Hello there! I am a Natural Language Processing researcher planning to post a new set of HITs soon. I have used MTurk in the past for my MSc thesis, it was a bumpy ride but more or less I am familiar with the platform and want to have a more fair experience for workers as well as ensuring quality. This time there will be HITs in different languages (spanish, english, french, arabic, chinese, german and italian). My main concern is how is the best way to ensure workers that take the task are fluent in those. I am aware qualifications can be a headache, and they limit access, and a language test may add too much complexity and time to the process. Another options is to use country specific Quals, to select that only workers from countries where those languages are spoken are able to take the task. However I am worried there may not be a pool of enough workers in those countries. Finally I can just make it very specific on the HIT title for which language the task is, and monitor the annotation to filter out those who clearly do not understand the task. Any suggestions to what would be the best approach is appreciated. The task takes around 1 min per HIT and there will be around 1000 HITs (x3 assignment each) per language. Oh and it doesn't involve any writing, it's 10 yes or no questions. Thanks!