## disentangling the experimental units and treatments in an experiment

Hi having some trouble understanding what the experimental units, treatments and observational units are in the following example

A company has 4 suppliers and wishes to determine the purity of raw material from each supplier. There are 4 batches of raw material per supplier randomly sampled, and three samples also taken randomly from each batch ,measuring purity( i.e 3 replications per batch for each supplier). This is a nested design as each batch is unique to the supplier, and the three samples are unique to the batch.

Treatment: entire description of what can be applied to an experimental unit
Experimental unit: smallest unit to which a treatment can be applied
Observational unit: smallest unit on which a response is measured.

The treatment would be the combination of batch and supplier, so in total we have 4x4=16 treatments, as we are comparing the differences between each supplier. But what exactly are the experimental units?- I would say treatments and experimental units coincide as each treatment can be thought of as being applied to each of the 16 batches (as the three replications from each batch are receiving the same treatment ( i.e one combination of batch and supplier)) - can we have this?- experimental units and treatments being the same.
Also ,am I correct in thinking the observational units would be the combination of sample, batch and supplier and thus in total we have 3x4x4= 48 observational units?- or do the observational units and experimental units coincide?

Many thanks