I have children in public schools in grades 3 through 8. This isn't surprising ...
... national tests given in 2009, showed that 40 percent of Minnesota eighth-graders and 43 percent of fourth-graders were either 'proficient' or 'advanced' in science."...
In other words, 60% failed.
I'm not concerned about science teaching in grades 1 to 4. Maybe someday, but not today. We do have a problem with teaching science in Minnesota's public schools between grades 5 and 8.
So what should we do about it?
We need old, cynical, hard bitten teachers nearing retirement to tell us what needs to change. I can only suggest a solution based on what I know about doctors. Since doctors and teachers have a lot in common, this may be relevant.
I was a real family doc once. I switched careers about fifteen years ago, but my wife still sees patients. There are a lot of superb family physicians, particularly in rural America but also in urban settings. On the other hand, radiologists make far more money for less work. The gap has grown over the past decades. Since most radiologists would be lousy family physicians, and most family physicians would die of boredom doing radiology, the impact isn't as large as it might be. Still, it's real. If you want to get more quality family physicians (or pediatricians, internists, etc), the cheapest solution is to pay radiologists less .
I suspect the same is true in teaching. Science teachers and reading teachers draw from somewhat similar candidate pools. The quality of science teaching will depend on the relative prestige and rewards between one domain and the other. If we want better science teaching in America, we need to make teaching science relatively more attractive than the alternatives.
Any comments from those who teach young students?
 You'll know you're paying too little when it becomes hard to read MRIs or get an interventional procedure performed. I suspect that would take a vast reduction in income.