Exact yet Subjective (Is this a figment of my imagination.)

[Triple Quad]

I've had this thought in my mind and I've tossed it around with some co-workers. We all work in method development. I'm new in method development in a real lab, fresh out of grad school.

My brief experience suggests to me that in the exacting world of chemistry trace analysis, there are yet surprising (to me) degrees of subjectivity.

One guy here is working from a 10 year old SOP and having trouble getting the method to work.

It comes down to this, give a certain project to 2, 3, 4 different analysts let them work independently. They will come back with three similar yet different methods. Most likely all will produce quality, precise and accurate results.

One guy simply prefers to standardize all his LC methods as much as possible with water and methanol. Another prefers, Water and acetonitrile.
The third who knows...

We all pour our blood, sweat, and tears into our projects to produce results that we can stand by. But, a supervisor with a different set of experiences, education, analytical preferences looks at the results and the method, then says the results are OK...but I would have done it differently. I like my way better.

Does this happen to anyone else?

When to push back and when to let the door swing in your direction various from situation to situation.

[Peter Apps]

Simple; run it both ways and compare the results.

Peter

[gcguy]

When developing methods for our QC department we have to be aware of what kit they have and what their technical capabilities are. This inevitabily leads to methods being a compromise of what the "best" method would be versus what "will do".

GCguy

[GOM]

Don't confuse precision with accuracy.

Regards,

Ralph

[Kostas Petritis]

A good supervisor should be able explain why the way he/she is suggesting is better. I personally like supervisors/managers that still maintain some hands-on experience. Otherwise it is easy to loose contact with reality and develop unreasonable expectations.

[Bruce Hamilton]

The only real issue should be whether the desired outrcome was achieved within budget and on time.

In general, there may be many routes to the same destination ( the required analysis specification ), but the work specification should constrain the options eg:-
- what instruments are available ( pumps, UV, RI, MS, ELSD detection etc ),
- what mobile phases are acceptable ( if the system uses normal phase, or a standard reverse phase system ( eg 0.1% TFA in H2O:CH3CN ), those could be preferred for instrument utilisation reasons ).
- what columns can be used ( some columns types are more sensitive to dying from junk than others ).
- what other rubbish is in the sample, thus determining the type of sample preparation.
- What analytical cycle time is required ( an IPC method may be limited to 10-15 minutes ).
- any constraints on material and labour costs,
- etc etc.

The analysts can use their experience, and that of others ( literature searches, discussions, etc ) to reach the desired outcome.

The customers usually only want to reach their destination quickly and cost-effectively, and they don't care about the 15 possible routes, or the various cul-de-sacs that method development analysts inadvertently managed to drive into.

Please keep having fun,

Bruce Hamilton

[tom jupille]

But, a supervisor with a different set of experiences, education, analytical preferences looks at the results and the method, then says the results are OK...but I would have done it differently.

A lot depends on the context and the tone of voice. I've been on both ends of that discussion over the years . It can be a great springboard to learning for both people as each explains their rationale.

I'll rephrase Bruce's comment as it would have come from one of my old professors: "There are m ways to skin n cats where m > n". There are enough variables in developing analytical methods that there is no way to do a global optimization in any reasonable amount of time. In order to get anything done, we have to take take short cuts and apply rules of thumb ("heuristics"). We get into this all the time when we teach our method development courses; there's no problem with different people applying different heuristics so long as they are applied consciously and have a rational basis.