For instance, when the interviews or focus group discussions were held and tape recorded in the mother language of the research participants which is not English.
The interviews and focus groups should certainly be conducted in the language (or languages) that the participants are most comfortable speaking. In my case, when working with Russian-speakers, I use Russian, a language that I learned as an adult for my research. Though I can participate in discussions and even lead interviews in Russian, I would nonetheless get a native speaker to transcribe the interviews in the language and then would work with native speakers to understand some of the nuances of what is said before translating sections for inclusion in publications.
The interviews and focus groups should certainly be conducted in the language (or languages) that the participants are most comfortable speaking. In my case, when working with Russian-speakers, I use Russian, a language that I learned as an adult for my research. Though I can participate in discussions and even lead interviews in Russian, I would nonetheless get a native speaker to transcribe the interviews in the language and then would work with native speakers to understand some of the nuances of what is said before translating sections for inclusion in publications.
Well, it is dangerous to just go on and use the local language - when you seek ethics approval to conduct research and especially in case of translation - you are required to provide the professional translation (both ways) - i.e. from English to the other language and from the other language to English - thus no misunderstanding of substitution of any concepts, phrases or words.
Of course - but I have considered that the original research is in English and responded accordingly in line with the ongoing procedure with ethics here.
Since a major focus of qualitative research and more specifically the interview tool is content, so local language use is an advantage with some restrictions. These restrictions could be explained in terms of validity and reliability of the measurement tool. For instance, content validity where a researcher needs to make sure that zero deviation has been achieved. In other words, the content and the targeted data has not been affected neither by the use of the local language in the first stage nor during the translation process in the second stage.