Sir, I think you should perform the experiments once again and if the results still doesn't fall in line with the research papers, you can back yourself and interpret the data accordingly with your own interpretations.

You need to check for yourself :)