This past summer Nature and Scientific American coordinated to publish special issues devoted to science and education. There's science that informs how teaching should happen as well as stuff on how science itself specifically should be taught, both of which may be of interest to readers here for evaluating and choosing schools.
All the articles from both publications can be found linked from this one page:
Here are some super-brief summaries of what I thought was relevant.
The brief 6-paragraph intro "An Education" explains why they did these special articles and motivates active (vs. traditional passive) learning, in which students are given tough problems and try to figure them out in depth, rather than just being told the answers and then needing to remember them.
"Why we are teaching science wrong and how to make it right" presents evidence that such active learning produces better understanding and memory. Project-based and hands-on-learning type schools from progressive independents to extreme outliers like Brightworks (and even public charter New School SF) would seem to be elementary school analogs to this kind of educating (vs. traditional school structures), but most of the discussion in this article centered around college-level teaching (with unclear validity to generalizing down in ages). It's hard to tell if the increasing standards demanded nationally of public education make it harder to teach this way because it takes more time to prepare for such teaching and adding more stuff to a curriculum that can't be missed leaves less time to not just tell students the answers. It's also worth noting that even at the college level, there is debate about the merits of this approach.
My take-home was to think that this type of teaching is vital as at least a component, but I didn't see enough to convince me that it has to be everything. It made me like in-depth progressive/project-based/hands-on problem solving even better, but even in schools stuck with teaching so much of a core curriculum that much of it has to be taught traditionally, there is probably room for some of this in after-school or supplemental programs, but PTAs might have to fund & choose/fight for it over other things.
"Reading, Writing and high-energy physics" discusses an approach to education (from preschool to university) whose effectiveness "has been verified by hundreds of empirical studies". The approach involves kids coming up with answers to questions (sometimes crazy answers like trees cause the wind by shaking their leaves) and then rather than teachers correcting them or adding new knowledge the teachers help nudge the kids to think of ways to test their ideas, such as asking if anyone has seen wind where there aren't any trees, or performing an actual experiment if possible. The kids then mimic the process of actual science (testing hypotheses). "The point is to spark questions, and a conviction that they can be explored rationally."
The first part of this article discussed early education, so it felt very relevant. I've seen analogs of some of these ideas at various independent schools. The Brightworks/tinkering "never say no" philosophy seemed similar but in the programs of the article it was never tell them they are wrong or give them the answer (science vs. engineering/building focus). Also, I remember reading in SFKFiles comments from prior years defenders of SF Friends School (and from the tour of that school itself this fall) discussions of teaching approaches to problem solving in math rather than just teaching the math itself. But few schools are going to teach like this all the time, and it's very hard to judge just how effective the technique is (let alone for us to judge to what extent different schools do this, or might in the future as evidence for it accumulates). [Locality note: The author of that piece writes for Nature from SF.]
"Body of Knowledge" advocates the outdoors/nature and unstructured play (vs. educational apps & traditional classrooms). Nature is immersive and kids are mini-scientists when free-playing there. Summary of the science: it's clear that time in nature is important to development but the science on the subject is pretty new, so hard to quantify how important or the details of how/why/etc. One study found that outdoor time improved 5-12yo's self-confidence and ability to interact with others & adults. Examples were discussed of how immersive natural environments lead to abstract kinds of knowledge such as math concepts (counting, categorization) not just direct knowledge of plants, etc.
I think many if not most people already agree that outdoor time, time specifically in nature (vs. just outdoors), and unstructured play (which overlaps but isn't the same) are all good. For many urban elementary schools, time in nature is only occasional trips and even outdoor time and unstructured play are pretty minimal, all to a degree that's clearly lamentable. There are some private schools with more of these but it's also clear that one gives up other things for many of those choices. The science doesn't seem at a point yet where it can help us practically make more informed choices other than vaguely factoring in outdoor- & unstructured-ness as one of many criteria in ranking schools. In the meantime, many schools give a lot of lip service to trips, to their small gardens, and to their plans to greenify their asphalt/cement-covered outdoor areas and clearly this is a big enough issue that most are moving in this direction, but some faster than others. The immediate take-home I got here was to plan to get my kids to nature as much as possible outside of school hours once they start K.
"Schools should teach science like sports" makes a great analogy that is better quoted than paraphrased: "Suppose you wanted to teach children to play baseball or softball.
How would you go about doing it? One approach might be to sit them down
and start having them memorize the rules of the game, the dimensions of
the field, the names and statistics of past players, and a host of other
facts. You would stop teaching them periodically to review the material
in preparation for multiple-choice assessment tests. The students who
showed a great aptitude for memorizing large numbers of facts could go
into honors classes where they would memorize even larger numbers of
facts. At the end of the process, without ever leaving the classroom,
how well do you think the children would be able to play baseball or
softball? More important, how many would even want to? Why have we thought that this process would work with teaching science to children?" You can probably guess the gist of the rest. The interesting part is that new K-12 science standards are coming down from a national standards-setting group since 2013 and 12 states + DC have adopted them already. So change seems to be coming. But it won't be at all SF schools at all grade levels tomorrow.
"Researchers find that frequent tests can boost learning" is an antidote to the idea that all testing is bad, an extreme that some progressive, project-based educational ideas seem to be taken to. Rather than throw out all testing, doing it in a way more consistent with research from cognitive science can actually improve retention and deepen understanding. The basic idea is that tests help solidify knowledge and connections to other related facts in the brain and shouldn't be used solely as assessment tools. In addition to a randomized controlled trial of the technique described (with great results), 2 other interesting things were mentioned later in the article: Use of the technique in one course in a certain study resulted in improved grades in other subjects as well. Also, the technique helped reduce achievement gaps between social-economic classes. Lastly, the article lists several ways in which current standardized tests are exactly the wrong kinds of tests to optimize learning based on the known cognitive science.
This sounds interesting, but I'm not sure I've learned much of anything
about testing at different schools to help use this info to better
evaluate schools. The publics for the most part do more of the unhelpful (and maybe harmful for other reasons such as stress, teaching to the tests etc.) testing, but I'm not aware of any schools that specifically use these techniques to optimize learning (but it could be this kind of detail just isn't discussed, especially since testing is such a dirty word to some parents). Interestingly, these techniques are quite a bit different than some of the ideas from the other articles or the standard "project based" or "hands-on" philosophies of many independent or charter schools.
There are other articles in the collection, but I pulled those out as the ones that seemed most relevant. Please share a summary if you read one of the others or your own take-homes from any work in this area, or what you see as relevance to specific schools in SF!