how do you separate batch variance from supplier variance
#test-results 2024-11-17
Sunday21 messages6 participantstimes are UTC
Highlights from this day
- warsaw_vial — if the vial has been through a bad transit you are testing the transit as much as the supplier 20:17
- LC_MS_Lena — post the lot, the purchase month, the lab, the number, and what you were expecting. that makes it useful later 20:21
- point_two_five — what interests me is when the gap goes the same direction every time. once is noise, three times is a pattern, still working it out 22:13
how many lots do you need before you can say anything about a supplier
if the vial has been through a bad transit you are testing the transit as much as the supplier
✅3
update on the earlier thing if you are testing to settle an argument you will not settle it with n of 1
post the lot, the purchase month, the lab, the number, and what you were expecting. that makes it useful later
medutest-report-d-0718.pdf
3 pages · 245 KB · not retained in the public archive
🎉1
log the boring results too. a channel that only records failures produces a false picture
update from 11 months ago: retested the same supplier, result came back better, logged both
the widest spread in my data is about 2.6 percentage points across five lots from one supplier
a single sample tells you about a single vial. that is genuinely all it tells you
one data point
log the lot number
thats a fill problem
the honest limits: hobby chain of custody is weak and any result here should be read with that in mind
do the results in this channel skew because people test when suspicious
selection bias
🧪2
what interests me is when the gap goes the same direction every time. once is noise, three times is a pattern, still working it out
when did you buy it
quantity testing is usually a separate line item and it is the one worth paying for
test it again
which lab
does anyone log the negative results, the boring ones