Piped Text for Different Languages Isn't Working Correctly | XM Community
Skip to main content

Hi everyone! I’m hoping you all can help with an issue we’re encountering regarding embedded data fields, branching, and a question randomizer.

Our survey can be taken in English or Spanish, so we have an embedded data field (Q_Language) capturing this information at the start of the survey. While we have used the Qualtrics translations feature for the majority of our translations, there are two questions in the survey where we want to adjust our question text based on the respondent’s answer at a previous question. We have assigned question text via embedded data fields depending on if the respondent selected “No, I don’t” or “I’m not sure” at a previous question and also if they’re taking the survey in English or Spanish. From there, the respondent is randomly assigned to an open text or ranking question, utilizing the embedded data fields they were assigned in the question text.

While this appears to work correctly in our testing, we have seen a large number of situations in which the question text appears blank, leading us to believe that our branching or embedded data fields are not set up correctly. Could this be due to the order of our branching logic and when the respondent’s language is recorded (note this block is towards the end of the survey)? Also, any ideas as to why this is working the most of the time but not always would be much appreciated as we want to replicate this issue so we can be 100% sure that this is resolved once we find a solution.

This is replicated for each answer option / language combination

 

Make sure you don’t have the embedded data and the question where you display the embedded data is not in the same group. 

Meaning:

If you set the embedded data based on the first question and if you want display the embedded data in the following question, 2nd question should not be in the same group or block. Create or move the following questions to the new block.

Is the question on which you base your displays and logic mandatory?

 

​@jbk These are all handled in separate blocks - we have separate branching for each combination and then separate branching for the randomizer and corresponding questions. We aren’t using any groups to separate these out differently though.

 

​@vgayraud Yes, that question is required.

In your screenshot above, I can also see the embedded data incentive_Question is under the randomizer. Is this embedded data is also part of display logic.

​@jbk No, incentives_question is not part of the display logic. That is for our tracking purposes so we know if the respondent saw the rank or open text option under the randomizer.

​@jbk No, incentives_question is not part of the display logic. That is for our tracking purposes so we know if the respondent saw the rank or open text option under the randomizer.

Is this embedded data working? Better you can move that under block “Future Purchase Intent” just to be on the safer side.

 

Coming back to your actual question, you said translation works in your testing right?

​@sklock,

I’m not sure what is causing your issue, but a few observations:

  1. You don’t need to set Q_Language, it is set automatically from the contact list (Is Language set for all you contacts?)
  2. It is best to avoid setting embedded data fields for piped text translation purposes. A better solution is to create a non-displayed multiple choice question where the choices are the piped text. Then just pipe the choice descriptions as needed. This makes the piped text part of the Qualtrics translation and the correct translation is shown when the language is switched.
  3. If you do use embedded data for piping purposes, you need to disable the ability of respondent to change the language (i.e., hide the language selector).

​@TomG Appreciate the response! We don’t use a contact list and instead send the surveys out via an anonymous link. So in this instance, we default to English, but allow the respondent to switch to Spanish if they’d like.

As for the embedded data fields, we’re looking to create custom question wording with piped text. Is there a reason why this wouldn’t work with Q_Language incorporated into the logic? It seems to be piping in correctly sometimes, but not always and is not specific to language selection or the previous response that was selected.

 

Remove ${m://Language} from Q_Language = ${m://Language}. It is not needed. I’m not sure, but it may be the source of your issue if m://Language is blank. If Q_Language were set to blank then your embedded data fields wouldn’t get set.

 

Leave a Reply


Terms & Conditions