We've been fooling around with a fresh approach for Information Technology, Facilities, and Faculty Support units in universities to gather feedback about facilities and technologies .  The problem: an institution, or system of institution, administers a set of learning spaces, each with a somewhat different array of features for supporting learning (e.g., projection, document cameras, movable seating, smart boards, wireless, clickers, ...).  The unit, or units, supporting these rooms need to know a) how valuable each of these features are so that they can budget for next year. They also want to know how to alter their plans for training and support. 

If there are more than a dozen learning spaces, interviews aren't practical. So we suggest surveying the faculty and students who use those learning spaces. Traditional surveys won't work if different rooms have different features. You can't ask everyone about "clickers" if 80% of the respondents don't even know what a "clicker" is.  So we suggest using a matrix survey. With a matrix survey, different groups of respondents (e.g., respondents using different classrooms) can be asked different sets of questions.  Only faculty teaching in room equipped for wireless will be asked if they are using wireless, for example.  And, if faculty say they are using clickers in a particular class, then students in that class will be asked about the value of clickers.

If you compare these two instructor feedback forms, both created with the same matrix survey but tailored for different classrooms (actually for all the instructors teaching in each of those rooms) you'll get a sense of what we can now do with a matrix survey:

Because it's a matrix survey, all the instructor responses will stream to the same database (even if the opinion sampling goes on over several terms), so that an institution could, for example, see if their smart boards were being used more productively as the years go by.

Here's a pair of response forms, both generated from a different matrix survey, this one aimed at students:

Each of these matrix surveys, obviously, is a brief proof of concept, ginned up in an hour or two each. I experimented with several different design options here, and I'm just getting started.