Exclusive Content & Downloads from ASQ
Heuristics to Analyze Bivariate Data
Summary: It has been known that free-hand drawing of a rough ellipse to circle a scatterplot can be likened to obtaining some rough approximation for probability contours of a bivariate Normal distribution. Given the approximate probability contours, one can arrive at a useful conclusion pertaining to the linear relationship by a bivariate data set. The whole procedure boils down to a simple visual heuristic. However, it provides the means to verify results quickly and renders cross-checks for conclusions about the data and computational flaws. We illustrate the procedure on a data set.
Other Ways to Access content:
- Topics: Data Quality, Design of Experiments
- Keywords: Correlation,Probability,Regression,Curve fitting,Heuristics
- Author: Deligonul, Seyda
- Journal: Quality Engineering