Paper Discussion

Experimental results often look strong on one particular setting but may not generalise. Robustness testing, such as varying parameters or data conditions, helps establish whether findings are reliable. How central should robustness analysis be to a solid manuscript?