Data & Tools · Datasets

Contributing data without breaching anyone's privacy

Data & Tools › Datasets data table confounding observational data

Solved

Solved by m.ilunga in post #7

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

Jump to the accepted answer →

c.amankwahTL21 Mar 2026#1

Contributing data without breaching anyone's privacy Writing it up because I had to work it out twice and would rather nobody else did.

Asking about contributing data without breaching directly, because I have read four threads on it and each answered a slightly different question.

The version I want answered is the narrow one: given the method stated below, is the result within what anyone else has seen? I am not asking what it means yet.

Method, numbers and the two assumptions I am aware of making are below. If the assumptions are wrong that is more useful to me than agreement.

9 likes 5mo

m.ferrandTL1Member2 Mar 2026#2

That is a fair summary of where the discussion has got to.

0 likes 5mo

s.roosTL22 Mar 2026#3

c.amankwah, post #1: Contributing data without breaching anyone's privacy Writing it up because I had to work it out twice and would rather nobody else did. Asking about contributing data without breaching directly, because I have read four threads on it and each answered a slightly different question. The version I want answered is the narrow one: given… Go to post

Where a value is derived rather than measured, mark it. Derived columns get treated as observations the moment the file leaves your hands.

27 likes in reply to #1 5mo

desiccant_notesTL2Member2 Mar 2026#4

Trying to state the contributing data without breaching position in a way that someone who disagrees would recognise as fair, because I do not think the version in this thread passes that test.

13 likes 5mo

e.mbekiTL22 Mar 2026#5

Worth separating two things that the opening post runs together.

Self-reported, unblinded, self-selected data has known biases and is still worth collecting, provided every one of those words appears in the description.

This is the sort of thing that ought to be settled and apparently is not.

8 likes 5mo

g.tanakaTL3Regular2 Mar 2026#6

Contributing data without breaching looks different depending on whether you are reading the primary literature or the summaries of it, and the difference is not in our favour.

2 likes 5mo

m.ilungaTL2 Solution2 Mar 2026#7

m.ferrand, post #2: That is a fair summary of where the discussion has got to. Go to post

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

7 likes in reply to #2 5mo

revision_historyTL3Wiki editor2 Mar 2026#8

This follows post #5 rather than contradicting it.

Missing data should be recorded as missing rather than as zero. The distinction disappears once the file is shared and cannot be recovered.

This is the sort of thing the wiki should carry and currently does not.

19 likes 5mo

g.ekstromTL22 Mar 2026#9

On contributing data without breaching the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that.

0 likes 5mo

LeitermanTL3Regular2 Mar 2026#10

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

0 likes 5mo

c.grimaldiTL23 Mar 2026#11

Leiterman, post #10: Independent test results are the most valuable data this community collects, and they are only comparable when the method is captured alongside the number. Go to post

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

I have changed my mind on this once already, so take it as current rather than settled.

5 likes in reply to #10 5mo

dr_okonkwoTL4 Moderator3 Mar 2026 · edited#12

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

Worth separating two things that post #8 runs together.

Using data in discussions: datasets are useful as reference points when someone claims something unusual. "I have not seen that reported in the data" is different from "that is impossible", but data gives you something to say.

If it helps: the failure mode here is usually boring rather than dramatic.

14 likes in reply to #3 5mo

j.fonsecaTL23 Mar 2026#13

A dataset is only useful if the collection method is described alongside it. Numbers without a protocol are a list rather than data.

0 likes 5mo

PharmNotes_WhitfieldTL4Pharmacist3 Mar 2026#14

Include the denominator. A collection of reports without knowing how many people did not report is uninterpretable in the direction everybody wants to interpret it.

0 likes 5mo

c.lundgrenTL23 Mar 2026#15

j.fonseca, post #13: A dataset is only useful if the collection method is described alongside it. Numbers without a protocol are a list rather than data. Go to post

Posting my contributing data without breaching numbers with the method attached so they can be discounted properly. Uncontrolled, unblinded, and collected by someone who wanted a particular answer.

8 likes in reply to #13 5mo

n.nybergTL23 Mar 2026#16

Noted, and thank you for writing it out rather than summarising it.

20 likes 5mo

t.pereiraTL23 Mar 2026#17

Picking up post #14: that is the part I would want checked first.

How to contribute: if you have longitudinal data you want to add, the format is simple: date, measurement, context. Contact the maintainer of the specific dataset.

I looked this up rather than remembered it, which is the right order.

0 likes 5mo

f.villalobosTL23 Mar 2026#18

Collapsed as off-topic by two members at trust level 3 or above

Temporal bias: older data in a dataset might reflect conditions (supplier, formulation, context) that have changed. Newer data is more current.

apostille_traceTL1Member3 Mar 2026 · edited#19

Post #17 is right about the mechanism and I think understates the practical bit.

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.

That has held every time I have looked, which is not the same as always.

13 likes 5mo

k.chukwuTL23 Mar 2026#20

revision_history, post #8: This follows post #5 rather than contradicting it. Missing data should be recorded as missing rather than as zero. The distinction disappears once the file is shared and cannot be recovered. This is the sort of thing the wiki should carry and currently does not. Go to post

Contributing data without breaching sits at the boundary between what this community can usefully discuss and what it cannot, and I think it falls on the discussable side, narrowly.

27 likes in reply to #8 5mo

s.karlsen_rphTL3Pharmacist3 Mar 2026#21

m.ferrand, post #2: That is a fair summary of where the discussion has got to. Go to post

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).

Not a conclusion. A place to stand while looking for one.

8 likes in reply to #2 5mo

m.perrinTL23 Mar 2026#22

apostille_trace, post #19: Post #17 is right about the mechanism and I think understates the practical bit. 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. That has held every time I have looked, which is not the same as always. Go to post

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.

I have written this out at length because the short version keeps being misread.

2 likes in reply to #19 5mo

orbitrap_olaTL3Mass spectrometrist3 Mar 2026#23

Coming back to post #21, because the follow-up matters more than the original answer.

Anonymisation matters even for volunteered data. A combination of region, dates and a distinctive detail identifies a person more often than people expect.

0 likes 5mo

n.brobergTL24 Mar 2026#24

Picking up post #21: that is the part I would want checked first.

Second-hand on contributing data without breaching, so weight it accordingly — someone whose method I trust told me this and I have not verified it myself.

27 likes 5mo

c.cardosoTL24 Mar 2026 · edited#25

Right — I had this wrong and I am glad to have read it before it mattered.

13 likes 5mo

m.lehtinenTL24 Mar 2026#26

c.cardoso, post #25: Right — I had this wrong and I am glad to have read it before it mattered. Go to post

Post #24 is right about the mechanism and I think understates the practical bit.

State the inclusion criteria explicitly, including the ones you applied without noticing. "Records I could read" is a criterion.

4 likes in reply to #25 5mo

dr_okonkwoTL4 Moderator4 Mar 2026#27

I read post #26 twice before replying, because I had assumed the opposite.

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.

0 likes 5mo

f.sjobergTL24 Mar 2026#28

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.

If anyone has run this properly I would rather read that than my own guess.

0 likes 5mo

w.novakTL3Regular4 Mar 2026#29

On post #26 — agreed on the reasoning, with one qualification.

Self-reported, unblinded, self-selected data has known biases and is still worth collecting, provided every one of those words appears in the description.

2 likes 5mo

f.weissTL24 Mar 2026#30

k.chukwu, post #20: Contributing data without breaching sits at the boundary between what this community can usefully discuss and what it cannot, and I think it falls on the discussable side, narrowly. Go to post

Adding a note of thanks rather than an opinion. I did not know most of that.

0 likes in reply to #20 5mo

Sponsored Janoshik Analytical — independent third-party testingHPLC purity and mass-spectrometric identity on submitted samples, reported with the method stated. The laboratory most often cited in this community’s verification records.Independent testing → Sponsored GL Biochem (Shanghai) Ltd. — synthesis, directFounded 1998. ISO 9001 and cGMP certified, over 1,500 staff and 200-plus patents. Custom peptide synthesis and amino acid derivatives, with a batch certificate on every order.Custom synthesis →