 No pressure. Wow, that, here we go. All right, thanks, Brian. My name is Lisa Cuevas Shaw. I'm the Chief Operating Officer and Managing Director for COS, and let me also add my thanks to everyone for being here. So we have evidence to suggest that there is substantial room to improve the credibility and trustworthiness of research. We also understand that we have a dysfunctional reward system that carries incentives for individual success and arguably for institutional success that are focused on getting research published, not getting it right. The good news, as Brian walked through, is that we have viable solutions to correct the reward system, which some, and many of you here, are implementing, shaping, co-constructing, constructing, and developing. But the key question, as Brian teed up, was what's a viable strategy for changing research culture and practice at scale under these conditions? So at a high level analysis, we know the following. As Brian noted, these current solutions are embedded in our highly complex, largely decentralized system. We've got multiple actors working at different levels, global, regional, national, institutional, individual, all with different needs, motivations, interdependencies, levels of risk tolerance, different levels of access, inequities exist, different community norms. Despite that heterogeneity, there are two key assets across this complex system that we have. So we know there's evidence of shared values around the intent and the principles underpinning science and also around open science. As well, there have been a lot of science, particularly recently, the legitimation of these values through policy change and various similar signals, which is wonderful. Another key enabler, and it seems simple, but it can't go without saying, technological innovation. It is easier to practically solution workflow changes at scale than it ever has been before. And that's not to suggest that it's not complex or that it doesn't take a sizable and sustained investment in infrastructure, but it's definitely possible to realize efficiencies through advanced and networked technologies. Therefore, we have an opportunity to promote self-organization among communities within the system if we pursue what Mazarin Banaji referred to as behavioral realism. This argues that interventions or efforts to change the culture have to be guided by knowledge and current evidence of how people actually think and behave, not how we want them to think and behave or how we think they think and behave. In the early 60s, Everett Rogers introduced the diffusion of innovations theory to provide a process framework for understanding how an innovation and idea technology spreads through a population over time. Based on the theory, there are five key factors that influence that rate of adoption. So you have something called relative advantage. Is the innovation better than the status quo? Will researchers, research stakeholders perceive it as better? If not, the innovation is not going to spread that quickly. Compatibility. How does the innovation fit with research stakeholders' past experiences and present needs? If it doesn't fit both of those well, it will not spread well. Complexity, how difficult is the innovation to understand and apply? The more difficult, the slower the process. Trialability. Can researchers and stakeholders try out the innovation first? Or must you commit to it, must we commit to it, whole cloth? If the latter, people will be far more cautious about adopting it. And then lastly, observability. How visible are the results of using it? If researchers or we adopt it, can the difference be discerned by others? Again, if not, the innovation will spread more slowly. So from the point of view of the researcher, let alone the research stakeholders, contributors, all of us, there's quite a bit to consider with these factors when thinking about adopting open science practices. We know and we empathize with the range of barriers and challenges that some of these factors present. But this is where it's useful to consider another aspect of the diffusion of innovation theory, which identifies five categories of adopters. You have innovators, early adopters, the early majority, the late majority, and the laggards. These categories are based on the time it takes us to adopt a new innovation and the degree of risk we're willing to take to adopt it. For the purpose of this change effort, we change the language, we modify it just a little bit to make it fit for describing research culture change. So innovation refers to open scholarship practice or structural change that enables open scholarship. Early adoption is what we described earlier. Mainstream means that most in a particular scholarly community have adopted the practice or practices. And then establishing that standard means that culture change is pervasive, and it has brought most of the laggards along. So that all sounds well and good. And Roger's theory and other compatible innovation diffusion theories for that matter have demonstrated this relative adoption curve time and again. We've all seen it in technology, mobile phones, internet, in public health campaigns, in the business world with new products and services. But as Brian noted, we actually need an implementation strategy here. And starting and scaling the adoption of open behaviors, those innovations by researchers requires that systems-based approach. And it needs to be participatory and inclusive as well. But no strategy for change can expect to activate everyone all at one time. So in applying diffusion of innovations, we aim to catalyze and empower innovators and early adopters as the beachhead for change in a scholarly community. And we aim to do that by providing open infrastructure that makes it possible to do the behaviors, conducting user-centered product and process development to make it easy to do the behaviors, supporting grassroots organizations to activate early adopters and make their behavior visible to shift community norms towards the behaviors, offering and co-constructing solutions with all of you with stakeholders, journals, funders, publishers, institutions to be able to nudge those incentives to make it desirable to do the behaviors, providing and promoting a policy framework for stakeholders to make the behaviors required. And that effective policy implementation needs that infrastructure to make it possible. It also needs that community buy-in to make sure that the behaviors are deemed good practice, not just a burden. So this image would suggest that this is a nice, linear, sequential process. But these five levels are highly interdependent. They're in development simultaneously and interactively. That's what makes all the work fun. And each is necessary, none sufficient on their own, to sustain the change effort. So we don't want to give the impression that the relationship between the diffusion of innovation theory and the theory of change is a neat one-for-one relationship between the rates of adoption and the levels of intervention. But generally speaking, there is some connection between making it actually possible and even better easy to catalyze innovators. And generally speaking, there's a relationship between making it required and the practice becoming a standard one. But overall, this process is very dynamic, highly interdependent. So to illustrate this in action, we can take one example of a specific innovation or open behavior we want to scale and map it to this implementation strategy. So Brian had introduced the concept of lifecycle open science, and we already talked about publication being the currency of advancement or reward in our system. So let's take a look briefly at how the community started and is scaling change of our publication norms through the lens of preprints. Preprints are versions of scholarly manuscripts made publicly available before any formal editorial and peer review process. Ultimately, they aim to make research more quickly available to enable discovery and reuse. So this chart here shows the growth of preprints over time across some broad research categories. You have the natural sciences, physical sciences, life sciences, social and behavioral sciences, and the arts and humanities. Preprints are effectively a standard in physics and closely allied areas, and they've been popular in economics for a long time as well. Essentially, this chart shows growth in preprint output overall, and also that there are some disciplinary silos. Preprints have been popular in some domains for a long time, but didn't necessarily rapidly spread to other disciplines without effort. However, with technology advancements, active community engagement, and COVID-19 as a catalyzing external event, we can see that they have dramatically accelerated in the life sciences. And you could argue, well, a lot of that would be COVID-19. But again, arguably, if the community was not ready, if there was not infrastructure in place, I don't know that we would have seen this bump. There wouldn't have been that readiness. So why, how has this been happening? We'll walk through some, not all, of the milestones just to illustrate the theory of change in action. So in the Make It Possible intervention, the creation of BioArchive, I think it was in 2013, as a recognizable place for life sciences to do preprinting was a major milestone. There were other existing preprint services, but they were generalists, or they didn't feel like a particular home for life scientists, so they were disciplinary specific and only accessible to a narrow range of life science. In the Make It Easy intervention, you have publishers such as PLOS integrating submission of preprint with submission of article back in, this was 2018. And that made it easier from a workflow perspective and also validated, at least to some degree, the posting of preprints. And at this point, there are many other publishers who have done that, that has become a little bit more of a norm with publishers and the research process. Not shown here, but just to say, improvements to BioArchive to make it easier and easier to post, choose licenses, address whatever concerns people have about preprints while they share or consume them is making it easier. We're not there yet in terms of removing all of the complexities and the concerns, but indeed there's progress that's been made. For Make It Normative, a wonderful action and milestone to point to is the formation of ASAP Bio in 2017 and all of its community building efforts involving key champions, high credibility exemplars, early career researchers, you can see their fellowship program there. There's been a lot of public information campaign work. On the right hand side, you can see the amplification and network effect of signaling and legitimizing preprints via social media, largely in Twitter. This has enabled norm setting as well. In the Make It Rewarding intervention, NIH changing its biosketch policy to be inclusive of preprints, thereby legitimizing and actually rewarding this work product has been a great milestone. Journals have been evolving policies to at least be tolerant of preprints, some more encouraging than others. And funders, as you can see here, Cancer Research UK, but there are many others, all actively promoting preprints. Finally, in the Make It Required, you have to show a bolder publisher move, you have E-Life only reviewing papers that have been preprinted. You also have different funders such as CZI, Welcome Trust requiring preprints for some of their funding. Now, that all makes it sound as though growth in preprints in the life sciences has been easy over these past five to 10 years, but this has actually been a much longer effort and there is certainly work to be done. So this is not to say this is all nice and neat and tidy. Naomi Penfold and Jessica Polka addressed some of the existing technical and social challenges of preprint adoption in the life sciences and you can see them here. Many are not unique to the life sciences, but their work did call attention to our need to be sensitive to and address various disciplinary boundary conditions as well. The issues that they call attention to are highly compatible with the five factors affecting innovation adoption that we pointed to earlier. So there's certainly a lot of work to be done here, many questions to continue to address around preprints, but this gives you a sense of how diffusion of innovations and this theory of change anchored in behavioral realism can be applied. And more importantly, how can we use that framing to continue and improve on efforts or observe what's working and why or why not? So we began with the key question of what's a viable strategy for changing research culture and practice at scale when we've got this very entrenched dysfunctional reward system. The combination of diffusion of innovations and theory of change provides a framework for an evidenced and actionable implementation strategy with very specific levers that aim to address the realities that individual researchers face in this complex social and political and economic system. So thank you.