Same trim was sent to two test labs, Why such a ‘big difference
Not homogeneous?
3 Likes
I know there’s a lab in canada that some lp’s send there flower to to get tested that always comes back a few points higher, I’m sure someone will catch on soon enough.
How’d you do your sampling?
2 Likes
The only lab LPs can use is A&L
Some are using high north labs.
4% variance is common imo
Homogeneity perhaps.
But also - measurement uncertainty can get you upwards of 20% different. So for 9.5 that would be 1.9% and for 13.5 that would be 2.7%. Its even worse between labs - because that uncertainty increases from 20% to 35% most of the time (there’s some good studies on interlab variance, if you are interested HMU)
When I’m doing large batches of things like trim (which are very rarely homogenized nicely) - I send in 5 samples and then take the average of the results for my mass balance and process expectations. And I almost always go with a single lab - or run my stuff myself in house. That way I reduce the variation in the sample preparation and instrument measurement uncertainty.
Last year…maybe the year before. Ha! I did a webinar on lab testing results and the things that cause these issues. It can be as simple as the reference standard is different and as complicated as one lab is using HPLC and the other is using GCFID.
5 Likes
A&L is the only health Canada certified lab, unless they’ve got a new one. But to my knowledge they dont
Are you even using a GC man?
This would be one of the reasons for getting in-house analytics early rather than late.
It’s much easier to understand the amount of variance in your input when you can run as many samples as you want to. You can (for instance) go grab a branch and test several flowers to assess variability. Blew my mind the first time I did that and had 20+% at the tip and less than 10% on the bottom. most of those buds would have ended up in the same bin…
If you really want to assess inter/intra lab variability you need your test samples to be as similar as possible (homogeneous).
I recommend this implement… Biomass Grinding at various scales? - #22 by cyclopath
4 Likes
Idk but there’s been a running theory that most cannabis dispensaries get their stuff tested at every local lab or at least 3 in the state and then pick the one that tests the highest.
Except that it’s generally the producer not the retailer that picks the testing facility (in mature markets).
Absolutely evidence of inflated numbers driving the market, to the point that labs not playing the game can see their share of the testing market shrinking and are forced to choose between accurate reporting and actually surviving…
2 Likes
That’s so fucked up
Yep. If you want to know it’s being done right, doing it your damn self is probably the best solution.
You’ll still have to play the third party lab game, but at least you’ll know who is cheating and who is just lousy at their job.
Also helps identifying the honest & competent labs too…
3 Likes
many states don’t let you send to more than one lab now. especially established METRC states. No more shopping.
In my state only one sample AND some rando dude comes and takes it. x.X
2 Likes
That’s true, but even in that scenario, if you try a few different labs for a few different batches, you’ll quickly find which ones are giving higher numbers out.
1 Like
Or you could just ask them about their recovery percentage from their validation. Either they did it and will give you the number… or they have no idea what you are talking about and you should look somewhere else for better science, IMO.
When I was in Arizona, there were two labs that had seriously high numbers. On inspection, one of those labs had not actually opened any of the samples we had sent! Talk about bogus numbers, never working with them again and so happy that the state was not yet mandating testing!
On inspection of the other lab - I found that they were running HPLC with a run of about 4 minutes. They were lumping all peaks into either CBD or THC - all of them. Chromatography, intentionally messy so they would push out higher numbers. Stopped using them too.
Then I instituted a quarterly check against the labs. Always looking for an internal RSD of 2.5% (which almost all of them could meet) and an intralab RSD of 15% (just like you see in the standards). Any who couldn’t cut it for whatever reason, got an audit by me. Sometimes that was good - helping people resolve issues. Sometimes it was just frustrating (because I had to reduce the number of labs we were using…)
In the end - I landed on three out of 11 labs I had tried working with. They all reported consistently within a few points of each other. I could rely on their timelines and customer service if issues came up. I have not found that same rapport with labs in Michigan yet. I hope to do so - mostly people don’t seem to have any idea about the kinds of basic scientific method questions I’m asking. The state is supposed to have checked all of them… but I have my doubts that anyone has done any validation at this point (because I haven’t seen any evidence to convince me otherwise). 
Anyway - I figure everyone should be doing their due diligence to vet their lab. But I know many do not - or don’t know how - or really don’t care. Such is life.
5 Likes
Very interested in inter lab differences
Thanks
@Future what about that Purple gadget at used at the GLG Denver? Would that be better than nothing for heads up potency reports?
my gumies seem to have a strong effect vs store bought? I wanna get it tested?