One more question as our first larger survey starts to take shape.
I have about 50 questions on my survey, most of which are multi-select. This leads to about 120 choices for diagnoses, and the user usually selects just one or two, so the form can be filled out in less than 20 seconds. This part is fine.
I have requirements to be able to fill out forms for the UN (50 choices) and the host country (122 choices), and I am thinking about adding a column for each question to match their forms. (Note that one choice could be in both reports.) I would calculate a 1 or 0 for each answer, and these can be easily filtered and summed by our end users. Also, as noted elsewhere, I am thinking of adding a note question for each response for a summary page.
Here is my question. Is there any file size limit that I need to be concerned about? Would 350 questions (mostly calculations) in the survey and 350 columns in the data file be excessive or cause any issues? We will look into automating the reports in our central office, which could reduce the XLS size, but that would be at a later date.
@wroos. Excellent questions. Actually the survey only has one free text question, and that is for “Other diagnoses.” (And we don’t do anything with that information other than marking it as an instance of ‘other’.) Therefore, every answer is a specific response and mapped to the UN form, the country form, or both. The mapping is complex, so I think that I will need to do this to have accurate and reproducible results. I do not know how to automate this process post-survey in Excel, but that would probably be a better solution, and I will be working with our IT department. Reports are produced monthly, so this process would allow our end users to copy and paste blocks of cells from the main spreadsheet to another matching the specific report. Then using filters for age, gender, and status, the columns can be summed and pasted into the final form (Still pretty low tech.)
I was mostly concerned about the survey size. We can improve the reporting mechanism over time, but I did not want to be creating technical issues. If you have specific suggestions for improvement, please feel free to contact me by private message.
FYI, and just for fun, here is an example of the mapping that is done. Blue highlights map to one form and red text maps to another. The choices under Dehydration relate to best practices, but do not map to either form
Unless and until your survey project does not collect a photo data or use a photo questionnaire i assume the said variables and columns would not be that large which would have problem in managing your data.