HughPickens.com writes: Claire Nee writes in the NYT that for psychologists, it's best to observe actual behavior, in real time, and afterward interview research participants. Yet for obvious ethical and safety reasons, it's almost never possible to observe a crime as it happens. Now psychologists have devised a simulated environment that can be navigated using a mouse or a game controller. and had willing, experienced ex-burglars to commit a mock burglary in it. Ex-burglars approached the task in a dramatically different way from a comparison group of postgraduate students, of a similar age as our experienced ex-burglars. Burglars entered and exited the house at the rear, while students, unaware of the cover that the side and rear of the house afforded, entered at the exposed front. Burglars spent significantly more time in areas of the house with high-value items and navigated it much more systematically than the students did. They also showed greater discernment, by stealing fewer but more valuable items. Most important, all participants burgled the real and the simulated houses almost identically (PDF). We concluded that using simulations can be a robust way to study crime, and in studying it this way, we will not be limited to just burglary. "A better understanding of criminal behavior will help us reduce opportunities for crime in our neighborhoods," concludes Nee. "By knowing what the burglar is looking for — what signals wealth, occupancy, ease of access and security in properties — we can make adjustments in awareness and protection."