600 is recommended


A sample size of 600 is recommended to balance feasibility, cost, and data robustness. A minimum sample of 300 and maximum of 1000 can also be accommodated.