Abstract
This article presents a hands-on web-based activity motivated by the relation between human health and ozone pollution in California. This case study is based on multivariate data collected monthly at 20 locations in California between 1980 and 2006. Several strategies and tools for data interrogation and exploratory data analysis, model fitting and statistical inference on these data are presented. All components of this case study (data, tools, activity) are freely available online at: http://wiki.stat.ucla.edu/socr/index.php/SOCR_MotionCharts_CAOzoneData. Several types of exploratory (motion charts, box-and-whisker plots, spider charts) and quantitative (inference, regression, analysis of variance (ANOVA)) data analyses tools are demonstrated. Two specific human health related questions (temporal and geographic effects of ozone pollution) are discussed as motivational challenges.
Acknowledgements
This study was partially supported by the National Science Foundation, grants 0716055, 1023115 and 0442992, and by the National Institutes of Health, grant U54 RR021813. We are also thankful to the numerous users, students, collaborators and instructors who have provided feedback, critiques and ideas about improvements of the SOCR resources in the past several years. Valuable contributions from Dexter Whittinghill improved the readability of this article. The critiques of the International Journal of Mathematical Education in Science and Technology editors also enhanced the manuscript.