Not At All Vital Stats 2024

keeping myself honest

It’s a little early to be grading myself for the year, writing-wise. Normally I wait until sometime in December, but I’ve got a bunch of other stuff to do right now.

My experiment with “The Lizard Thing” shows that the hybrid model of part-free / part-paid was not working for me.

Not sure what to do about that, but open to suggestions.

Outside of Substack, I’ve almost tied my record of 7 magazine / anthology acceptances in 2021, when I had way more free time, with about the same number of submissions and fewer total rejections.

2022 was an especially rough year, ego-wise. Most of those were repeated rejections of the same pieces, though, not new works. Screenshot by me, from The Submission Grinder. As always, thanks to Diabolical Plots for maintaining this as a free service.

Are those numbers statistically different from each other? I doubt it. As my students were learning this week, low sample sizes make results harder to interpret. They were flipping coins 100 times per pair of partners, and about once per class, we got a 60/40 result, which gives a chi-square value slightly above 4. If we choose the standard 1/20 chance of being wrong purely by luck (in other words, being 95% sure but no more than that), then we would reject the null hypothesis that the coin is fair in that specific case.

However, if we conveniently ignore the fact that the students were flipping different coins at their different lab tables and merge all their efforts into one big data set of 600 flips, then we get a much lower chi-sqare value.

0.24 is much less than 3.84

We have much less reason to believe the coin(s) are rigged. That’s the power of sample size.

Randall Munroe of xkcd harnesses the power of sarcasm and repetitive pedantry in Significance.

Personally, I wish we did have a rigged coin or two, randomly placed somewhere in the lab, and the students’ job was to find it. That would be way more fun. Students love catching us in mistakes. My personal policy is that when they do, I say,

Good job! You got me!

Then I let them wonder whether I did it on purpose.

We spend an entire week teaching meiosis, and students really struggle with it, mostly because they hate repetition (outside of video games) and they are so insistent that they can skip steps in the lesson. They can’t, any more than they could with a recipe, but kids born with a fast-forward button in their hands simply don’t believe that.

Of course multitasking is a real thing, they think.

Of course I can just imagine a complicated multi-step process in my head, from one example.

They believe that being able to repeat something back like a parrot, using the same words or slightly different words, is the same thing as understanding it. In other classes I make explicit use of Bloom’s Taxonomy to get them used to thinking about depths of processing (or heights, to use a more consistent visual metaphor).

Often shown as a pyramid. I like this version because there’s a left-to-right flow of time implied.

The result is that since they don’t really understand that chromosomes are independent of one another, they struggle with Punnet-square genetics the next week. We never even get to the more interesting case of using the squares to find a rare crossover event. We used to let the students play with FlyLab JS, but again, the overhead of learning to use yet another interface to repeat the same experiments prevented them from getting to the more interesting cases.

Of course the majors are going to do it all again in Genetics 392, but they’ve been seeing this stuff since 6th grade. Seeing it is not enough.

Polyhedral Peer Pressure

I learned about independent events from rolling dice at the D&D table, over and over again, in the course of telling fantastic stories with my friends. Understanding probabilities and taking advantage of them to win battles with monsters was a badge of honor in our nerdly community. Here’s a random example by Keith Ammann from The Monsters Know What They’re Doing on “Why These Tactics?”

A creature with a feature that gives it an advantage (or gives its enemy a disadvantage) will always prefer to use that feature. If it can’t, it may even shun a battle altogether. On average, an advantage or disadvantage is worth approximately ±4 on a d20 roll; with midrange target numbers, it can be worth as much as ±5. It can turn a 50/50 chance into 3-to-1 odds, or 3-to-1 odds into 15-to-1 odds . . . or the reverse. Advantage and disadvantage are a big deal.

He’s talking about 5th edition D&D, not the 2nd edition we were using in undergrad. My point is that our social environments make us willing to invest our time and effort into specific socially approved topics and tasks. My kid knew the stats of 750 Pokemon by heart, effortlessly, and yet actively fought against learning the times tables, even with the help of both parents and Schoolhouse Rock.

What Is Meiosis For?

An entirely typical description of this process lists and defines the stages without ever saying why cells would do this ridiculous, complicated dance of the chromosomes.

Illustration showing the nine stages of meiosis. Copied from the link above. Image credit: Laura Olivares Boldú / Wellcome Connecting Science

There’s a tiny little grayed-out note that says “recombination occurs,” which is really burying the lede. Meiosis is a reshuffling of the genetic deck, absolutely essential for us slow-breeding humans. Otherwise the microbes (some of whom can divide every 20 minutes) would quite simply eat us alive. That’s the important part; that’s what we should lead with.

Why do we not do this? Some logical intuition about building blocks? Some hope of building suspense towards a narrative payoff? Mostly not. Mostly it’s just blindly repeating how we were taught decades ago.

There are people experimenting with better ways.

• Teaching meiosis with the DNA triangle framework: A classroom activity that changes how students think about chromosomes

  • scroll down for similar articles

• Development of a meiosis concept inventory

  • Concept inventories started in physics in the late 1980s as a way of uncovering and diagnosing the vast diversity of student misconceptions about a particular issue. They were influential there, and spread somewhat into biology by the time I started teaching in the mid-2000s.

I have done a great deal of teaching and learning research over the course of my third career. I have published almost none of it¹, in part because nobody I know reads those academic journal articles, especially not classroom teachers like me who have 85 lab reports to grade every week. It seems much more likely to me that an interested reader will stumble across it here.

Not that reading it is any guarantee of adopting the innovation. Changing people’s behavior is a lot of hard work. Even as someone who is personally willing to drive two hours each way to attend a half-day interdisciplinary professional development conference, I’m embedded in a big machine whose solution to most problems is to switch publishers or homework software providers. I don’t necessarily have the ability to implement innovations in team-taught courses.

Which is one reason I jumped so enthusiastically into running the Biology department’s internship program, and the Freshman Seminars program, and why I’m developing my own upper-level courses in Science Writing and Science Policy, because there are no “standards” for those things. Nobody else wants to do those things, and hopefully that means they will leave me alone and let me do them the way I want.

I will leave you this week with an episode of Nature, one of my kid’s favorites, called My Life As a Turkey, where Floridian Joe Hutto (following the example of Austrian ethologist Konrad Lorenz) imprinted himself on a bunch of wild turkey hatchlings, allowing him to adopt the role of their mother and offering him a unique window onto the local wildlife. Even very skittish species gave him a pass whenever he was in the company of his brood.

Happy Thanksgiving!

1. This exercise, which I presented about on Friday at the SPARC conference last Friday, came out in 2013, based on work I did at the local community college in 2007-8. I’ve written about later versions of the same idea here and here.