I am currently designing a questionnaire whose variable names include A101, A102, A103, A104, A105… A126.
But my mentor had added a new question between A104 and A105 and assigned A105 to this question, which means that the original A105, A106, A107… A126 should be converted to A106, A107, A108… A127, respectively.
(Plus, variable name that I need to modify is like this：
Survey Solutions Designer does not have a feature that allows you to load a batch of variable names to assign to questions.
Instead, Survey Solutions Designer offers you a search option where you can find all existing matches in the questionnaire with the name of your indicated variable and then simply replace it with the corresponding value.
My solution is manually and I do the variable name change question by question, just as I explain it to you here
No.
Imagine loading the variable name in a batch, how would you indicate in that batch that a variable name is for a specific question if the questions have no identifier that makes them unique?
Variable names are not seen by the interviewers or respondents.
If you explain this and the fact that it will require a manual changes to a large number of variables, perhaps he/she would be sympathetic to introducing new questions as A201, or A105a if ordering matters somehow.
Don’t forget that you will also need to change all the usages of variable names in all the conditions, which is a great opportunity to introduce a problem/bug to an already tested questionnaire.
There is no batch renaming of variables, and I can’t imagine how it may have been implemented, since you would somehow need to explain all the changes to the computer, what has been added, deleted, reordered.
From other environments, the closest action i can think of is refactoring for variable name e.g. in Visual Studio. It is still done manually for each variable, but takes care of all the usages in the code. Survey Solutions doesn’t have this.
I’ve started learning programming long ago, in the times when BASIC still required the programmer to manually assign line numbers to every line of code. In those days we were numbering the lines as 10, 20, 30, … rather than 1, 2, 3,… precisely to be able to insert something between the already entered lines if code. The same approach can be used for numbering questions, but giving them meaningful names is still the best solution, (and free from the described renumbering problem).
Because my mentors often revised the questionaire with Excel, their first instinct to cope with that is what I descirbed in my first post .
“Giving single question meaningful names is the best solution”, which you have mentioned in your answer post, have inspired and confuesd me at the same time.
So when you have a questionaire with two or three hundreds questions or more, do you have any tips or tricks to give every single question a meaningful variable name rather than number them？
I will be extremely grateful if you can give me a questionaire template that you designed.
For me, it seems inevitable to number variable name when I was given a questionaire with hundreds of thousands of questions:sweat_smile:.
Not sure what is causing the difficulty here. Most people are fine with reading totalIncome, ageOfHead, numChildren. Of course the specifics of your survey may be something that I am not aware of.
Yes, that happens to me often when my quiz has new questions.
I solve it as @sergiy mentioned by adding descriptive names for the variable names. In case of following a numbering, as is your case, then what I do is add the variable with the new numbering but I add a description to the name, for example A105_Age and the next variables A105_Phone.
Note, that in this process we have to loose all the diacritical marks, which may or may not affect the possibility to recognize the original meaning.
You can consult in the forums dedicated to programming or database systems, on how it is customary in the Chinese community of programmers to name variables using English characters.
When field survey ends, the first thing for my colleagues is to clean raw data and they had often complained that variable name plagued them a lot.
They had to reference Stata’s label (by the way, setting the text of the question as a Stata’s label by default is great) to figure out which variable name is the corresponding question on the questionnaire.
So it sounds great and convenient that naming variable following Pinyin system (or Mandrin).
However, because the identical Pinyin characters in different context may have totally different meanings, confusion may also arise during the process of cleaning data.
So could you give me some specific programming or database systems whose rules to name variable is similar to which Survey Solutions applys? I am so sorry that I don’t major in CS and don’t have experience in coding
.