I am not endorsing anyone for the LAUSD election on Tuesday. My friend from Teach For America, Nick Melvoin, is running in Board District Four, and I told him “I hope that either you or Steve wins on Tuesday.” He told me that there is a high likelihood of that happening.
Just for fun, I thought I might create a model to predict the outcome of the election. I know – I have a very weird sense of what is fun.
I used my own special recipe of variables and created a multiple linear regression to make a predictive model based on the primary results. This model is completely unscientific and completely off the cuff. I am completely aware of the weaknesses of this methodology…
But who cares, here are the results of my model.
In Board District 4, my model predicted a very tight race – one that is within the error range. This means that even according to my model, it is too close to call.
In Board District 6, however, my model leans heavily toward Kelly Gonez. My model relies heavily on online interactions – and Gonez got a big boost online from some well-connected endorsements and an article in Teen Vogue. That might be throwing off my model quite a bit.
Anyway, I am totally willing to eat my words on Wednesday morning when this is completely wrong. But I had fun doing this – I guess thats what makes me a school data nerd.