Sep 16, 2011 15:28 ET
MRIA Responds to “Evaluating the Polls: An Open Letter to Ontario’s Journalists”
TORONTO, ONTARIO -- (Marketwire - Sept. 16, 2011) - MRIA has been asked to comment on “Evaluating the Polls: An Open Letter to Ontario’s Journalists,” released by Ipsos Reid on September 14, and is pleased to do so.
MRIA agrees that all polls are not created equal. One key factor for differentiating between reputable polls and others, which may not have been conducted in accordance with statistical science principles and industry best practices, is whether the firm belongs to a self-regulatory association. In Canada, MRIA members are required to adhere to the association’s Code of Conduct and Good Practice, which includes specific provisions related to the conduct and reporting of polls. Further information on what constitutes a reputable poll, how results should be reported, and a directory of the association’s members can be found at www.mria-arim.ca.
Resources for Discerning Journalists
To assist journalists in reporting on polls appropriately, MRIA collaborated with The Canadian Press in the summer of 2010 in modernizing the “Reporting on Polls” section of The Canadian Press Stylebook. A copy of the revised section is appended and can be found here. At the time, MRIA also published “Assessing the Quality of Online Polls: Nine Questions for Journalists to Ask”, which is appended and can be found here.
MRIA acknowledges that IVR (interactive voice response or touch-tone telephone) polling is still in its infancy in Canada. MRIA supports innovation in the industry, however, and notes that members who use this technique have engaged in testing to ensure IVR’s accuracy before they began using it in publicly released polls.
Also, in a published paper titled “An Evaluation of the Methodology of the 2008 Pre-Election Primary Polls,” prepared by the American Association for Public Opinion Research’s Ad Hoc Committee on the 2008 Presidential Primary Polling, the authors (a panel of survey research experts) indicated that “all of the final pre-primary polls were conducted by telephone, using either CATI or IVR systems. We found no evidence that one approach consistently out-performed the other – that is, the polls using CATI or IVR were about equally accurate.”
In 2009, online research supplanted telephone surveying as the dominant research methodology in Canada, and its share of the total market has continued to grow since then.
Online research polls using convenience samples have, for a number of years already, produced results that have proven to be reliable predictors of outcomes, including election results.
Online survey participants are self-selected, meaning that a margin of sampling error cannot be calculated or quoted for online panel research studies. Using margin of sampling error to describe the accuracy of online polling is misleading, and prohibited under MRIA’s Code of Conduct.
Weighting, to adjust for potential biases in survey data and produce reliable results, is routinely done with a variety of methodologies, including telephone and online surveys. Weighting adjusts poll data in an attempt to ensure that the sample more accurately reflects the characteristics of the population from which it was drawn and to which an inference will be made. Census data is often used to bring the sample in line with the characteristics of the population. Weighting does not involve any changes to the actual answers to survey questions.
MRIA supports innovation in how poll questions are asked, provided that the results are carefully reported in light of the question design. There is more than one right way to conduct research and apply analysis, and MRIA members are not afraid to take experimental approaches to help their clients and the public reach a deeper understanding of the issues.
In a spirit of innovation, advancement, openness and transparency, MRIA encourages industry debate about the use of new and emerging survey methodologies.
MRIA’s Research Agency members all face the same challenges, and all use statistically sound, albeit sometimes different, methodologies to listen, observe, analyse, interpret and deliver accurate, actionable insights to clients. Through innovation, professional development, and certification, and the rigorous execution of research studies by thousands of dedicated employees across the country, our industry continues to be a world leader in the field.
A directly relevant MRIA statement on the accuracy of election polling, published in February 2011, is appended and can be found here.
Surveys and Polls Serve Democracy
Polls strengthen Canada’s democracy by giving voice and influence to Canadians on products, services and issues that directly affect their lives.
As the single authoritative voice of the Canadian marketing research industry, MRIA and its members stand behind this message, 100% of the time, 20 times out of 20.
You Speak. We Listen. Things Improve.
ce communiqué de presse est disponible en français
For further information, contact
Brendan Wycks, MRIA Executive Director,
at 1-888-602-6742 or (905) 602-6854, ext. 8724 or email@example.com
Anne Marie Gabriel, Managing Editor of Vue magazine
at (905) 602-6854 ext. 8723 or firstname.lastname@example.org