Go read the preprint and compare the differences between Apple and IBM results with the standard deviation measured in scores. In Experiment 1, with no delay, the difference is 1.6 and the standard deviation is 3.3. This tells those of us who work with statistics that the result isn't meaningful, even ignoring the systematic errors mentioned in a previous post (Apple and IBM exposures occuring on different days!!). The same sort of statistical and dseign tomfoolery carries on throughout the article. I suppose publishing in psychology is a lot easier than in physics.
'"The congressional policy and agency practice of relying on the marketplace instead of regulation to maximize consumer welfare has been proven by experience (including the Comcast customer experience) to be enormously successful," concludes Comcast VP David L. Cohen's thinly-veiled warning to the FCC, filed on March 11. "Bearing these facts in mind should obviate the need for the Commission to test its legal authority."'
1) How does the Comcast user experience relate to marketplace control when Comcast is lobbies for and relies on government support to drive build local monopolies?
2) How does the Comcast user experience say anything positive about the status quo, considering that Comcast is notorious for faking speed test results (with accelerated transfers for the first 10-20 seconds, sufficient to trick a speed test but not very useful for most meaningful transfers), blacklisting customers with no warning, and of course breaking customers' p2p transfers, as is the case in the FCC issue that prompted this response from Comcast.
The FCC knows these things and Comcast knows that they know, so I'm not sure who the target of this statement might be. This might be an attempt to divide congress and the FCC on this issue and encourage them to fight each other.
Slashdot Mirror
Go read the preprint and compare the differences between Apple and IBM results with the standard deviation measured in scores. In Experiment 1, with no delay, the difference is 1.6 and the standard deviation is 3.3. This tells those of us who work with statistics that the result isn't meaningful, even ignoring the systematic errors mentioned in a previous post (Apple and IBM exposures occuring on different days!!). The same sort of statistical and dseign tomfoolery carries on throughout the article. I suppose publishing in psychology is a lot easier than in physics.