In this lesson, we're going to continue our discussion of how to prepare for fieldwork. We're particularly going to focus on language considerations here. First off, there are a couple of different factors that you should think about when you decide what approach you want to take for your language training. One of which or perhaps the most important one would be how many languages will the interviews be conducted in? If you're just doing it in one language, the requirements are going to be very different than if you're planning to deploy your survey in an area where multiple languages are spoken. Second consideration is whether or not the languages are commonly read. Meaning, will you be able to find interviewers who can actually read the language fluently. In some cases, you might be able to find some interviewers that can speak the language fluently, but might not be familiar with how to read or write in the language. Some languages do have a written version of them, but the specifics of how to read that language can be particular. So is it possible to find interviewers who can actually read the language when it's written. One final point is, are the languages in which the interviews will be conducted, spoken fluently by their trainers. This can include any local dialects. Basically, any language that you'll be carrying out interviews in, you need to have someone who is capable of training the interviewers in that language. Ideally, on your team, you would have a trainer who can speak that language. If not, we'll talk a little bit about how to handle those situations. Some common guidelines that are important for you to follow as you go through your language considerations is first, interviewers must absolutely be trained in all of the languages in which they'll administer the questionnaire. Don't assume that interviewers will be able to translate the questions both correctly and consistently on the spot, even if they speak that language of interest, making those translations just off the top of their head can be really difficult. Make sure that they have an understanding ahead of time of how to do the translation in all of the languages that they're going to actually be administering the questionnaire in. As much as you can, try to make sure that your interviewers are trained together. Now this might seem it's conflicting with the idea that each interviewer should be trained in all the languages that they'll be administering the questionnaire, in the particular case when you're actually going to be in a region that speaks multiple languages. If you're in that situation where there will be multiple languages, one way to handle it, and the best way to handle it is that interviewers should be trained together in one common language. Usually, the official language of the country, for example Swahili or French before you split them up into different language specific groups. That way you make sure that everybody has the same base information and then specific people who need to speak specific languages have access to that specific information as well. There should also be intensive communication between the language training groups to make sure that really the same information is being relayed to everybody, even if they're not being trained in the same language at the same time. One final guideline, if the languages is commonly read, in that case, the best approach is to translate the questionnaire into the language or the languages, and then back translated. You'll translate it, for example, if you're doing it in a region that speaks both English and Swahili, you would translate from English to Swahili, and then back translate back to English just to make sure that you've done a very consistent translation. This should be done by language experts, people who have experience and expertise in doing this work. In the case that the language is not commonly read, you might want to instead just focus on translating some keywords. For example, pregnancy, injection, immunization, fever, diarrhea, some specific terminology that's relevant for your survey, and include those particular words into the questionnaire. In that case, your interviewers won't have to make any guesses or approximations for those particular words, they'll have them right there. Even if the whole questionnaire isn't translated, having those keywords can be really helpful in the field. Certainly back to the question of trainers and the trainers ability to speak languages. If the language is fluently spoken by the trainers, then ideally they should just go ahead and carry out the training in that language directly. If you're considering any local dialects, it might be useful to bring in experts in these dialects to ensure that both the questions and the response options are being translated appropriately. In that case, you'd actually be reaching out to someone with expertise in the dialects of interests. That person would come in and do a little workshop with your interviewers to make sure that they can translate the questionnaire appropriately. If one or more languages are not spoken fluently by the trainers, then you will absolutely need to bring in some external language experts to lead this process. In many countries, the best way to handle this might be to reach out to a language or literacy institute that exists in many different countries. They would be able, either they would themselves have that stuff or they might have recommendations for people that you could reach out to come in and help you with this work. In addition to translating either the questionnaire or just the keywords, depending on what your situation is, language experts actually can help you go through the questionnaire question with trainers and the data collectors and agree on how to appropriately translate each question, as well as the response options. You have to remember as you are thinking about your logistics and so on, you will need budget and time in order to allow for this language translation process to go forward. When you're considering who to select to be an interviewer for your survey, fluency in those survey language should be at least one of the criteria for selecting your interviewers. This said, it's important that you be mindful of the fact that different people interpret the word fluent and proficient slightly differently. If it's feasible, it's preferable to always test the interviewers during the selection process. If you have someone on your team who speaks that language, maybe they can have a short conversation with each interviewer to make sure that they are as fluent as they claim to be. Or another way to handle that is to give them a random survey question and ask them to translate it into the language of interests on the spot to assess firsthand what their level of fluency is. When you're carrying out the language training, you can also use this time to assess how your interviewers are performing in that language training and use that to help guide what decisions you need to make in terms of final selection. If the interviewer can't conduct the interview in the survey language, then they cannot be sent to the field. That will necessarily lead to some incorrect data if they are not able to communicate the questions and the response options correctly. That's actually very key data quality consideration for you to keep in mind.
