Percentiles of weight in tons in the
xauto data
Roger B. Newson
Methods
We used the Mata-based SSC package rcentile to estimate
percentiles 0 to 100 by 25, with confidence intervals clustered by the
variable firm, meaning that we are sampling firms from a
population of firms, instead of models from a population of models. A
100\(\alpha\)th percentile of a
variable \(Y\) is defined informally as
a solution in \(\theta\) of the
equation \[
C_Y(\theta)=\alpha,
\] where \(C_Y(\cdot)\) is the
centre ridit function of \(Y\).
Results
The percentiles are plotted in Figure 1 and tabulated in Table 1.
Figure 1. Percentiles of weight in tons in the xauto
data (23 firms, 74 models)
Table 1. Percentiles of weight in tons in the xauto
data (23 firms, 74 models)
| Percent | Percentile | (95% | CI) |
|---|---|---|---|
| 0 | 0.880 | (0.880, | 0.880) |
| 25 | 1.120 | (1.035, | 1.365) |
| 50 | 1.595 | (1.325, | 1.685) |
| 75 | 1.800 | (1.655, | 1.940) |
| 100 | 2.420 | (2.420, | 2.420) |