 So, a standard MRI scan maps structural features of the brain. A functional MRI scan measures blood flow to various areas of the brain, and this is a great help because it allows us to measure levels of activity during decision making. So we know from previous research that managers often make decisions on the basis of their emotional reactions rather than on the economic characteristics of the decision alternatives. And this can be costly for firms. And economic theory would traditionally prescribe performance-based incentive contracts to reduce these sort of costly managerial decisions. But psychology theory suggests that emotion is hardwired, and so it's not clear exactly how performance-based incentive contracts should work in these emotional decision contexts. So we use a combination of FMRI brain imaging technology and traditional behavioral methods to show three things. Just consistent with psychology theory, we find that emotional reactions seem to persist under a performance-based incentive contract. But these performance-based incentives also induce managers to process information more analytically when emotional reactions are potentially costly. And as a result, consistent with economic theory, we find that performance-based incentives reduce although they don't eliminate costly decisions based on emotion. So what are the benefits of FMRI scans for accounting research? So of course you can use multiple methods to address any research question, but we think that FMRI gives us a unique advantage in this context. We initially came to FMRI because we were seeking to test a theory that required measures of processing style. And the behavioral measures based on self-reports that were out there just weren't particularly satisfactory. And so we wondered if we could learn more by directly measuring brain activity. And from a big-picture perspective, FMRI was the tool that allowed us to demonstrate that these two competing perspectives, the psychology perspective that emotion is hardwired and the economic perspective, the performance-based incentives reduce managerial decisions, costly managerial decisions could coexist and we could disentangle those effects. Really the choice to use a new methodology should be driven by the research question. There are significant resource needs for FMRI in terms of researcher time, scanner time, participant time, and funding sources aren't always clear. But if you do decide to work in this area, be prepared to make a large upfront investment. So learning the language of neuroscience isn't easy, learning the biology of the brain, relevant prior research, the limitations of the technology, all of that requires a large upfront investment. And finally, most accounting researchers who work in this field won't be experts in it. And so they'll have to find neuroscientists to work with. And that's not always easy because as in most of the hard sciences, neuroscientists tend to be attached to labs. And they're not always willing or able to work outside of those labs. But it's really essential to find the right person. So my colleague, Annie Farrell, and I wouldn't have been able to undertake this project without the help of our co-author, Josh Go. What do you think, Ibarra, that goes from here? So, you know, FMRI allows us to open up the black box of cognitive processing. And it's not always desirable or necessary in accounting research. But I think it can be helpful in a couple of areas. So there's a recent paper in the Journal of Accounting Organizations in Society in which the authors very astutely, I think, map the evaluative versus reactive functions of the brain to the reporting versus control functions of accounting. And so from that perspective, I think FMRI can help us to understand how both how people use accounting information in a relatively deliberate way, but also how they react to it in a relatively automatic way. You know, some say the process isn't particularly important. That it is the outcomes that matter. And that theoretical models are not designed to describe the underlying process, but rather to predict outcomes. And I have some sympathy with that point of view. But I think ultimately, you know, just understanding the underlying process is going to help us to increase the precision and accuracy of our models.