I was with the kids at a large and obviously well-funded science museum today (I am not talking of the excellent Mayborn Museum in Waco), a place of much noise, flashing lights, and so on, all designed, presumably, to make science fun to kids. It was great fun for the kids. But it didn't make science fun for the kids. The kids had fun while doing things that might have had value for a scientific education had the excitement of noise, flashing lights and too many other kids not been distracting them. Or, as one might somewhat unfairly put it, they had fun doing things that would have had a value for a scientific education had the fun not distracted them from it.
There are two little points here, one for science education and one for both science education and for philosophy. First, to do something flashy—pressing a button and getting a result—is only science if it is done in the appropriate context of theory and/or careful observation. To learn to enjoy doing something flashy like that is not at all to learn to enjoy science. There is no correlation between doing flashy things and doing science. Some scientific experiments are flashy, involving booms and sparks, and some involve noting how varying x correlates with a weak but statistically significant variation starting with the third decimal place of y in a sufficiently large population.
Second, even if what the kids did were science, to have fun while doing science is not the same as to have fun doing science. One can have fun while doing science by listening to enjoyable music on one's iPod while doing an experiment, but that is not having fun doing science. For the doing of science to be fun to one, the fun must come from the doing of science, and must do so, as Aristotle would say, non-accidentally. It must be the fun proper to science, and must come from science in the right way. The fun of getting flashy effects is not one of the main forms of fun proper to science. The fun proper to science is the fun of observing patterns emerging from careful observations, of making and/or testing bold hypotheses, of seeing the mystery of the ordinary, and generally the excitement of the intellectual life.
Now, granted, one can often make a child (or even an adult) come to experience the fun of an activity A by having her engage in A while enjoying B. Thus, I guess, to enjoy certain kinds of music, one needs to hear a lot of it, and hear it in a positive mood. So one might feed someone chocolates while they listen to that music, both to make her stick around and to create a pleasant association. So making kids have fun while they're doing science, even if the fun they have is not proper to the science, is in principle a viable strategy. But that requires that the kids be actually doing science while having fun, and that one be aware of the distinction between having fun doing science and having fun while doing science, so that one does not congratulate oneself on "having made science fun" when one has merely had the kids have fun while having science.
Part of the point here parallels Peter Geach's point about the good. One cannot simply define "good" and "baseball player", and then conjoin the definitions to get the definition of "good basketball player" (if one could, the inference from "x is a good basketball player" and "x is a baseball player" to "x is a good baseball player" would be valid). The same is true of "fun" and "science". An activity isn't an instance of "fun science" just because it is fun and science.