22.3 Dependent Errors and Time Series
Outside of controlled experiments, it isn’t possible to guarantee observations are independent of one another. Observations in a random sample that appear unrelated could be linked by an unknown, lurking variable. Without data that measure that lurking variable, plots don’t reveal the dependence. An exception, however, happens with data collected over time. Lurking variables are often correlated over time, allowing us to detect their presence even if we don’t know exactly what they are.
Detecting Dependence Using the Durbin-Watson Statistic
The key diagnostic plot for the simple regression model graphs the residuals on the explanatory variable. By removing the pattern that relates Y to X, the scatterplot ...
Get Statistics for Business: Decision Making and Analysis, 3rd Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
