Training On How To Ask “Disability” Questions On Censuses And Surveys

Written by Daniel Mont

Category: Implementation

22/08/2019 - 22/08/2019
A researcher gains informed consent from a research participant in Bangladesh. From Community Eye Health.

High quality data depend not only on designing effective questions and testing them appropriately, but on appropriate administration in the field. Therefore, it is very important interviewers are properly trained.

One question that often arises is, “What special training is needed for asking the Washington Group questions?” The answer is, less than you think beyond the usual (or typical) training that interviewers should receive for any data instrument.

As with any survey or census data collection, interviewers should be instructed to:

  • treat all respondents with respect;
  • read the questions exactly as written – this includes not only the question text, but the response categories as well; and
  • record the response given by the respondent and not make any assumptions about what the response should be.

Remember, the issue of disability arises in all surveys, even if questions on disability are not part of the questionnaire, because a person with disability could potentially be the respondent for any survey. So, training for all surveys should talk about how to interact with a person with a disability, for example:

  • treat people with disability with the same respect as any other respondent;
  • do not make assumptions about their capabilities;
  • accommodate people with any hearing difficulties by finding a quiet, well-lit space, or using a sign language interpreter if needed;
  • accommodate people with communication difficulties by speaking slowly, if necessary, speaking clearly, and being prepared to repeat questions or answer categories if needed;
  • accommodate people with vision difficulties by making it clear when you are addressing them;
  • accommodate people with intellectual disabilities by not treating them like children, making sure they understand you, repeating questions and answer categories if necessary, and being patient and respectful. Often people with cognitive issues are concerned about answering “incorrectly” and do not want to provide an answer the interviewer does not agree with. Therefore, it is especially important to not react in a way that suggests you disagree with an answer or find it unexpected, or through one’s tone or gestures imply that a certain response option is the obvious choice.

When administering disability surveys, training on how to interview people with disabilities should receive extra attention. Often times interviewers may be shy about asking about disability because of the shame and stigma that can be attached to it, or simply because they are uncomfortable or feel the questions are personal or will offend the respondent.

Interviewers should be trained to be comfortable asking about disability issues and to treat these questions as they would other questions on the survey. Interviewers need only be told that these questions are asking about difficulties people may have doing these selected activities for whatever reason. They are not to presume by looking at the person that they know what those difficulties are. For example, a person in a wheelchair might be able to walk. They may just have a lot of difficulty doing so, and therefore use a wheelchair much of the time – or even occasionally when the need arises.  In fact, in the training of interviewers there is actually no reason to go into great detail about models of disability in the first place.  By putting too much attention on models of disability, you may inadvertently be creating more anxiety among the interviewers about these questions.  When interviewers administer the questions in a relaxed manner, respondents are more likely to respond similarly.

It is critical that interviewers be instructed that the word “disability” should never be mentioned in the interview process. Doing so leads to under-identification of people with disabilities, as explained in much material on this website. There should be no introductory phrase that mentions disability either in the questionnaire and interviewers should not insert such an introduction on their own.

Training is very important in countries where multiple languages are used, especially if the standard practice is to have interviewers translate the questions during the course of the actual interview. Best practice would be to have official, tested translations but sometimes that is not possible (see the translation protocol blog entry). In those cases, it is particularly important that the interviewers are trained to interpret the questions in a way that identifies difficulties doing these activities, and not to refer to disability, impairments, or medical conditions.

One issue that arises with the Washington Group questions, is whether or not the interviewer should always read all four response categories for all the questions. After the interviewer asks the second or third question, the respondent may answer before the interviewer gets a chance to read the response categories, because the respondent is familiar with the response options and answers appropriately.

If respondents don’t immediately offer an appropriate answer, the default should be to always read the response categories. But if the respondent starts answering with appropriate responses, like “some difficulty,” then there is no need to strain the respondent’s patience by reading all of the response categories. Of course, if the respondent simply says “yes,” then the interviewer should say, “would that be some difficulty, a lot of difficulty, or unable to do?”

While analysts and users of the data need to be well-versed in models of disability and how to use the collected data within the frameworks of the social model of disability or the ICF bio-psycho-social model, the interviewers need only to:

  • understand very specifically what the question is asking;
  • ask the question as written; and
  • record the response, regardless of whether they think it is the “right” one or not.

And of course, to treat all people with respect, but naturally this applies to all data collection.