4 ms·

News media ought to pay more attention to the confidence intervals associated with these metrics. Uncertainty is always present in statistics and the revisions

News media ought to pay more attention to the confidence intervals associated with these metrics. Uncertainty is always present in statistics and the revisions to these numbers rarely get reported (it was reported in this case to fit the tone of the story). News articles like these have a powerful psychological effect on the public, so shouldn't the media do a more responsible job by saying "oh by the way, it's 38,000 jobs +- 50,000"?

eloff 10y ago
Probably they figure they have to dumb it down for people to understand it. Which is hugely patronizing, and underestimates people I think. I wish the media would take their jobs more seriously in general, because, while there are scattered exceptions, by and large they distort facts, twist things, and misrepresent things to the point where it's nearly impossible to get an accurate view of the reality. Which is a complete shame, because they really have a powerful effect on people's views and those people go on to vote for idiot politicians and just plain wrong policies. I know in politics everyone has an opinion, but that doesn't mean all opinions are equally valid. More often than not the popular opinions are so off base with reality that it's sad, or scary, or both. And if you think it's bad in the US - look at Venezuela - a prime example of an ignorant and poor populace supporting bone-headed policies and politics until the country is reduced to a shambles. Democracy can only be as good as the voters, unfortunately - and the media bears a heavy responsibility there.
joobus 10y ago
> For example, the confidence interval for the monthly change in total nonfarm employment from the establishment survey is on the order of plus or minus 115,000. http://www.bls.gov/news.release/empsit.tn.htm http://www.bls.gov/news.release/empsit.tn.htm So 38,000 +- 115,000. In other words, they don't really know if employment went up or down.
- __derek__ 10y ago
  Thanks for digging that up. I thought it was around 100,000 and was about to go look for the actual number.
dgemm 10y ago
I have often thought the same thing - they are essentially reporting on something with a quite low signal to noise ratio, but they don't know what that is. And yet many important decisions are based on the monthly numbers.

eloff 10y ago

Probably they figure they have to dumb it down for people to understand it. Which is hugely patronizing, and underestimates people I think. I wish the media would take their jobs more seriously in general, because, while there are scattered exceptions, by and large they distort facts, twist things, and misrepresent things to the point where it's nearly impossible to get an accurate view of the reality. Which is a complete shame, because they really have a powerful effect on people's views and those people go on to vote for idiot politicians and just plain wrong policies. I know in politics everyone has an opinion, but that doesn't mean all opinions are equally valid. More often than not the popular opinions are so off base with reality that it's sad, or scary, or both. And if you think it's bad in the US - look at Venezuela - a prime example of an ignorant and poor populace supporting bone-headed policies and politics until the country is reduced to a shambles. Democracy can only be as good as the voters, unfortunately - and the media bears a heavy responsibility there.

joobus 10y ago

> For example, the confidence interval for the monthly change in total nonfarm employment from the establishment survey is on the order of plus or minus 115,000. http://www.bls.gov/news.release/empsit.tn.htm http://www.bls.gov/news.release/empsit.tn.htm So 38,000 +- 115,000. In other words, they don't really know if employment went up or down.

__derek__ 10y ago
Thanks for digging that up. I thought it was around 100,000 and was about to go look for the actual number.

dgemm 10y ago

I have often thought the same thing - they are essentially reporting on something with a quite low signal to noise ratio, but they don't know what that is. And yet many important decisions are based on the monthly numbers.