Benford's Law
gives a distribution that turns up surprisingly often. Given the small sample
size of figures here, I'm not entirely sure how it would
help.
Nonetheless:
We have 23 numbers in total. I've removed
those phones for which no damages were awarded, and ordered the others in order
of first digit. See table below:
Galaxy S II (Epic 4G
Touch).100,326,988
Exhibit 4G . . . . . . . . . .1,081,820
Epic 4G. .
. . . . . . . . .130,180,894
Fascinate. . . . . . . . .
.143,539,179
Indulge . . . . . . . . . . .16,011,184
Continuum . . . .
. . . . . .16,399,117
Nexus S 4G . . . . . . . . . .1,828,297
Galaxy
Tab . . . . . . . . . .1,966,691
Galaxy S (Showcase) . . . .
.22,002,146
Galaxy S II (Skyrocket) . . .32,273,558
Replenish. . . . .
. . . . . .3,350,256
Galaxy S II (AT&T). . . . . .40,494,356
Gem.
. . . . . . . . . . . . .4,075,585
Infuse 4G . . . . . . . . .
.44,792,974
Droid Charge. . . . . . . . .50,672,869
Mesmerize . . . .
. . . . . .53,123,612
Galaxy Prevail. . . . . . . .57,867,383
Galaxy S
4G . . . . . . . . .73,344,668
Galaxy Tab 10.1 WiFi . . . . .
.833,076
Captivate . . . . . . . . . .80,840,162
Galaxy S II
(T-Mobile). . . .83,791,708
Vibrant . . . . . . . . . .
.89,673,957
Transform. . . . . . . . . . . .953,060
TOTAL. . . . . . .
. . . .1,049,423,540
By Benford's Law, approximately 30% of these
figures should start with a 1. That adds up to 6.9 - very close to seven. Eight
of them start with a 1.
17.6% should start with a 2. That's very close to
four numbers. Only one starts with a 2. (Going back to the original verdict
form, though, I notice that there were three that started with a 2; the other
two appear to have been eliminated by the corrections).
For the full
digit-by-digit list, see table 2 below:
First digit . . Expected . . .
Found
1 . . . . . . . 6.924. . . . . .8
2 . . . . . . . 4.050. . . . .
.1
3 . . . . . . . 2.874. . . . . .2
4 . . . . . . . 2.229. . . . .
.3
5 . . . . . . . 1.821. . . . . .3
6 . . . . . . . 1.540. . . . .
.0
7 . . . . . . . 1.334. . . . . .1
8 . . . . . . . 1.177. . . . .
.4
9 . . . . . . . 1.052. . . . . .1
This seems like a pretty
close match to Benford's Law to me. There's a few more 8's than one might
expect, but I don't think that's significant.
I'm still not sure what, if
anything, that proves, though. [ Reply to This | Parent | # ]
|