The philosophy of some people is the welder only needs to be capable of depositing sound weld metal.
I have seen shops that tack weld the test assemblies in advance so the welder does not have to set up his own test coupons. The machines are set for him. The test assemblies may even be in the test position before the welder shows up to test (sounds like an ATF to me).
There are circumstances where the welder is only required to deposit weld metal in a specific position, of a certain type, a certain size, and for a certain length. The parts arrive at the workstation in a fixture ready to weld. That production welder does not have to see the welding machine, it is already set to weld. That scenario is exactly what was described to me for a shop located in Mexico. The welder only knows how to make the one weld and is only expected to make that one weld.
With that scenario in mind, how would you qualify the welders with one last thought in mind. The turn-over rate is 200% per year.
It is the Fredrick Taylor method of manufacturing.
My opinion is that the qualification test should be reflective of what the welder is expected to do in production. If the welder has to layout and tack weld components on the shop floor, then the welder should layout and tack weld the test assembly. If the welder is expected to set up the welding machine as production conditions dictate, then the welder should set up the welding machine for the qualification test. There is little to be gained b qualifying a welder under a set of ideal conditions only to find out the welder cannot work under production conditions. I do not believe in sending the welder to a laboratory to be qualified. I test the welder in the shop where he is employed using the same equipment he will be using in production.
I guess it depends on the goals established by the entity testing the welder. Is the goal to turn out as many "qualified" welders as possible? There are some schools that have qualification rates of nearly 100% at graduation. Of course the same welders cannot run a fillet weld because all they practiced in school was 1-inch groove welds. There are a couple of labs in the area that have pass rates way above 80%, but the same welders can't make a decent weld in the field. Technically, both situations just described are meeting the letter of the code, but they are not doing anyone any favors by qualifying people that are not competent. That's why I test welders on the site. Are all welder qualification tests equal? No, not by a long shot. They are fair, but not the same because not all welding jobs are the same.
Best regards - Al