20% of them switched to not approving. That’s 1/5.
But iPod charts are fucking trash. You can’t include confidence intervals on interpolation splines like that. Those don’t look like restricted cubic splines or b splines, and that chart is not a dense lowess chart.
They strangle the fuck out of that data to hide that they have like 8 data points.
Rural Americans keeping the curve nice and flat
20% of them switched to not approving. That’s 1/5.
But iPod charts are fucking trash. You can’t include confidence intervals on interpolation splines like that. Those don’t look like restricted cubic splines or b splines, and that chart is not a dense lowess chart.
They strangle the fuck out of that data to hide that they have like 8 data points.