input output x1 -> +---------+ -> y1 x2 -> | Process | -> y2 x3 -> +---------+ -> y3 QA QC
The above process model could for instance be for an injection moulding process. On the input side we would have some process parameters like temperature, pressure, humidity, and timing. On the output side the dimensions of the part being moulded, which we can call attributes of the part. That is, process parameters on the input side and product attributes on the output side.
If the input is stable then all the output is stable and it follows that variation in input causes variation in output.
Quality Assurance (QA) is done by checking the process parameters and since this happens before the product is produced this assures the right quality. Quality Control is done by checking the product attributes after the product is produced and control is established by sorting in non-conforming and conforming products.
By including all process parameters on the input side (e.g. the maintenance standard of the injection moulding machine, the ambient temperature, the quality of raw material) the process in the model is only the sum of the input process variables and thereby reduced to cause-and-effect.