The Angus Reid Forum conducts its polling entirely through online research— surveys or studies conducted on the Internet via a personal computer, tablet or smartphone. We have made the decision to conduct our research solely through online means because traditional telephone based approaches to collecting survey data have become increasingly problematic in the past decade. Unprecedented refusal rates, technology barriers and declining coverage have all made telephone interviewing a challenge in much of the developed world. This type of research is quickly becoming little more than “phone spam”. Pew Research noted response rates had dropped to six per cent in 2018 – something that has driven that organization to also shift to online. Alternative approaches are needed by organizations and institutions that require an accurate reading of the public mood, online research fills this gap.
The appeal of online methods is that they provide a far richer environment for measuring public opinion. The online environment presents respondents with highly visual, interactive, and engaging surveys, ensuring that people provide thoughtful and true-to-life responses to questions tackling complex subject matter. Pictures, audio and video clips become part of the experience for respondents, who are not asked to remember something they might have seen or heard, but presented with the opportunity to see and hear it.
Finally, online panels allow researchers to delve into areas of respondent perception and experience that are simply impossible to penetrate via the traditional telephone survey. This includes detailed surveys of health conditions, personal experience with abuse, and detailed inventories of household possessions.
Unlike most telephone surveys where the relationship between study participants and researchers is fleeting, anonymous and largely devoid of commitment; most successful online research requires transparency, physical or symbolic incentives and trust.
What’s an online panel?
Sending online surveys to a random sample of the population is not possible in an era of anti-spam regulations. Unlike traditional telephone based sampling practices, where random calling is considered a methodology of preference, effective online research requires a significant investment in establishing respondent participation and engagement. Potential respondents must be invited into the survey process using a wide variety of methods and channels. Since the investment associated with this invitation process can be significant, it is increasingly attractive methodologically to invite potential survey participants to a situation where they agree to respond to more than one survey. Those who agree to consider taking multiple surveys are referred to as “panelists” and the group which they belong to, in the eyes of a research practitioner, are referred to as “community panels”. These communities can consist of groups as small as several hundred members to vast communities in the tens of thousands.
The methodological advantage in the online world is that panelists typically share background demographic information which forms part of their “profile”. This profile is an essential tool for sampling and analysis purposes, meaning that respondents will not be asked to continually answer questions about gender, age and where they live, etc.
The Angus Reid Forum recruits via a widespread invitation approach and a double opt-in screening procedure. Willing respondents are recruited through targeted banner ad placements in an extensive array of websites and through partnerships with non-governmental and charitable organizations. This approach ensures an appropriate demographic balance that captures the diversity across all sub-segments of the population.
These community panels are maintained through advanced sampling techniques and frequent verifications of personal identity, contact information, and demographic characteristics. Relying on a combination of sampling regions based upon configurations of electoral districts and past voting trends, the Angus Reid Forum panels reflect the general population by continually verifying and recruiting so that the socio-demographic characteristics of each sampling region match actual sub-populations according to both the census and electoral data.