I think one of the reasons this is confusing to many, is that beginner circuits rarely have more than one voltage. Every project I assembled as a beginner had but one input voltage. This may not be true of everyone, but I think it's common enough that the idea of having another voltage, and it being relative to something, makes it elusive.
As I think about this, there are a couple of ideas that I personally would use to explain relative voltages.
The process of learning how to use a voltmeter lends to this discussion. Learning about electronics, I would measure the voltage across an LED or resistor and sometimes got the leads backwards. The voltmeter would show a "negative" voltage. Reversing the leads "fixed" it. The value was unchanged. It became quickly apparent that it was a matter of perspective. "Am I looking at this with respect to ground, or respect to the positive voltage terminal?" It's the same voltage, with different points of reference.
Expanding on that, I think it simplest to show a schematic with two batteries in series, and explain that connecting a meter in the various combinations will produce two voltage values, sometimes positive, sometimes negative. One could then add to that schematic with a couple of resistors or lamps to show that it's possible to provide one voltage or the other to those components.
Another concept that helped me understand voltage potential and how it can be relative, is the age-old question about "Why don't birds get electrocuted on power distribution lines?" The answer is that they are at the same voltage potential as the wire they are touching. As far as the bird is concerned, everything is fine. Relative to ground the bird is at a lethal potential, but since it's not touching ground, there's no real danger.
It might be useful to construct an example from the bird-on-wire example: Suppose the bird is on a 10KV line, and a nearby wire for some other purpose carries 5KV. With respect to ground, both lines are dangerous. But what if the bird on the 10KV line were to make contact with the 5KV line? It would still be a 5KV shock. Less than what would happen if it were to bridge the 10KV line with ground, but still deadly.
The point of this is to show that the voltage from these lines is relative to some frame of reference. With no other information, someone with a meter might think the 5KV line is "ground" and the other is a 5KV power line. You could, in theory, say that a third wire connected to earth is, in fact a negative 5KV power line. (Misleading, but the point is that calling a particular line "ground" is basically establishing it as a point of reference.)
Finally, an analogy that isn't perfect but might be of use, is to imagine people on balconies of a high-rise building. If someone on the 80th floor were to jump to the 79th floor, they would probably be OK. Their "falling potential" to that floor was only 1; but clearly (with respect to actual ground) their falling potential is a much deadlier 80. What's important is where they are jumping to - that point of reference.
