I need your recommendation about suitable statistic test for the following scenario:
1. I produce Solution A (original & modified) and I want to compare their pH against a gold standard solution.
2. I want to compare the pH between original and modified solution (Solution A original vs Solution A modified)
3. I used Solution A (original & modified) to detect certain chemical in plant. In order to evaluate whether the solution can detect or not, I ask 5 persons to observe the chemical reaction on the plant and compare them with gold standard one. They need to give the following grade, "Poor", "Good" and "Excellent". I want to know whether Solution A (original & modified) is good enough as gold standard, so what is the suitable test?
I'm thinking about independent t test for no. 1 and paired t test for no 2. For no. 3, I'm thinking about weighted kappa where I combine all the grades and compare between Solution A (original) with gold standard and Solution A (modified) with gold standard.
So, is my choice of statistic test correct? or is there any better test I can perform?