Low Down Poor Abused Data Blues

I’m having one of those days. One of those “I have a meeting that will finalize the analyses for a paper, let’s just check that ONE thing that didn’t matter in early versions of this model, so, it shouldn’t matter now, and OMGWTFBBQ everything changed!” kinds of days. Well, OK, not everything – just the new piece of the story that intrigued me the most because it was so delightfully intuitive. It’s made me feel a little blue. Well, a lot blue. So, fingers flying across the keyboard to fit new models, I flipped on some Muddy Waters. At the same time Miriam G suggested the Low Down Poor Abused Data Blues. While I’m no Robert Johnson, Muddy Waters, or BB King, clearly, this must be done. I’ve got a first verse. I’m welcome to contributions:

This is me. Right now. Except replace the guitar with a laptop furiously churning through R code.

Low Down Poor Abused Data Blues

Oh my data’s got autocorrelation,
eats up my treatment effect.

Oh my data’s got autocorrelation,
eats up my treatment effect.

But the decrease in sample size, it makes my analysis such a wreck.

2 thoughts on “Low Down Poor Abused Data Blues

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>