...both before and after the Blackberries were added as part of the large grant. I'm not totally comfortable talking about all of it, but I will mention a couple things.
As far as the size of the grant - imagine paying a cell phone bill for 175 teens for four years and ask yourself where a big chunk of that went.
The participants themselves were made aware of the level of scrutiny their interactions received every year. In addition to gaining parent consent Every Single Year at a yearly data collection visit, we got student assent and made them aware that their stuff was being logged. Because they had already been involved for many years, they were used to the idea of being observed and were comfortable with it.
As far as privacy is concerned, access to raw data is tightly controlled. Before just about anyone got a look at the data, all the identifiers that the investigators could think of (including the identifiers for the teen participants themselves, who are guaranteed a certain amount of anonymity) were stripped from the logs. Although, yes, the communications of other teens who had neither given assent or consent were captured, any time it left the secure archive it was made anonymous. It might be creepy if we knew who they were, but anyone involved in working with the text didn't really have exposure to the teens themselves.
As far as the images are concerned, we didn't archive any multimedia, just the text, at least when I was involved. For precisely the legal reasons everyone has brought up.