In this unit kids learn what it means to find a line of best fit, how to find it with technology, how to interpret the correlation coefficient, and how to make appropriate predictions. They should also know that there is non-linear regression, though that will get practiced along with each new function family later.

- There’s a premium on regression problems that we can actually check the answers to. Here’s one, via CME.
- Bowman Dickson has a post where he’s collected lots of sources of data. In the past I’ve used gestation period/size and international life expectancy.
- The MAP has an activity where students are tasked with deriving a metric to measure the strength of correlation with.
- PCMI has some good problems on lines of best fit, including some a series that guide students through developing a method for calculating the line of best fit.
- A collection of student responses to a rich task, which is helpful both for the task and for the student responses, which sniff out some conceptual issues kids have with linear regression.

Still looking for:

- An introductory activity that will help kids get the idea that not all regression is linear, but that doesn’t require students to understand other function families yet.

Here’s everything posted under ”Regression,” including stuff that didn’t make the cut.

### Like this:

Like Loading...

Have you seen Sarcasymptote’s post on messing with kids notions of linear regressions? http://sarcasymptote.wordpress.com/2011/05/25/mathematical-analysis-and-its-discontents-or-wtfcydwt/ Good stuff.

Thanks! Now I know about it, and I posted it on the site.