VISUALIZING & USING DATA

ESSENTIAL CONCEPTS

  • Data can be visualized in a variety of ways. When considering what visualization to use, a statistician should consider the data set, how to make the data story clear, and the audience they are preparing for. Sometimes, visualizations are intentionally created to mislead the viewer or to “lie” with the data.


  • The goal of analyzing (univariate/bivariate) data is trying to construct a “true story'' about the contextual relationship. Doing this well often involves:

    • examining and/or interpreting the overall shape of the data in the context

    • choosing from and/or interpreting various measures of “typical-ness” (measures of center in 1-variable data; line of best fit in 2-variable data)

    • choosing from and/or interpreting various measures of variability (standard deviation/IQR/etc in 1-variable data; correlation coefficient in 2-variable data)

UNIT GUIDE

Visualizing and Using Data (IM1, Unit 3)