@johnmyleswhite kicked off my morning with a provocative, challenging tweet and long twitter conversation:
One meme I wish would die off: the belief that we can teach high school students statistics without teaching them calculus.— John Myles White (@johnmyleswhite) March 27, 2017
Ironically, one of the most prominent proponents of the StatsFirst idea is the person I learned Calculus from: (Art Benjamin: Teach Stats Before Calculus)
I think the strongest part of his claim, based on comments at other points in the thread, is that there are instances when knowing a little bit more is worse than not knowing it at all. I think he is specifically imagining people making medical and other life&death decisions based on a high school level of statistics knowledge. I can well imagine that it is possible, given such a circumstance, that it might be better for the doctor to know a little bit less, to know that they don't know enough, and therefore to trust the advice given them by experts a little more.
Framed like this, it's an empirical question -- a question of statistics, rather. How many people are faced with decisions like this, where expert level knowledge of statistics would enable them to do all the reasoning themselves, but novice level knowledge of statistics would make them very likely to make critical mistakes of reasoning. What is the cost of this naive statistical reasoning? What would it take to educate everyone who ever makes such decisions able to do the reasoning themselves? What is the marginal benefit/cost of increasing the number of people able to do this work themselves?
I wrote my first real tweetstorm in response, putting physics in the place he puts statistics. Because many of the same things have been said about physics education, and about the move to get more people into the door by teaching Physics First curriculums and by doing algebra-based physics at the high school and college levels for everyone but engineering & physics majors.
- I am not a statistician, or a stats pedagogue -- cannot speak to the specific issues in stats education, but the argument (2/18)
- has been made before, about physics education. And while the issues in physics ed require careful thought, it's clear to me that (3/18)
- the value is entirely on the side of getting more people in the physics door, not fewer: (4/18)
- Physics education cures a number of misconceptions about the world, and creates some others. (5/18)
- Is this an argument for less physics education? Or for better physics education? (6/18)
- Does allowing more people to take any physics class at all (by dropping calculus co-req) empower too much naive physical reasoning? (7/18)
- Do we burden higher education with too many hopeless cases by getting them in the door with calc-free physics... (8/18)
- when we should be weeding them out with as tough an entry as possible? (9/18)
- Are bridges being built with calc-free hs physics knowledge? Do we have a responsibility to ensure that only the fully qualified... (10/18)
- get to the next steps, where they will learn how to build good bridges? (11/18)
- Does the difficulty of developing intuition about physics (mechanics, E&M, relativity, quantum, etc.,etc.) suggest that we (12/18)
- shouldn't teach it at all unless the student is fully pre-committed to a physics career? (13/18)
- @@@@ that. I say get them all in the door. Teach responsibly. Teach with humility. Pay attention to ed research. (14/18)
- If physics has a problem teaching students to reason responsibly, then physics needs to do its homework, figure out how to do that, and (15/18)
- teach its teachers how to teach. There are big communities doing just that in physics (physport.org), and math, and chemistry, (16/18)
- and I think there's one in statistics too. (17/18)
- Challenges are not a good argument for less education, they are an argument for better education. (18/18)
With statistics, it probably is true that one is more likely to be faced with life&death decisions, which statistical reasoning could illuminate, which one is more likely to get wrong with a little education than with zero. I think that probably doesn't happen much in physics. But what does happen, a lot, is people with a little experience in physics, or math, or education, or science in general, getting in over their heads when they try to use that in a different situation, one for which their educational experience wasn't intended to prepare them.
Consider how many people think they know a lot about climate research because they have personal experience with the weather.
Or the number of people who think they are experts in education because they were students once.
I think the antidote to this issue is teaching that is more focused on epistemology than on content. How do we know what we think we know? should be the central issue in any content-rich field. That will give students more experience with the humility appropriate to the human condition. It will make them more able to handle the uncertainty of new situations. It will help them be ready to create entirely new knowledge when the time is right. And it will help them be ready to place skeptical trust in experts, when their own knowledge isn't up to the task.