Well, I did it again.

I was going to follow up my discussion of predictive modeling talking about the {marginaleffects} package, which lets you summarize the average effect of changing one variable in your analysis.

My problem? I let my perfectionism win over my desire to communicate.

It’s an old habit. Sometimes you put something out in the world, and it is judged harshly, especially by those who are critical. Talking about model interpretation is filled with people like this. When you teach model interpretation, there is always someone who seems to wait for you to trip up. I don’t know what joy they get from this (probably very little), but they seem to rejoice in telling someone they’re wrong.

I know I can’t let these people win. They don’t take chances, so I shouldn’t give them any weight.

This is just to say that I know the next newsletter is late by a couple of weeks. I think I finally understand how to talk about {marginaleffects}, and I reached this point by not being hard on myself.

I see a lot of adult learners be so hard on themselves. They say things like “I’m not making enough progress,” or “I don’t know why I can’t understand this; everyone says it’s simple.”

The answer is that what you’re trying to learn is inherently hard, and you have to be gentle with yourself to progress.

When I teach, I am often fighting with the ghosts of teachers past who have taken up room in their student’s brains. These teacher’s effect’s linger even decades later. Teachers who said horrible things to their students, such as they were too stupid to understand something. I try to give students a measure of grace and forgiveness, and show them the joy of learning something again.