Hi,

I have the following question. In an Online survey (LimeSurvey) we want to investigate life and working conditions in nursing homes. We want to collect data about the staff and the residents. Since the residents cannot answer themselves we ask each staff member to give answers about one resident concerning quality of life and care. Furthermore we want to avoid that different staff members answer about the same resident, since these residents would be doubled in the dataset. Therefore our idea is to include one (the first) question, where the staff members answer which resident they refer to. If another staff member would then answer for the same resident, they would receive a message that this resident has already been answered and that they could answer for another resident.

Is this possible in LimeSurvey, if yes how?
Do you see another way of solving our problem?

Thank you for your help!

Herdener

How would the first question look like? If you provide a list of residents already (no openended question), then you might be able to create quotas with the value of 1. So when someone has respondent "resident 1" already, and another one chooses "resident 1" the quota full message will be triggered (you can choose the message and the link/ where to forward).

Give it a try, but it will only work if you have a question with the list of residents. But you will need this anyway, because if you try to do it with an open ended question you will have so many different versions of the name of the residents that it wouldn't work anyway.

Maybe you could consider using tokens. Each resident would be assigned unique token and then the respondent would have to enter the token to start survey.

Depending on your setup Elisa's solution with the tokens could be the best and easiest one. You asign a token to each resident, e.g. use the room number, or ID that already exists and is known by the personal.

Thank you for your advice. Our idea was also to go for room numbers and not for names. I expect that it will not be possible to get the room numbers in advance. Maybe we just have to sort out the doubled files in SPSS.

Well, the room numbers should be pretty simple. They probably follow some kind of pattern anyway and if you know at least the patter (e.g. floor + room), then you can create a list of room numbers, even if it is not 100% accurate. Assuming that the pattern is first digit for the floor and 3 digits for the rooms you can create a list for the first floor from 1001 to 1100, for example. If there are only 50 rooms, this is no problem, the last 50 will simply not be filled. If there is no room number 13 (for whatever reason), then this doesn't really hurt. But of course, ideally you would receive a list of rooms before, which would make it easy to create a token list, with the room numbers. Even better would be a list with room numbers and resident names, because then you could use the room number as token and once they inserted the token and are in the questionnaire you could insert the name from the token attribute list via EM and could ask if the name of the resident is correct, just to double check.