Acknowledging rather than arguing. The reasoning holds as far as I can follow it.
2 likes 11mo
Data & Tools · Datasets · continued
Aggregated purity results across four services, with caveats — a second dataset posts 31–60
This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1.
I read post #28 twice before replying, because I had assumed the opposite.
I have three months of notes on Aggregated purity results across four and the honest summary is that the trend is real and the week-to-week numbers are noise. I nearly drew the opposite conclusion from the first fortnight.
9 likes 11mo
NHuddleston, post #15: Publishing the raw records alongside the summary is what makes a dataset checkable. A summary alone asks for trust that nobody has earned. This is where my knowledge stops and I would rather mark the edge than blur it. 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.
Not a strong opinion, just a consistent one.
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.
I have written this out at length because the short version keeps being misread.
5 likes 11mo
Everything in post #32 holds. The case it does not cover is the one I have.
Where a value is derived rather than measured, mark it. Derived columns get treated as observations the moment the file leaves your hands.
Not a conclusion. A place to stand while looking for one.
14 likes 11mo
dexa_twice_yearly, post #13: Units in the column header, always, and the same units down the whole column. Mixed units in one field is the commonest defect in shared spreadsheets here. Go to post
Second-hand on Aggregated purity results across four, so weight it accordingly — someone whose method I trust told me this and I have not verified it myself.
j.sandvik, post #38: Second-hand on Aggregated purity results across four, so weight it accordingly — someone whose method I trust told me this and I have not verified it myself. Go to post
Units in the column header, always, and the same units down the whole column. Mixed units in one field is the commonest defect in shared spreadsheets here.
Collapsed as off-topic by two members at trust level 3 or above
Sample size is not the only thing that determines what a dataset can support. Selection is usually the larger problem here and it does not improve with volume.
Two people can read the same figure differently here and both be reasonable.
Good question, well framed, and I would like to see it answered properly.
23 likes 11mo
g.ekstrom, post #28: The most useful dataset this community could hold is boring: lot, supplier, service, method, date, result. That is enough to answer most of the questions people ask badly. Go to post
Picking up post #39: that is the part I would want checked first.
On Aggregated purity results across four the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that.
I had written a reply contradicting post #42 and deleted it. Here is what survived.
Aggregating first-hand accounts does not produce evidence of the kind a trial produces. It produces a description of who chose to post, which is a real thing and a different thing.
The interesting part of this is the exception, and I do not understand the exception.
3 likes 11mo
Combining results from different testing services into one column loses information that matters. Keep the service as a field.
Worth one more sentence than it usually gets.
0 likes 11mo
g.pemberton_uk, post #11: Thank you for the correction. I would rather find out here than later. Go to post
Small methodological point on Aggregated purity results across four: repeating a measurement is cheap and resolves most of what is being argued about here at no cost to anyone.
Coming back to post #45, because the follow-up matters more than the original answer.
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.
On reflection I would soften that slightly.
1 like 11mo
Post #45 is right about the mechanism and I think understates the practical bit.
Temporal bias: older data in a dataset might reflect conditions (supplier, formulation, context) that have changed. Newer data is more current.
If that reads as pedantic, it is, and it has saved me twice.
0 likes 11mo
Wendelboe, post #17: Adding the measurement that post #16 says would settle it. What I want from this Aggregated purity results across four thread is the list of things that would need to be true for the claim to hold. If we can write that list, we can check it. Go to post
Posting my Aggregated purity results across four numbers with the method attached so they can be discounted properly. Uncontrolled, unblinded, and collected by someone who wanted a particular answer.
t.dumitru, post #7: Aggregating first-hand accounts does not produce evidence of the kind a trial produces. It produces a description of who chose to post, which is a real thing and a different thing. Go to post
Adding thanks rather than a view. I do not have a view worth the space.
taper_table, post #21: Everything in post #19 holds. The case it does not cover is the one I have. The reason Aggregated purity results across four keeps being re-asked is that the answer is conditional and people quote it without the condition. It is not that the answer is unknown. Go to post
Post #52 is the version of this I will quote in future. One addition.
The bit of Aggregated purity results across four that nobody enjoys is that the answer changes depending on what you are trying to decide with it. Say what the decision is and the thread will converge.
Where I part company with post #51, and it is a narrow parting.
Reframing Aggregated purity results across four slightly, because I think the disagreement is about the question rather than the answer. If the question is "does it happen", yes. If it is "how often", nobody here knows.
12 likes 11mo
Leiterman, post #29: I would put moderate confidence on the mainstream reading of Aggregated purity results across four and no more. That is not scepticism for its own sake; it is where the sourcing actually stops. Go to post
Confirming post #56 from a second method, which matters more than confirming it from a second person.
The arithmetic on Aggregated purity results across four is the easy part and it is where the errors are, which is an uncomfortable combination. Show your working and someone will catch it.
r.mcalister, post #42: Picking up post #39: that is the part I would want checked first. On Aggregated purity results across four the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that. Go to post
No disagreement from me. Posting only so the question does not look ignored.
Picking up post #56: that is the part I would want checked first.
A codebook describing each field takes twenty minutes and is what makes the file usable by anyone but you. Most shared datasets here do not have one.
This is the sort of thing that ought to be settled and apparently is not.
18 likes 11mo
On post #59 — agreed on the reasoning, with one qualification.
Aggregating first-hand accounts does not produce evidence of the kind a trial produces. It produces a description of who chose to post, which is a real thing and a different thing.
This is the sort of thing the wiki should carry and currently does not.
0 likes 11mo