XPost: misc.phone.mobile.iphone, comp.mobile.android   
   From: jollyroger@pobox.com   
      
   Chris  wrote:   
   > Jolly Roger  Wrote in message:   
   >> chris  wrote:   
   >>> On 09/12/2015 05:46, Jolly Roger wrote:   
   >>>> Paul M. Cook  wrote:   
   >>>>> On Tue, 08 Dec 2015 20:36:17 -0800, Savageduck wrote:   
   >>>>>   
   >>>>>> Your assumption is that accident rates should rise with cellphone   
   >>>>>> ownership.   
   >>>>>   
   >>>>> That is exactly the assumption I would make   
   >>>>   
   >>>> It's good you admit you are making this huge I'll-advised assumption.   
   >>>>   
   >>>>> if all these are true:   
   >>>>>   
   >>>>> 1. Cellphone use is distracting & distractions are dangerous   
   >>>>> 2. Accident rates compiled over the past 50 years are reliable   
   >>>>> 3. Cellphone ownership skyrocketed during a certain time frame   
   >>>>> & we assume a percentage of owners use them while driving.   
   >>>>   
   >>>> Of course you're ignoring the many factors unrelated to cell phone use   
   that   
   >>>> may decrease accident rates, because that's the only way your silly   
   >>>> argument can work.   
   >>>   
   >>> Exactly.   
   >>>   
   >>> Even if the accident rate did go up it wouldn't be proof that cellphone   
   >>> use causes accidents. It only shows that there is a *correlation*   
   >>> between the two. As we all know correlation does not imply causation.   
   >>> The correlation could be down to a totally unrelated effect. Just like   
   >>> ice cream sales don't cause shark attacks:   
   >>> https://simple.wikipedia.org/wiki/Correlation   
   >>>   
   >>> The only way to truly test whether one effect causes another is to test   
   >>> the effect explicitly. For example, take a suitably large number of   
   >>> people who drive (it might be 100 or 1000 or 10,000) monitor their   
   >>> driving for a suitable period of time (a week or month) both with and   
   >>> without a cellphone. Then test to see if there is an   
   >>> observable/statistical difference in the rate of accidents (or other   
   >>> measure such as 'near misses') when the people drove while using a   
   >>> cellphone vs when they didn't. Job done.   
   >>   
   >> Accident rate isn't applicable. You measure response times, driver errors,   
   >> and so on through direct observation. And that's exactly what studies have   
   >> done. Here is just one of several such studies:   
   >>   
   >>    
   >   
   > Maybe, maybe not. Paul is specifically taking about accidents.   
      
   He's fixated on accident rates because it's the only way his misguided   
   argument works.   
      
   --   
   E-mail sent to this address may be devoured by my ravenous SPAM filter.   
   I often ignore posts from Google. Use a real news client instead.   
      
   JR   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)