These aren't "stats" they are just Squartrade's summary data
The stats themselves aren't terribly interesting to me (perkj brings up an excellent point about having all the other manufacturers' data lumped into single categories) but I still think the case issue is legit.
-snip-
Just random food for thought
Remember these aren't "stats". Square trade provided no tests of differences, just their summary data from their report records. These data are not an independent survey of phone failures. That would require randomly calling up smartphone owners and trying to balance your sampling across different phone types. Since they didn't do this, statistical tests may not even be appropriate. I have to say though, the fact that all their figures are made in Excel just further belittles my impression of the companies mathematical sophistication.
1.6% of 1,000,000 phones = 16,000 phones
16,000 x ~$500 phone = $8,000,000
Care to pay the difference on that "statistical dead heat"?
Point made in that the bottom line the insurance company feels would actually have little to do with actual statistical differences.
As for those interested in why the other companies' phone models were combined, see the full report for which models they looked at. Here's an excerpt:
"
Motorola: Droid, Droid X, Cliq
HTC: Nexus One, Evo, Droid Incredible, Aria, Hero
Blackberry: Curve, Bold, Storm, Torch, Pearl
Other Smart Phones includes phones from Samsung, LG, Palm, Nokia, Sony-Ericsson, and other minor manufacturers.
"