Item non-response

Hi everyone,

I'm a rookie in quantitative data analysis. I'm working on a survey of student responses across a number of schools. There are about 8 questions, and in the data there are YES's NO's but also quite few item non-responses dotted about. (I'm pretty sure it was a box for YES and NO, so some didn't cross either).

I'm doing some volunteer impact research and so I want to choose a defensible methodology to deal with item non-responses, but ideally I would choose a methodology that maximises the positive (yes) answers. I want to help this charity get funding.

So, across all the schools, I've quantified answers to each separate question in this way: percentage of YES *amongst those who answered*. So I'm not counting item non-response as NO. I think this is okay (?), and if anyone has a term for this approach, that would be helpful to know when it comes to presenting my methdology.

However, would it be problem if I used this to count all the answers, across all the questions (percentage of YES *amongst those who answered*)? My worry is that I would be ignoring the participants who put no response for some questions, but then answered other questions. Seems like I should somehow take account of them? I'm not sure, it just felt wrong to me intuitively, maybe someone can help?

Thanks in advance!