IJMR special issue

The USA has a new president-elect this week, but was this the result that the USA pollsters expected?  The polls were showing a close race for the White House, but the majority predicted a slight Clinton lead, which turned out to be false.

So, have we reached a crossroads in the extent to which survey based research methods are able to predict accurately the mood of the public and provide a clear picture of the likely outcome of key political events?

 Methodological issues were identified in the inquiry conducted by the British Polling Council /MRS inquiry into the polls after the 2015 election, which I've also covered extensively in IJMR Editorials. Despite statistical considerations such as the margin of error that mean that any polling estimate is subject to a degree of uncertainty, polls are still expected to predict the right winner even when the outcome appears to be close. At the same time, policy makers look to polls and surveys to provide an indication of attitudes towards current or future public policy.

Can public opinion only be measured by the traditional pre-coded questions, as used in surveys measuring voting intentions, or should other questions be devised to probe opinions?  Can techniques like crowd sourcing, qualitative research and sentiment analysis, which can be less easy for journalists to report, help to throw light on how opinion is developing? Crowd sourcing methods have, for example,  been tried by some researchers, leading to mixed results (see Martin Boon's double award winning paper, 'Predicting elections; a 'wisdom of crowds' approach', IJMR Vol. 54 Issue 4).

Some researchers have already turned to social media to observe opinion and identify the issues of importance to the public, tracking how attitudes change over the campaign period, and identifying key sources of influence. Social media – qualitative data on a quantitative scale - can track the volatility that appears to be becoming the norm in political campaigns today, sifting through vast quantities of data to provide a constantly changing picture of public opinion.

Do research techniques that involve the synthesis of different research methodologies, and/or the development of complex models, integrating big data and market research as, for example, developed in India in their last national election, give us more reliable guidance about public opinion and the likely result of an election than the traditional voting intention poll? How can we successfully fuse data from such very different sources to produce a more holistic, and accurate picture of opinion? 

Finally, are we researchers also failing to recognise our responsibility to society by continuing to use methods that deliver inaccurate predictions, thereby creating a potentially serious misleading form of social feedback?

In summary, how can we measure and report on public opinion in a way that is not misleading for policy makers and informs public opinion?

We therefore feel that the time is right for a special issue on public opinion research methods. Other than a series of papers on measuring public opinion around the time of the 2010 UK general election, the last time we did this was back in early 2004. We hope that we can attract papers that between them provide a truly international perspective on the theme, including experiences from developing economies.

We are also very pleased to announce that Professor John Curtice (University of Strathclyde) and Nick Moon (GfK) will be joint guest editors.

This special issue, 'Challenges in accurately measuring public opinion', is provisionally scheduled for January 2018, and therefore we have set the closing date for submissions as July 31st, 2017. So, if you feel that you have something to say in this field, please visit the IJMR website for how to submit a formal paper, a Forum article or a Viewpoint on this special issue theme: https://www.mrs.org.uk/ijmr.