regular6 Mar 2026#61

a.novak, post #45: Combining data from different sources: datasets from this site are not directly comparable to published trials because the populations are different. They are worth reading separately, not merged together. That is what I would do. It may not be what is correct. Go to post

Everything in post #57 holds. The case it does not cover is the one I have.

Where a dataset is used to support a claim in a maintained document, the version used should be cited. Otherwise the document and the data drift apart silently.

Reporting the observation and leaving the explanation open deliberately.

1 like in reply to #45 5mo

d.nilsenTL26 Mar 2026#62

Narrowing post #61, because the general version has more than one answer.

Collection methods: ask how the data were collected. Longitudinal tracking over months is stronger than retrospective recall. Prospective measurement (done while experiencing something) is stronger than memory afterward.

A qualification I should have led with rather than closed on.

0 likes 5mo

OkaforTL3Regular6 Mar 2026#63

Limitations of datasets: all community-collected data has limitations. The population is self-selected (people in this community are not representative of all people using these compounds). Reporting bias is real (remarkable outcomes get reported; mundane outcomes do not).

22 likes 5mo

i.oseiTL26 Mar 2026 · edited#64

Two things can be true about contributing data without breaching at once: the mechanism is plausible and the evidence for the size of the effect is thin. Most of the argument here is people defending the first against attacks on the second.

10 likes 5mo

g.haalandTL3Regular6 Mar 2026#65

a.zamora, post #41: Useful. I have added it to my own notes with the date on it. Go to post

Outliers should be visible in the published data even if excluded from the analysis, with the exclusion rule stated in advance rather than after seeing them.

I have separated what I observed from what I concluded, which does not always happen.

0 likes in reply to #41 5mo

il.dumitruTL26 Mar 2026#66

That is a cleaner way of putting what I was circling around.

30 likes 5mo

f.abrahamsenTL2Member6 Mar 2026#67

Worth separating two things that post #65 runs together.

Worth separating contributing data without breaching as a question about the compound from contributing data without breaching as a question about the documentation. They get answered by different people and only one of them is answerable here.

15 likes 5mo

e.coelhoTL26 Mar 2026#68

s.vanhecke, post #58: Self-reported, unblinded, self-selected data has known biases and is still worth collecting, provided every one of those words appears in the description. Go to post

Independent test results are the most valuable data this community collects, and they are only comparable when the method is captured alongside the number.

6 likes in reply to #58 5mo

s.grahameTL2Member6 Mar 2026#69

Date every record. A dataset assembled over two years without dates cannot distinguish a change over time from a change in who was contributing.

5 likes 5mo

e.roosTL26 Mar 2026#70

Small correction to my own earlier position on contributing data without breaching. I had the units the wrong way round, which changes the conclusion by an order of magnitude and therefore changes it entirely.

0 likes 5mo

i.lehtinenTL26 Mar 2026#71

s.roos, post #3: Where a value is derived rather than measured, mark it. Derived columns get treated as observations the moment the file leaves your hands. Go to post

Post #68 is right about the mechanism and I think understates the practical bit.

Reproducibility: if sharing data, include enough context (compound, dose, timeframe, method) that someone reading it understands what it represents.

3 likes in reply to #3 5mo

unit_conversionTL3Regular6 Mar 2026 · edited#72

Helpful, and easy to find again, which is half of what a good reply is.

11 likes 5mo

This topic was referenced in

Cleaning a self-reported dataset and what you throw awayData & Tools › Datasets · 2 replies
Sample size in a voluntary survey: the selection problemData & Tools › Datasets · 5 replies

Suggested topics

Topic	Participants	Replies	Views	Activity
Revisiting: Sample size in a voluntary survey: the selection problem Data & Tools › Datasets data table confounding site feedback observational data Solved Revisiting: Sample size in a voluntary survey: the selection problem — setting out what I have, and where I think it stops being reliable. I have spent a fortnight trying to pin Sample size down and I want to…		2	56k	9mo
Contributing data without breaching anyone's privacy — what changed since Data & Tools › Datasets observational data confounding site feedback data table On the subject in the title: Contributing data without breaching anyone's privacy — what changed since Working notes rather than a conclusion. What changes if the standard account of Contributing data without…		2	58k	17mo
Cleaning a self-reported dataset and what you throw away — what changed since Data & Tools › Datasets observational data site feedback data table confounding Solved Posting this under the heading it deserves: Cleaning a self-reported dataset and what you throw away — what changed since Everything below is what sits behind that. What changes if the standard account of…		2	2k	1h
A community side-effect dataset, with its response rate and biases Data & Tools › Datasets confounding site feedback data table observational data Solved A community side-effect dataset, with its response rate and biases — setting out what I have, and where I think it stops being reliable. Community side-effect dataset: setting out the arithmetic in full,…		2	5.1k	19h
Aggregated purity results across four services, with caveats — a second dataset Data & Tools › Datasets data table confounding site feedback observational data Closed Aggregated purity results across four services, with caveats — a second dataset — setting out what I have, and where I think it stops being reliable. Posting a small dataset on Aggregated purity results…	+127	151	42k	10mo

Related topics — sharing the tags data table, confounding, observational data

Topic	Participants	Replies	Views	Activity
List price versus negotiated price versus what you pay Access › Pricing formulary data table compounding Under review On the subject in the title: List price versus negotiated price versus what you pay Working notes rather than a conclusion. List price versus negotiated price — I have the observation and I do not trust my…	+99	115	11k	3mo
Area percent versus weight percent: the confusion that causes most arguments — what changed since Analytics › HPLC & UHPLC gradient method data table HPLC Area percent versus weight percent: the confusion that causes most arguments — what changed since Writing it up because I had to work it out twice and would rather nobody else did. What changes if the…	+30	34	40k	13mo
Waist circumference technique and its repeatability Outcomes › Progress reports data table long read effect size DXA Waist circumference technique and its repeatability — setting out what I have, and where I think it stops being reliable. I have spent a fortnight trying to pin waist circumference technique down and I want…	+55	60	11k	5mo
Shoulders, tangent skims, and defensible integration Analytics › HPLC & UHPLC peak integration UV & DAD detection HPLC gradient method Posting this under the heading it deserves: Shoulders, tangent skims, and defensible integration Everything below is what sits behind that. Setting out an integration problem with the actual figures, because…	+16	20	11k	13h
Immortal time bias in a claims-database study Evidence › Study critique risk of bias observational data disputed confounding Closed Immortal time bias in a claims-database study — setting out what I have, and where I think it stops being reliable. Asking about immortal time bias directly, because I have read four threads on it and each…	+29	34	28k	11mo