if the vial has been through a bad transit you are testing the transit as much as the supplier, still working it out
#test-results 2024-09-15
Sunday13 messages5 participantstimes are UTC
Highlights from this day
- peak_split — if the vial has been through a bad transit you are testing the transit as much as the supplier, still working it out 13:12
- peak_split — unrelated but 96.8 against a certificate of 99 is fine. the same supplier landing low four times in a row is a question 13:43
- peak_split — my sheet has 8 lots from JEEP with a spread under a point, which is tight, someone check my working 13:45
unrelated but 96.8 against a certificate of 99 is fine. the same supplier landing low four times in a row is a question
my sheet has 8 lots from JEEP with a spread under a point, which is tight, someone check my working
if you are testing to settle an argument you will not settle it with n of 1
thats not great
the widest spread in my data is about 2.6 percentage points across five lots from one supplier
sorry to jump in sample handling on your end absolutely affects it. draw cleanly, ship fast, do not let it sit warm
does anyone log the negative results, the boring ones
selection bias
whats the worst result anyone has logged and what happened next
how do you separate batch variance from supplier variance
log the lot number
| Evidence | Counts? |
|---|---|
| Independent report, lot you bought, lot number legible | yes |
| Independent report, no lot number | no |
| Supplier certificate only | no |
| Screenshot of a report | no |
| Blinded pooled submission | yes, for the pool |
| A friend who said it was fine | no |
🎉6🧊17📈18
a single sample tells you about a single vial. that is genuinely all it tells you