I mainly use 3 factors when conducting a DOE, mold, melt temp and pack pressure. The biggest driver for dimension variability has been mold temperature, followed by pack pressure.