I have a doubt how to assess the statistical difference between ,,products. For my MSc thesis I had to ask people how they perceive the texture of some food products, both via two online surveys and a sensory test. These were performed via CATA method. differsmber of participants differ between the online and sensory and also within products (e.g. 64 people evaluated apple in the online and 20 in the sensory, 62 evaluated mozzarella online and 19 in the sensory test). I have now two contingency tables presenting the number of selections: example dataset
What I want to do is to assess if there is a statistical difference between the same product (e.g. apple) online vs sensory. Do you have any suggestions on how to do it (considering also the error correction)?