Jump to content

Coding and references in survey data

Recommended Posts

I have some survey data in which one of the questions we asked people about how their projects are funded. I have coded the column of this data by different nodes to represent the different types of funding using text search (over 2000 responses I had to automate it!) - grants, contracts, donations, etc. The responses are open-ended so sometimes they mention the word grant or funding multiple times.


An example is:

"We received funding from ABC and also were funded by XYZ"


If I code for text search fund*, this response generates two references, but it is from one respondent/ case. How can I do the query/ coding so that it only comes up as one reference (for the whole response)? Is there any way to merge them so that it can be seen as just one reference?


Hope I'm making sense and would be grateful for some help!!




Link to post
Share on other sites

Hi Jill


Just wondering how you are running the query and creating the nodes initially. Rather than using a * wildcard you could run the query with Finding matches set to Stemmed words which would return you similar words in the single query. You could then code them to the single node and save you the problem of trying to merge them later on or set up queries to deal with multiple nodes etc.


For example if you were to do a Stemmed word query for something like "Develop", the query returns: Develop, Developed, Development, Developer.


Hope that is helpful.

Link to post
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...