The last presentation at ELA I attended was a quite interesting one. A couple of researchers from the University of Cincinnati presented on a study of school principal's perceptions toward off-campus speech. Their findings will hopefully be out in an article soon, but the general idea was that place of the creation of the speech was a significant predictor of a principal's disciplinary reaction to the speech. In effect, if a student created the speech at home, principals were less likely to regulate whereas if the speech was created at school, the school was significantly more likely to regulate. Watch for an article soon that gives much greater detail on their findings.
But, here is the thing. The "place" of the speech's creation actually matters very little in the legal analysis. The legal analysis is pretty much wholly concerned about the effects of the speech. Tinker, in particular, set up a student speech analysis regime that is wholly concerned about the effects of speech on the schools. This is why we are concerned with whether the speech's outcome will have a material and substantial disruption on the school environment. Now, this comes with some cavets. The Bethel and Hazelwood analysis IS concerned with the creation of the speech (not lewd, vulgar or offensive and not school sponsored, respectively). This can play into the off-campus speech analysis, but not nearly as much as the Tinker issues and I highly doubt that is what is accounting for the researcher's findings.
This "place" issue is important because of the geographical effects of the Internet. When it comes to published speech, the Internet has gotten rid of geography in a lot of ways. Internet speech is neither here nor there, it is both. This blog post, for instance, is in San Antonio's airport right now where I am writing it, but it is just as much in Lexington or Cincinnati or your school or my school or a student's home. It is in all those places as soon as the server request transfers the data through the Internet to a computer (and with RSSing, that happens a lot without someone even intentionally sending a request). Thus, the effects of the speech (what Tinker is concerned with) can also be in all those places. So, using place as a determinate factor really doesn't make sense. The "nexus" test and the "disruption" test (if those are 2 separate things) only cares about the effects - not the creation - and since those are the two most important tests for off-campus speech, the researchers findings really are instructive that principals are not understanding the fundamental analysis at play here and are likely making mistakes in disciplining student speech.
Anyway, it was excellent research and I look forward to reading the article on it.