How to deal with analyzing optional survey data

Question

Consider this scenario. There is a website that ranks every type of sandwich. For each type of sandwich, users fill out a survey about the following: A) how tasty it is (on a scale from 1-5); B) how filling it is (on a scale from 1 to 5); C*) if it is spicy (-1 or 1).

The website displays the average value of the answers to Question A, the average value of the answers to Question B, and the sum of the answers to Question C.

Question A and Question B are mandatory. Question C, however, is optional. I have the number of user-submitted ratings for each sandwich, and I have the sum of the answers to C.

I want to compare the spiciness of different sandwiches. Since different sandwiches have different rankings, I cannot simply compare the sum of the answers to Question C. I also cannot weight it as "Sum of Answers to Question C/Number of Ratings" since Question C is optional and some people won't answer it.

What would be the best way to go about comparing the spiciness of different sandwiches under these constraints?

Answer 1

Your best bet is to change the way you ask question C and the way you measure it. Giving a yes / no response for C will not let you gauge the actual spiciness so you need to adjust it to a scale and allow individuals to rank it. While all of your questions are fairly subjective, C is the most so as tolerance to spice plays a big role in if one thinks it spicy or not.

I suggest using a 1-5 scale, as you have for the others, and give comparative suggestions for each level of the scale e.g. 1 is eating bread while 5 is equivalent to eating a habnero pepper.