Analyze related surveys with Stats iQ | XM Community
Skip to main content

Hi,

I have two longitudinal surveys.

I have done the setup of the two surveys like described on https://www.qualtrics.com/support/survey-platform/common-use-cases-rc/pulling-data-into-a-second-survey-longitudinal-surveys/

So I have two datasets, both containing the field “Login ID” by which the two surveys can be matched.

Is it possible, and if so how, to run Stats iQ on both surveys, using that Login ID to relate the answers to eachother.

So for example, in Survey 1, I have the field Login ID and the question “how satisfied are you with A”.

And in Survey 2, I have the field Login ID and the question “how satisfied are you with B”.

I want to know if there is a correlation between A and B.

 

I used a mock dataset, where I purposely had the answers on A and B that belong to the same Login ID be similar, so I know that there should be a correlation shown on this mock dataset. I added 51 answers on both surveys.

 

I have tried adding the two surveys into a Dashboard,

Both with the Data Mapper and the Data Modeler,

and in the Data Modeler both tried using a Union and using a Join,

but whatever I do, Stats iQ keeps saying that it can’t relate or do a regression:

 

51 rows (50%) of the sample were missing at least one cell of data. Stats iQ automatically imputed the cells to enable the rows to be included in the model. To learn how data was imputed for each input variable, explore the model in detail below.

Automatically impute missing data for your sample

The regression could not be performed because there was no data, either due to restrictive filters or variables having many missing cases.

To fix this, relax your filters and remove variables with a lot of missing cases from the model.

 

Is there any way in which I can still do this?

Thanks!

@JoycaV A good indicator showing that you’ve Joined the dataset successfully is the number of response you have. For example: You have 51 responses on A survey with 5 field and 51 responses on B survey with 6 field. If you join them successfully base on a common ID, you still gonna have 51 responses at the end on 11field.
In your case, I’m guessing youre having 102 responses at the end. So the A side is missing the other half and viceversa (that explain the 50% missing data). There’s no row contain both A and B, they dodge each other perfectly.

If Joining them in Qualtrics is too complicated. You can export them out to excel and use “VLOOKUP” to join the data or Import both of them to ChatGPT if you have it and ask the AI to merge the dataset. I’ve done dataset contain hundred of thousands data with ChatGPT without breaking a sweat.
Import the Excel back to Qualtrics as Imported Data and do your analyze.

Hope this help

Eventually it seems my data join worked after all, as I tried again today. Stats iq mentioned first it was adding and updating some fields. So I guess that after I had applied my data Join, it just took a while for those edits to come through until the stats iq.

 

@JoycaV Oh yeah, the data update part. Sometime it’s automatic, but sometime it’s not and you have to do it manually. That confuse me also

Leave a Reply


Terms & Conditions