Election polling plays an important role in democracies by providing, to the best extent possible, empirical insights into public opinion at a time when political leaders and the public are paying heightened attention to how communities should be governed.
At MQO Research, we consider it a responsibility to collect and disseminate this information rigorously and transparently. By doing so, we contribute to an informed public discourse about governance and policy-making, especially in smaller jurisdictions, where such data is often scarce. Along with our industry colleagues, we strive to round out an informed public conversation about the leaders and issues that matter to the people they affect.
Why Conduct Election Polling?
At MQO, we often conduct our own independent, self-funded election polling (and we explicitly indicate if the research was commissioned by someone else), which costs us time and money. So, why do it?
Our reason for conducting election polling is to offer important information to the public during a time when attention to governance is at its highest. Our audience are not just voters but also residents who, while they may not vote, have a stake in how they are governed. High-quality polls provide better-than-anecdotal evidence about the popularity of political parties, leaders, and policies, and they delve into the issues affecting people’s daily lives, gauge reactions to government actions, and measure satisfaction levels with public initiatives.
Polling data is also important information for politicians, decision-makers, and community leaders. It offers a window into the minds of their constituents, helping them understand public sentiment on policies, governance, and pressing issues. We believe that this feedback loop is essential for responsive and effective governance, and we’re proud to play a small role in supporting it.
Risks and Responsibilities
Conducting election polling isn’t without its risks. There’s always the chance of being wrong, which can stem from the inherent challenges of accurately capturing public opinion coupled with the tendency for minds to change. Polling is hard; it’s not a simple matter for a variety of reasons. People are messy, and getting people to be forthcoming and accurate about the way they think is tough enough. Moreover, methodological errors, sampling mistakes, and interpretation errors can all affect the accuracy of polling information.
We also face the risk of touching a nerve—reporting facts about impactful issues can lead to public backlash. However, our commitment to facts and rigorous methodology helps us navigate these challenges and be confident in the work we do. We have to be thick-skinned about this, which is a lot easier to do when we know that we have done everything we can to be as accurate and thoughtful as possible in data collection and analysis, and in the way we talk about the results of our work. The crux of our craft is to provide accurate representations of public opinion, even if the findings may be uncomfortable for some.
Adapting to Changing Times and Combating Threats to Data Quality
As technology and communication methods evolve, so must polling techniques. The extent to which people answer their telephones changes over time, including both landlines and cell phones. The proportion of people who have landlines is different than it used to be. However, more people have cell phones, which can be a great way to reach people because we have good information about where those people really are and they don’t need to be sitting at home for us to reach them.
We blend landline and cell phone methodologies to cover the population in the most representative way. We also incorporate online surveys, sometimes blending them with telephone methods, depending on the nature of the research. We acknowledge that people who take surveys, especially those who do so often, can be a little different than people who don’t. We make methodological choices with eyes wide open, understanding that each method has its own set of challenges and benefits, and we design our research and produce our analyses with these considerations in mind.
Data quality is a significant consideration, especially with the (necessary) rise of online survey methodologies. We face challenges like people attempting to game the system for compensation or bots attempting to fill out surveys. This is an ongoing “arms race,” and it’s one of the main reasons that conducting quality research is challenging: We have to be ever more clever and technologically savvy in identifying, blocking, and removing invalid responses while not inadvertently excluding valid respondents. To combat this, we continually evolve our data validation techniques and blend our methods.
Being Careful, Having Standards
Rigour, innovation, and paying careful attention are the attributes that separate good market research from excellent market research. It involves the design of the research upfront and being extremely astute when it comes to validating and protecting the validity of the data. There is always a degree of subjectivity involved in every step—from research design to data collection, data validation and cleaning, analysis, and finally interpretation and reporting back—and this is where the value of expertise shines brightest. As a professional market research firm tested over nearly four decades, we know we’re in a position to collect and distribute this information in a rigorous and high-quality way.
MQO Research is an accredited agency member of the Canadian Research Insights Council (CRIC). This accreditation obliges us to adhere to stringent industry standards that we wholeheartedly support, including those concerning transparency in public opinion and election polling. We report sample sizes and field dates, disclose if research was commissioned by a third party, and always show the extent of data weighting used. We appreciate these transparency measures as they help the public and the media to make informed choices about the nature of the data presented to them.
Data Weighting: Art, Science, and Necessary Evil
Data weighting is a legitimate practice used to adjust survey results so they better reflect the actual population demographics. This involves knowing the population proportions on meaningful variables like age, gender, region, income, educational attainment, and others, depending on the nature of the sample and the research. If, for example, we have half as many respondents from a group as we’re supposed to have, we might weight their responses more heavily to compensate. However, doubling the weight of a group’s responses would be fairly extreme, and excessive weighting can introduce inaccuracies—you can run the risk of putting more credence than you should on a small number of responses in a given group, which can throw things out of whack. Therefore, at MQO, we place significant emphasis on sampling techniques to closely mirror population proportions, minimizing the need for extensive weighting adjustments. We try to be as accurate as possible by limiting the extent of data weighting by taking the time and care to collect responses from a sample that reflects our population to the best of our ability.
Industry Collaboration and Continuous Improvement
We acknowledge that no research is perfect. We’re always hedging a little bit on our results; we always allow a margin of error (or equivalent) around our results, and we know that it is important to expect the unexpected. That’s why we value the participation of other reputable and accredited firms in public polling. We consider them to be industry colleagues, and together we round out a profile of public information, contributing to a more comprehensive understanding of public opinion. Each time we engage in public opinion research, we strive to be adding a little bit more to the conversation—our own independent, best-as-we-can contribution to better public understanding of the society that we live in.
High-quality election polling is both challenging and essential. It requires a delicate balance of methodological rigour, adaptability, and ethical responsibility. At MQO Research, we’re committed to overcoming these challenges to provide reliable information that enriches public discourse and supports a better-informed society.