Creating variables with bucketing prior to data collection | XM Community
Skip to main content
Our questionnaire is primarily composed of items on a 4-point bipolar agreement scale. Additionally there is an "I don't know" response included.



The key metric we want to report in our Qualtrics public web reports is percent agreement for each item which combines the "strongly agree" and "agree" choices. This percentage excludes the I don't know answers. We also want to report frequency of all responses including "I don't know" so I am not excluding it from analysis.



Once data is being collected I know I can create a variable with bucketing that combines the "strongly agree" with "agree" and "strongly disagree" with "disagree". This then shows up as an embedded data in the reporting section. But there is no option to do this prior to data being collected and this created variable doesn't carry over if I create a copy of the project.



Would there be a way to create a similar "bucket" variable some embedded data in the survey flow prior to data collection and so that it can copy for future iterations of the survey? Or am I missing something in the web reports that allows this sort of analysis without needing to create a new variable?
@romerton You can can use branching logic to create the bucketing as embedded data in the survey flow:



E.G. if you variable name is NETS you would have three branches

1. If Question is strongly agree or agree then NETS = agree

2. If Question is strongly disagree or disagree then NETS=disagree

3. If Question is don't know then NETS=don't know



Under Options for the embedded data field you will want to change it from Text to Text Set so it will create a frequency for these three values when you are analyzing.



See sections "Embedded data options" and "Setting Values in the Survey Flow"



https://www.qualtrics.com/support/survey-platform/survey-module/survey-flow/standard-elements/embedded-data/#EmbeddedDataOptions
@bstrahin that makes sense. I did not know that about text sets.



It will be a gnarly survey flow since this is for over 100 items like this. I will look at the exported .QSF file with a few of these built and see what the code looks like and if it is something that I can build and then import.

Leave a Reply