Aspects of Disclosure


Our 9 pillars of disclosure are by far the most comprehensive in the industry.

1. Accuracy Disclosure


As a first step in transparency, it is important that polling firms disclose their accuracy record. Forum Research has consistently been the most accurate firm in predicting Canadian elections.

Forum Most Accurate Firm in Predicting Provincial Elections

Forum had the most accurate prediction for the 2015 Federal Election and 6 provincial elections

“Our record in predicting the outcome of the 2015 Federal Election and in Canadian Provincial elections is unmatched, with our proprietary Forum Poll IVR polling methodology proving its accuracy again and again. As we continue to refine our techniques, and prove our methodology works, we will continually put our methods to the test, which is why we’ve predicted twice as many elections in the last 8 years as our nearest competitor,” said Forum Research president Dr. Lorne Bozinoff.

Read More: Forum Overall Accuracy Release

Research Firm Most Accurate 2nd Most Accurate 3rd Most Accurate Elections Predicted
Forum Research Inc. 7 4 3 15
Insights West 3 1 - 5
Abacus Data 2 1 - 6
Leger 2 1 1 6
CRA 1 - 2 3
Angus Reid - 1 3 7
Nanos - 1 1 2
Insightrix - 2 - 3
Ipsos Reid - 1 3 7
EKOS - - 3 7
CROP - - 1 2
Praxis - - 1 1
Return on Insight - - 1 1
Oraclepoll - - 1 2
Campaign Research - - - 1
Hill & Knowlton - - - 1
Mainstreet - 2 - 6
Justason - - - 2
ThinkHQ - - - 1

By-elections

By-elections provide the Forum Poll tm with an opportunity to demonstrate just how accurate our Interactive Voice Response (IVR) polls are. Of all the by-elections where other polling firms attempted to predict the vote, Forum was the most accurate in four out of the six. Of the most recent by-elections, the Forum Poll tm accurately predicted the winning candidate 16 out of 19 times, or 84% of the time.

Read more: 2014 By-Election Accuracy Summary

2. Client Disclosure


Forum Research is retained by many leading organizations. Our client list is available here.

3. Data Access Disclosure


The most important element of transparency is making the raw data available to external researchers. Forum has placed its raw polling data in the Data Library of the Department of Political Science at the University of Toronto. All data is accessible to any Ontario university faculty or student member.

You can view the data archive at   The University of Toronto Political Science Data Archive

4. Information Disclosure


We support the disclosure of all relevant details of each individual survey and as such our news releases always include a Methodology section with the following information:

  1. Data archive location
  2. Field dates
  3. Margin of error for the total sample
  4. Margin of error link for all sub samples
  5. Name of sponsor
  6. Sample size
  7. Weights used

5. Methodology Disclosure


The Forum Poll tm is conducted by Forum Research, with all results based on an interactive voice response (IVR) telephone survey. Our surveys are based on randomly selected samples of respondents 18 years of age and older. The accuracy of the results is based on the sample size for each study. Subsample results (such as age, gender, etc.) are less accurate. Margins of error for samples are available here. Where appropriate, the data has been statistically weighted by age, region, gender, and other variables to ensure the sample reflects the actual population, according to the latest Census data. This research is not necessarily predictive of future outcomes, but rather, captures opinions at one point in time. When predicting popular vote, actual election results may differ from Forum’s predictions due to:

  1. Changes in voting intentions after the survey was conducted
  2. Parties and/or candidates success at getting their voters out
  3. Differences due to sampling error.

6. Non Random Online Sample Disclosure


We only conduct polls for which we can calculate a margin of error. In order to calculate a margin of error we require that the sample be randomly generated. We do not consider opt-in online panels to be random samples and therefore do not conduct polls using online panels.

7. Paid (recruited) vs. Voluntary Samples


Forum Research actively seeks to attain unbiased survey results. Part of that pursuit involves strictly using voluntary respondents. All of our respondents are volunteering their information to us. None of our survey respondents are paid for their answers, and thus have no bias or ulterior motives related to financial or material gain.

8. Professional Training Disclosure


Forum Research has a team of experienced professionals who conduct our polls. Our polling team is headed by Dr. Lorne Bozinoff. Dr. Bozinoff has a Ph.D. in Marketing from the University of Toronto backed up by over 30 years in the Research and Customer Experience marketplace. He is a recognized expert in the areas of survey research, questionnaire design and advanced statistical analysis. He has managed survey research projects covering a range of topics from graduate outcomes/satisfaction, employer/business satisfaction, student satisfaction, customer satisfaction and employment outcomes, to public policy and corporate image.

Before joining our firm, Before joining our firm, Dr. Bozinoff was Executive Vice-President of the Gallup Poll. Prior to that, he was Senior Consultant at The Coopers & Lybrand Consulting Group. Previously he was Marketing Research Manager at Bell Canada and before that he was a faculty member of The Pennsylvania State University where he taught courses in survey research and consumer behaviour

Professional Designations

Dr. Bozinoff has the following professional certifications and as such is bound by their respective Codes of Conduct and Ethical Practices:

  1. C.Sci. – Chartered Scientist
  2. C.Stat. – Certified Statistician
  3. CMC – Certified Management Consultant
  4. CMRP – Certified Marketing Research Professional
  5. P.CRM – Customer Relations Management
  6. PRC – Professional Research Certification

Professional Affiliations

  1. American Marketing Association
  2. Canadian Association of Management Consultants
  3. Intelligence Association (Gold SealMember)
  4. Marketing Research
  5. Marketing Research Association
  6. Royal Statistical Society (Fellow)

9. Weighting Disclosure


While we feel that our weighting methods are a highly proprietary element of our research, in the interests of transparency, we have shared our weights with senior editors at both the Ottawa Citizen and Toronto Star.

For more information on survey weights please read this article: A Discussion on Survey Weights