Solved
Keeping track of randomized items for further use in the survey
Hi,
I have 20 blocks in the survey with each containing some text. The idea is to randomly select 2 out of these 20 blocks and show to the person. The person is later asked a question about these two two blocks and has to choose one of them. Is it possible to somehow connect the answer (i.e., "The second text was more interesting") to the actual block that this text was shown, as it is a crucial part of this design?
Thanks.
Best answer by mattyb513
If you need it within the survey, you would probably use Embedded Data. Set an unique Embedded Data field within each of the randomized blocks and then you can access that later with piped text.
If you are looking to access it after the fact, you can export your randomized viewing order when exporting your results. That will give you a column showing in what order the questions were viewed, from which you can derive the block order. https://www.qualtrics.com/support/survey-platform/survey-module/block-options/question-randomization/
Of course, you can also access the ED from your data as well.
It might be as simple as logic that looks like this:
!
View original
+6If you need it within the survey, you would probably use Embedded Data. Set an unique Embedded Data field within each of the randomized blocks and then you can access that later with piped text.
If you are looking to access it after the fact, you can export your randomized viewing order when exporting your results. That will give you a column showing in what order the questions were viewed, from which you can derive the block order. https://www.qualtrics.com/support/survey-platform/survey-module/block-options/question-randomization/
Of course, you can also access the ED from your data as well.
It might be as simple as logic that looks like this:
!
Thanks, that looks like something I need! However, I can't really figure out how to set a unique id for each block. And should you do it _before_ the randomization? Since judging from your screenshot, the randomization block already includes FirstBlockSeen variables, while it seems that Block1/2/3 variables were created beforehand, right?
+6The unique ID is set manually to whatever you want. In my case, I was setting it to "Block 1" literally.
You don't need to do it before the randomization because you only care about what was seen 1st and what was seen 2nd.
The randomizer will choose the Group and then run through it top to bottom. First it checks to see if you have seen a 1st block, and if not, sets that with the unique identifier for this block.
Then it checks to see if you _have_ seen a first block. We do this because there is no else condition unfortunately. It does the same as above, but backward.
Finally, we show the user the block.
You would have 20 of these groups, where the only thing differing group to group would be the unique identifier ("Block 1" in my case) and the actual block element shown.
Oh, that's makes things much clearer, thank you!
And, if you don't mind, one final question – would it be possible to refer to these blocks (the ones that were shown) later on in the survey by their embedded metadata? So that I could later on once again show these exact blocks again to the same person?
+6Hmmm... I suppose you could.
It would be a lot of logic, but basically you would do branch logic where it would check:
```
if EmbeddedData FirstBlockSeen = 'Block 1'
OR
if EmbeddedData SecondBlockSeen = 'Block 1'
then show block 1
```
This would be repeated for each block.
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.