Arrea a lenha: setembro chegou e já é carnaval em Olinda; confira agenda de prévias
Section_11_Statistics_Data_Polls_and_Surveys.txt
11.1 Introduction Accuracy,bingo estourado em copacabana impartiality, clarity and credibility are as important when numerical values and data are deployed in the BBC's output as they are in the rest of its journalism and other output. Content makers must be rigorous in applying due scepticism. When the BBC commissions or carries out research itself, investing the BBC's authority, the audience must be able to trust what is reported. Statistics and data are an important source of information which can result in revealing and significant output as long as the right questions are asked and there is an informed awareness of the pitfalls. Data can include information of any kind, not just statistical data, and can be acquired through many different ways such as FOI requests, leaks, scraping and open source investigation. Using data acquired by hacking may only be justified in exceptional circumstances and would require careful consideration of the public interest. Any reliable producer of such material should be able to explain the methods and workings on which it is based. Using statistics, data and numbers can create a sense of certainty for the audience but they are often imperfect representations and may contain inaccuracies and measurement error, so their uncertainties should be factored in. Polls, surveys, questionnaires, phone and online votes, vox pops and focus groups can all have a useful and fruitful role in finding out what different people and groups in society think or experience. The BBC must be clear in distinguishing between the illustrative and the measurable, and precise in its use of language, to ensure the integrity of the BBC's journalism and content. Similarly, when data – and analysis of that data – is the basis of output, content makers must ensure that the techniques and tools used are robust. This section should be read in conjunction with the guidance: Reporting Statistics.
