Explaining different process variances

Bmi

New Member
#1
Hi everybody,
The situation:
I'm stuck with the following problem: some measurements are done on a number of batches. These measurements are assays and are done on a number of batches giving different concentrations, batch concentration. The assays themselves also produced control concentration, independent from the batches. Every assay has one control concentration and 2 to 3 batch concentrations (same test setup done for different batches).
The problem:
Both batch concentration wrt batch number (ordered to reflect time), control concentration wrt assay number (also ordered to reflect time) and batch concentration wrt assay number (as said, multiple values for one assay) shows great variance and an increasing positive trend. (Significant positive linear regression slope coefficients).
The question:
Are there any tools I can use to see which variance is attributed to which process knowing that batch concentration is NOT independent from the assays. Both batch and control concentration were normalized with another variable called peak area, that's why batch depends on the assay and control just depends from the assay.

I want to test the hypothesis whether the batch variance is mainly explained by the assay variance, ie. Variance in all the data is (mainly) not caused by the batch manufacturing process.

Thank you!