We were commissioned by the Electoral Commission to test the question for the referendum on Scottish independence.The overall aim was to assist the Electoral Commission in its independent assessment of the intelligibility of the Scottish Government’s proposed referendum question: Do you agree that Scotland should be an independent country? Our testing used a combination of depth interviews (with 203 participants) and focus groups (with 62 participants in 10 groups). Fieldwork was carried out between 17th November and 15th December 2012. The results of our testing played a key part in the Electoral Commission’s overall assessment of the question. Our conclusions and recommendation are shown below. The full report of our research can be read here (PDF) The Electoral Commission’s full report can be read here (PDF) Conclusions and recommendations The proposed question, ‘Do you agree that Scotland should be an independent country?’, was found to be easy to understand, clear, simple and concise. Participants were able to answer it in a way that matched their intentions. Further, almost all participants understood the broad concept of ‘independent country’ (even if they felt there were many unanswered questions about what it might mean in practice). Specifically, they understood that it meant separation from the rest of the UK. The main problem with the proposed question was that it was widely perceived to be biased – by those who supported independence as well as those who opposed it – because it was felt that the phrase ‘Do you agree…’ might lead people towards a ‘Yes’ vote. Overall, although participants thought that the question was biased towards ‘Yes’, they did not feel it would have a major impact on the result and would probably only affect those who were very undecided, unsure or “easily led”. In light of this issue, we recommend that version 2, ‘Should Scotland be an independent country?’, be used in the referendum. As was the case with the proposed question, it was found to be easy to understand, clear, simple, concise and participants were able to answer it in a way that matched their intentions. It excludes the problematic phrase ‘Do you agree…’ and, as a result, it was widely perceived to be more neutral than the proposed question.