You Are Interested in Studying the Effects That Gambling Has

You are interested in studying the effects that gambling has on temporal discounting (i.e., devaluing rewards a person might get in the future relative to smaller rewards "right now"). Accordingly, you decide that there is no better way to research this relationship than with a field experiment in Las Vegas. However, you are still unsure how to operationalize your variables, and how to achieve some sort of random assignment to conditions. Specifically, how could you control as much of the experimental context as possible, given that Las Vegas is such a crowded and lively place? What could you use to your advantage? Is there a way you could assess temporal discounting using unobtrusive measures - if so, how; if not, why not? How could you demonstrate that the act of gambling caused a change in people's temporal discounting "ratio" (i.e., their weighting of future rewards relative to immediate rewards changes after gambling)? Should you just use laboratory-validated gambling simulations and temporal discounting measures in your field study - why or why not?