When to use multilingual function vs separating into multiple experiments for multinational survey?

You should use the multilingual function when your survey is identical across countries.

Sometimes, it is better to include all languages into one experiment for the following reasons:

  • Makes it easier to look at aggregate results (average scores across countries)
  • You can use the segmentation function to look at results by country.

However, this can only be done when the survey is identical across countries:

  • When there are no questions on the brand list (as countries typically would have different brands).
  • When you are not looking to do subgroup analysis in conjunction with a split-sample component (having a randomisation block that shows X number of questions out of a larger N number of questions).
  • When prices are comparable.

Otherwise, it would be better to separate into multiple experiments to get cleaner, more focused results, you can do this by duplicating the experiment and then change the inputs where necessary.