The IJB profiles our collaborators from time to time. Our inaugural piece features Neil Seeman, whose work was integral to the recently-published Generation Distress investigation. Seeman is the founder, chief executive officer, chief privacy officer and chairman of RIWI Corp. He leads overall company strategy and RIWI’s work in global public health security –– and in global security data solutions to assess, predict, prevent and eradicate violent extremism. Neil invented RIWI’s core intellectual property. He is the author or co-author of hundreds of articles in major media around the world, more than 30 peer-reviewed journal papers and several books and monographs.
- What drives your work?
I am focused on measuring and documenting the truth about what people in all regions of the world think, feel, need and want. This higher-level goal requires an exacting commitment to continually refining our technology platform to gather the authentic opinions, observations and fears of “quiet voices” –– traditionally referred to as “underrepresented groups” in public health. For us, quiet voices include people of all types whose opinions are not captured in traditional mechanisms of measuring “citizen voice,” such as election results, focus groups, social media analysis, official labour and wage statistics, banking or other financial transactions data, online panel surveys, mail-based household surveys, telephone surveys, or face-to-face interviews. Understanding, for example, the statistical relationships between what quiet voices in rural parts of China anonymously report to the RIWI platform and what quiet voices in rural America report to us on the very same issues can yield insights into commonalities of concern that bind people across countries. Only by measuring changing perceptions of truth can we then measure what policies, campaigns or activities work to unite people together –– or wedge them further apart. In public health, this kind of mapping exercise can result in extraordinary insights. What does the same type of person with very similar health characteristics, health system access challenges, stigma challenges and similar beliefs and values in a faraway part of the world share with another? How can we find common solutions to empower both people? So we are in the business of “Big Data” and “Smart Data” for global learning, the next evolution of data analytics.
- How does RIWI’s model of gathering trend intelligence differ from traditional methods?
The RIWI machine, built on the architecture of the internet (the Domain Name System, or “DNS”), learns every day and functions based on people stumbling into a RIWI domain that is no longer, or never was, commercially active. There are hundreds of millions of these domains, they change and rise in number every day, on any device. RIWI controls access to a huge, changing dynamic pool of such ‘clean use’ domains. The domains host surveys that may arrive at you from broken links to hypertext on diverse blogs or digital media you consume. That’s because once a web domain goes commercially vacant all the links associated with that domain potentially fall into the fast-growing RIWI ocean of domains capable of inviting you to be RIWI’ed or subject to a RIWI ad test or survey. You anonymously choose to respond. Unlike traditional online survey approaches, the technology’s algorithms ensure that anyone on the web in an area of interest has an equal chance of being randomly exposed to the questions. Also, unlike government or panel surveys, all data are gathered anonymously, reducing social desirability bias and thus eliminating a major barrier to participation. Furthermore, respondents are not incentivized to participate in any way. We randomly engage a new, random set of unique respondents each day. So far, more than 1.6 billion interviewees from 229 countries and territories have participated in a RIWI ad test or survey.
- “The company has specialized in gathering public perspectives on stigmatized topics like access to mental health services, as well as terrorism, gang activity and the treatment of women, girls and LGBTQ communities around the world. Walk us through some of those findings.
In 2011, we went on record for tracking the abrupt change of sentiment on the streets of Egypt prior to the fall of the Mubarak regime. That took me and the world by shock. In 2017, RIWI predictors (people who answer RIWI questions, this time run in partnership with Viacom) successfully predicted the outcome of the Australian same-sex marriage referendum. For us, a lot of this work for which we are recognized internationally in the area of socially stigmatized topics comes both from commercial demand and from decades of expertise studying the many elements of anonymous interaction in an online setting that are relevant to bias mitigation. Stigma is a cross-cutting lens that applies to all our data collection work. Whether it is in the field of finance, humanitarian aid impact or consumer goods trends, our systems and methods ensure that the “RIWI machine” is bias-aware and built to reach as diverse an audience as possible in order to mitigate the bias that inexorably affects the analysis of all events.
- Most recently, RIWI collected the opinions of more than 6,000 Canadian and American university and college students for the IJB’s inaugural project Generation Distress. Why was this project of interest and what did you discover?
When I learned that the IJB was examining mental health challenges on university and and college campuses in Canada and the U.S. as a form of investigative journalism –– as opposed to a critique of the wide range of varied initiatives across North America trying to address unmet student health needs –– I was immediately drawn in. For this is, at root, a journalistic investigation that requires a depth of curiosity and research that is actionable. Equipped with the data RIWI could provide to the IJB, the ecosystem of journalists and academic experts could thereby help identify a baseline of mental health challenges on campus, track changing prevalence over the course of a year (i.e., before and after COVID-19 had unleashed itself across the globe) and identify proactive suggestions from students for students –– for example, about ways of improving access to mental health services for those affected. The act of designing this exercise as a journalistic investigation, and without assigning blame to any institution –– in the service of measurement and honest reporting –– made the work immediately accessible and amenable to changing the public conversation about mental health on campus from a topic that seemingly only affects people in the shadows, to a topic that is solution-oriented, data-driven and now known to affect so many. Clearly, the data reveal, too, that some are far more at-risk than others, and this investigation helps us figure out who, where, and why –– without casting blame or adding to the cloud of stigma.
In this work with the IJB, we discovered in 2020 that there was a sharp spike in reports of both anxiety and depression at the end of March and into April when most universities and colleges migrated to online learning and the pre-exam period began. The changing prevalence of these conditions, especially depression, fell during the post-exam period. However, disclosures of anxiety and depression rose again steadily from July through October, coincident with schools announcing online learning for September. In August, students reported the highest levels of anxiety and depression over the period of data collection as they faced the fall semester of online learning.
Other university-based researchers are already investigating other aspects of the data in preparation for journal publication. Some of the questions under inquiry include:
- Do attitudes toward college/university mental health services differ between U.S. and Canadian students?
- What are the relationships between mental health service use and campus attitudes, common stressors, and negative outcomes?
- Is the relationship between attitudes and outcomes mediated by common stressors?
- Do attitudes and stressors interact to predict outcomes?
To see the RIWI analysis and interpretation of the mental health work, visit here.
- What was your experience with this model of collaborative research between academics and journalists?
There was, from the outset, a “bias to action.” This is what makes me love working with people trained in public health and in journalism. There is a deserved sense of urgency. There is a healthy desire to ensure the evidence layer for urgent public health crises is non-partisan, robust, open-source and open to examination. On the RIWI side, we found it a pleasure to work with the IJB ecosystem of journalists whom we helped guide, but also whom we learned from, in terms of how best to make sense of the data. Further, the process enabled us to meet with university-based researchers of varying areas of focus who wish to leverage the data to investigate their own particular areas of inquiry. This evolved into a mutual knowledge translation experience which led to continuous improvement in how and why we collect data, and how to make the process more impactful for the end-users of data and for knowledge producers and consumers.
- How does this work connect with the groundbreaking mental health work of your parents, University of Toronto Professors Mary Seeman and Philip Seeman?
This journey reminded me of what my parents –– long-time University of Toronto researchers Mary Seeman, OC, MD, DSc, Professor Emerita of Psychiatry, and Philip Seeman, OC, MD, PhD, DSc, Professor Emeritus of Neuropharmacology –– taught me about the foundation of knowledge. Knowledge-seeking is based on humility and asking questions –– always asking questions –– of diverse people with different views, and listening diligently. It is very hard to listen and ask questions well, and this is a process of continuous learning.