whats the underfill rate people are actually seeing
#test-results 2024-10-03
Thursday31 messages9 participantstimes are UTC
Highlights from this day
- VialBot — Inter-lab diff for lot KP-0925: 99.4% vs 98.6%. Within expected range. 18:18
- salt_bridge — my sheet has 5 lots from WXT with a spread under a point, which is tight, n of 1 obviously 19:25
- VialBot — Testing queue: 8 submissions open, 32 awaiting dispatch. 19:54
- monoisotopic — i told GGPeps about a result they would not have liked and they did not argue. credit where it is due 20:00
if you are testing to settle an argument you will not settle it with n of 1, still working it out
log the lot number
🤝9👍1🧪15
do you test every order or sample it
one bad test does not condemn a supplier. one good test does not clear one. both halves get ignored equally
a single sample tells you about a single vial. that is genuinely all it tells you
unrelated but how do you separate batch variance from supplier variance
quick one does anyone log the negative results, the boring ones
one data point
🤝1
thats a fill problem
😂1📉1
a result without a lot number and a purchase date is an anecdote with a pdf attached, from memory
if the vial has been through a bad transit you are testing the transit as much as the supplier, take that with a pinch of salt
test it again
Inter-lab diff for lot KP-0925: 99.4% vs 98.6%. Within expected range.
does a bad result on one lot condemn the whole supplier
certificate said 99 and the independent came back 98.6, is that a problem
[edited]is quantity testing available as a separate thing or bundled
my result contradicts the supplier certificate, do i tell them
how much does sample handling on my end affect the number
my sheet has 5 lots from WXT with a spread under a point, which is tight, n of 1 obviously
🙏1
i test roughly every tenth vial. it is a sampling strategy, not paranoia
the widest spread in my data is about 2.6 percentage points across five lots from one supplier, someone check my working
thats within range
weigh the vial before you reconstitute. it costs nothing and catches the common failure
genuine question what interests me is when the gap goes the same direction every time. once is noise, three times is a pattern
Testing queue: 8 submissions open, 32 awaiting dispatch.
i told GGPeps about a result they would not have liked and they did not argue. credit where it is due
a point or two between the certificate and an independent result is inside what two labs disagree by
the results in this channel skew because people test when they are suspicious. that is selection bias and it is real, happy to be corrected
slightly off topic but i have never regretted spending the money on a test. i have regretted not testing twice
98.1 against a certificate of 97.4 is fine. the same supplier landing low four times in a row is a question
selection bias