Are they equivalent across countries and languages and how have they been translated? Clearly, the framing of the question and the response categories are key to being able to make comparisons. Aside from looking at the question stem it is important to also look at how the response categories are set out. Different modes have different formats - for example as to whether don't know and refusal are offered explicitly as response options to the respondent or are only used in response to a spontaneous use by the respondent. Similarly, individual prompts, where the interviewer reads out each response category and the respondents' replies to each, are only possible where an interviewer is administering the questionnaire. Clearly, on self-completion questionnaires the respondent can see the full range of response options before answering including whether don't know is a valid response. There is a vast and growing body of literature on the effect the layout and structure of the questionnaire has on the respondent so having a good look at the questionnaire for any differences in presentation is an essential starting point.