Choosing a method for a two-sample statistical test

#1
Hello everyone,

I am trying to find a best method to analyze the difference between two datasets. The thing is I have a separate software where I calculated environmental impacts from production of a certain chemical substance. However, not all information was available about the actual content of the substance, so I made an assumption based on the similar substances about the amounts of the missing content. The assumption was to take an average value out of 8 different concentrations found in the open sources and use them in my calculations. And right now, I have two different samples of n size =19 with the results of my calculation before introduction of the assumed average value and after.
Summarizing all things, I would like to know what method I should use to identify a statistical difference of these two samples of the same size (n=19) under the assumptions: introduction of the average value has no significant influence on the results, or it does have significant influence on the final result. How should formulate my Ho and H1 and should I use t or f test. And should I anyhow involve bootstrapping in my calculations. I know it might be confusing because I am confused myself and a lot of methods mixed up in my head right now.

Thanks!