I recently sent a survey out that had some quantifiable data, such as how many times have you talked with or worked with in the past 6 months? I used a scale type question and got responses. However, when I generated the relationship graph, it doesn’t show the responses or data? Am I doing something wrong?
1 Like
When you generate networks from a survey question that collects numerical data (i.e. a numerical question type or a numerical slider question type) you will find that that numerical data is added as attributes on the edges for the generated network. Within the explore view please navigate to Size and then Edges. If you then select Weight you should see your edges have been sized by the weight.
Once edges have been sized by the weight you can then use edge weights for those metrics where it is applicable (for example the three different types of Degree).
Thank you for that! With regards to the matrix of choices, how does that data get represented? For example, I asked a question, how often did you work with X over the past year and had options 0, 1-2, 3-5, 6+ but that data did not show up. Should I have used a numerical slider? And if so, what is a matrix if choices question used for and how is it displayed?
You could indeed have used a numerical or numerical slider question there but a matrix question is also an option. You generally use a matrix question when you have categories that you want the respondent to select from (e.g. Daily, Weekly, etc. or Agree, Disagree, etc.). With a matrix question, the response will also be available as an attribute on the edge but it won’t be a numerical attribute. Rather, it’s a categorical attribute. That means that you won’t see it if you navigate to Size then Edges because only numerical attributes are available there. You will see it though if you navigate to either Filter then Edges or Color then Edges.
It’s also possible to export your network and, in Excel, convert the categories on the edges into numbers if you so choose. You can then update the network data and will have the numerical attribute available to you.
It’s also worth noting that if you allow respondents to select multiple values per row in the matrix question then you will have multiple attributes (one for each column) in the network. If you allow only one selection per row then you will only have one attribute for each edge.