Monday, December 21, 2009

Using neighborhood data to create statistical models

High school math teacher Matt Townsley encourages students to use data that interests them as much as possible.  In a recent project, students were asked to answer the question, "Is size a useful predictor of a house's assessed value?" Data from websites such as the Johnson County Assessor were made readily available for students to access and research the ins and outs of houses on their own streets and in their housing developments.  Additional resources were posted on his website for students to use as they created a second statistical model with Iowa City or Cedar Rapids houses as data points.

Students were asked to produce a 1-2 page double-spaced report summarizing their findings:

  • A brief summary of the data in each mini-study (Solon neighborhood and CR/IC).  This included a response to the following question using data, graphs, regression lines, significance tests and correlation coefficients to backup your opinion - "Is size a useful predictor of price in your neighborhood? ...and in Iowa City or Cedar Rapids?"
  • A prediction of the students' house's assessed value based on square feet using his/her neighborhood model and the Iowa City or Cedar Rapids model.  Does this match-up with the assessed value according to the websites?  Explain a few possible reasons why or why not.
  • Finally, include a paragraph explaining any possible errors observed in the way data was collected or the process in which this study took place

Scatterplots, regression line equations and prediction intervals using Microsoft Excel were the norm for each student.

This project parallels quadrant D learning as written in the 9-12 Data Analysis/Statistics & Probability Iowa Core Curriculum essential concepts.

If you have any questions or comments, contact Matt at

No comments:

Post a Comment