Do not go softly into that good R²
While ordinary least squares regression can be an unreasonably effective tool, it can pay dividends to dig deeper. The original chart shows a simple linear fit between AI subscription rates and private payroll changes, yielding a modest R² of 0.2111. However, running the diagnostic plots revealed a couple of