On Interruptions

Last month, Ben Brooks ran a poll on his website asking readers their attitudes toward interrupting other people using various writing devices. I reached out to Ben and he provided with me the poll data for analysis.

With my initial, and somewhat hasty, pass through the data, I sent the results back to Ben, which he posted.

The thing is, in my haste, I goofed a bit. I had originally ran chi square tests and identified the cells with the largest deviation as being “significantly different”. Max Masnick, an epidemiology PhD student, correctly pointed out that a chi square test only identifies that something is different; it does not identify what is different. For that, you need some additional testing.

I recently went back, revisited the data, and updated my analysis. And Ben kindly posted that update.