I am conducting a sociocentric study with multiple name generators. I noticed that once an individual is nominated on one name generator they are not available for nomination in subsequent questions. I have read a way around this in the documentation i.e using a Name generator (quick add) interface.
I have done this but my challenge with the Quick Add is that it does not allow searching the roster during nomination. My population includes people who struggle to remember the details of their social contacts with accuracy so I have added other identifiers such as nick name, area of residence as a solution. Is there a solution to this?
I found this statement in the documentation " *We recommend that you add all the nodes you can to the alter pool before adding additional details so that you can add these details to all the nominated nodes rather than zigzagging through the protocol" which I thought could be a solution but I am not sure how best to implement it. Is there a way of allowing participants to nominate all their social contacts (using the Name Generator for roster data) without assigning them to a specific name generator initially and then later on using the side panel to allocate the individuals to specific name generators?
Welcome to the community I’m Joshua, and I lead the design and development of the software.
Correct. In Network Canvas, nominating a node from an roster adds it to the interview network. Once a node is in the network, there is no longer a need to show them in the roster. This is an intentional design decision.
I wouldn’t think of this as a workaround, but rather as a different workflow entirely. The quick add name generator is a free recall task. You can use it with roster data (via a side panel), but the intent behind that feature is for it so be used in scenarios where you have a partial or incomplete roster, and you want participants to be able to add missing/additional nodes while also retaining the ability to add from a roster.
If you have a complete roster, you should use the dedicated roster interface.
I think your issue comes from the fact that you are mixing up name generation and attribute generation. You want to be able to say something like “Who provides emotional support?” and then store the attribute provides_support along with the nodes that were nominated. This isn’t how we intend the software to work.
There are two solutions:
Separate out nomination and attribute assignment. Have a roster stage where prompts are based on some more generic question, followed by one or more name interpreter stages that assign attribute data. This is what is being referred to in the documentation passage you quoted.
Use the name generator quick add with two side panels. Panel 1 shows your roster, and Panel 2 shows your interview nodes (“people you already mentioned”). You can then use the additional attributes feature to combine attribute and node nomination. However the drawback is that you won’t be able to search or sort the roster, and the interface will be more cluttered. You also cannot disable the ability to nominate alters outside of your roster.
Thank you for the helpful response. I still need clarity on a few things.
My name generators look something like this “Who are the people on this list who could give you information or advice in relation to an important life decision, such as a health or moving to another place?” I have 5 of them, each asking for nominations of people providing support to the participant in different domains. Is this what you are referring to as attribute assignment?
At the moment, I have prepared a more generic name generator (using roster data)- I am able to search as desired. This is followed by 5 name generator stages using the Quick add (with side panels). I have assigned a variable to each of of the name generators to manage to isolate them when I collect the data. I have an name interpreter stage asking about the frequency of contact, etc. What would you advise?
I would say that those “domains” are implicit attributes, so yes. You aren’t assigning a variable to track those attributes, but they are being implicitly recorded via the stage/prompt that the node is nominated on.
Really, the two options above are yours to choose between.
The two side panel approach will give you the behaviour that you want, in that you will be able to “see” (and re-add) nodes that have already been nominated - albeit with the drawback that you won’t be able to sort or search the main roster, and the size of the panel is much smaller.
The approach where you do the roster nomination via a more general name generator (something like “Who has provided you with support of any kind? E.g emotional, financial, material”), and follow that up with one or more name-interpreters will keep the search and sort behaviour, but at the cost of changing your name generators.
I’m Joshua, and I lead the design and development of the software.