Here are some stuff I've read recently:
Is your laptop cooking your ...?, Reuters
As I'm writing this, I'm doing a very bad thing to my fertility -- this is if you believe story time by the people involved in this study. How should you read this article? No control group. 29 total participants (tiny sample). No mention of how they were selected, or why "young" was a necessary criterion for selection. Most important of all, the study measured the increase in temperature, but the link between temperature increase and fertility is unproven fiction. Such a link would be complicated to establish as I can imagine that some sustained level of use may be required to kill sperms.
How the Internet correctly predicted 36 of 37 Senate races, Randall Lane in Business Insider
In which Mr. Lane from the Daily Beast boasted about its "election oracle." What's good: innovative use of online text mining to help make election predictions; nice application of statistics. What's bad: the headline claim of "the Internet" predicting is badly misleading; scientists used data collected from the Internet to make predictions. Should you believe this? Accuracy rate has a denominator: as he discussed, it depends on whether you count all races or only close races. But what about races so close that the results are within the errors of measurement? What does it mean to have a predictive algorithm that predicts with "high accuracy" such essentially tied races?
Interest payments are not profits, Dean Baker
Baker reacted to this ridiculous NYT article that showed up over the weekend, proclaiming profit-taking in the Fannie/Freddie bailout. Say: your brother has fallen on hard times, and comes to you for help; you agree to provide him a loan, and at the family rate; turns out you don't have excess cash lying around so you go to the bank to get a loan; your brother pays you interest; you think you've made a profit from this arrangement. So many things wrong here. You have not gotten your principal back yet. And the profitability of this arrangement was determined long ago: it is the differential between the rate you are paying the bank and the (family) rate your brother is paying you. Lost in all of this: the government does not have free money; there is a cost to money-printing.
Comments