Is distraction good or distraction bad? Here are two studies that might lead you to opposite conclusions.
The first study (below) shows that attending to instructions about a motor sequence task (explicit instruction) resulted in a much poorerperformance for test subjects, than tasks performed without the lessons (implicit).
So being distracted and not listening to a teacher in this case, might be better than if you paid attention to what an instructor was having you learn. Maybe this is why some folks who excel with hands on learning (like engineers, lego builders, etc.) may ignore instructions, and just prefer to go to work.
In another study paradigm (Poldrack), though (classifying shapes with mildly challenging auditory distracting task), multi-tasking and distraction were seen as culprits: distraction by a second task impaired subjects' ability to extrapolate from their learning - so the net conclusion were that conceptual learning is suboptimal in a multi-tasking environment.
If one really thinks about these different studies, though, one can see that distraction itself is not a simple matter. The Poldrack paper was not just distracting, it was somewhat challenging (count the number of high tones). Also the nature of the distraction (sound, counting, visual sorting) used competing modalities.
What does this mean? Distractions are not alike. And tasks aren't either. If your distracting task is somewhat challenging (e.g. trying to listen to your spouse discuss the merits of various cell phone plans while you're driving for the first time in Boston), and perhaps the modes are different (watching, reacting, listening), learning and extrapolation are particularly likely to suffer.
On the other hand, motor sequence tasks that may become automatic, may be particularly well-suited toward implicit instruction or non-competing distractions.
Because today students are multi-tasking more than ever, a nuanced view of multi-tasking is important. Many kids (and grownups) use multi-tasking (music most often) to be in a pleasant state of restful attention, and most will tell you that they prefer familiar and favorite music when brain work is required. Setting up a computer or stereo system listening to music might be fine, but trying to do a new physics problem set while chatting about the latest classroom gossip is not.
Multi-tasking affects the brain's learning systems
Modulation of Competing Memory Systems by Distraction - Abstract only
Eide Neurolearning Blog: Our One-Track Minds: We're Not Really Multi-Tasking...
Eide Neurolearning Blog: The Multitasking or Music-Tasking Generation?
Explicit vs. Implicit Motor Sequence Learning