Monday, November 29, 2010

I'm Making Toast

My last post claimed that the relationship between the settings on a toaster and the time spent toasting is not a linear relationship.  I worked this out using Dan Meyer's data from his blog dy/dan.  Dan responded that one data point changed the distribution.  I'm pretty sure that he is right, but judging from this comment by Dan, neither of us will rest until this is settled.

I've decided to collect my own data.  I'm recording the air temperature inside of the toaster to make sure that I start each trial at 80 degrees Fahrenheit.  I'm also setting the toaster randomly.  I have a super-geeky way to do make the random selections, as you can see below.

This is a slow process, as it takes 10 to 15 minutes for the  toaster to cool.  Hopefully I can finish collecting my data before the sacrificial loaf goes stale.  I'll have to write another NSF grant if it does.