>> Click Here for Historical Wall Post Listing <<
Econintersect: Every day our editors collect the most interesting things they find from around the internet and present a summary "reading list" which will include very brief summaries (and sometimes longer ones) of why each item has gotten our attention. Suggestions from readers for "reading list" items are gratefully reviewed, although sometimes space limits the number included.
Is it fair to blame the West for trouble in the Middle East? (The Conversation)
Cost to rebuild Gaza? $4 billion, political will (Yahoo! News)
British hostage released in Libya after five months (Yahoo! News)
Jordan tackles homegrown Islamists as it joins attacks on Isis (Financial Times)
Report: ISIS plots to seize Iran’s nuclear secrets (Al Arabiya News)
There are 14 articles discussed today 'behind the wall'.
Today we have a lengthy rant on proper use of significant figures in economic data.
Do not miss "Other Economics and Business Items of Note", the final section every day.
Please support all that we do at Global Economic Intersection with a subscription to our premium content 'behind the wall'.
There are between 75 and 100 articles reviewed most weeks. That is in addition to the 140-160 articles of free content we provide.
You get a full year for only $25.
Let's change the units to 100 btu/$ GDP. The three numbers are now 72.56140, 73.29424 and 74.02718. Remember now, these are indistinguishable based on the ability to measure GPD. Lets round the numbers to three significant figures: 72.6, 73.3 and 74.2. These are identical because we cannot measure GDP with sufficient accuracy to tell the difference. Based on the assumed accuracy of measurement we know that 73.29424 is more than 72.5 and less than 74.3, but we do not know any more than that.
So Buchanan's point is that the seven significant figures in the EIA tables represent absolute nonsense. There is no information delivered after the first two significant figures. To present the data without misrepresentation the EIA table entry should be 7300 and there should be no decimal point after the second zero because a decimal point indicates that the second zero is a significant figure and there are only two.
Often the first non-significant figure is included in reporting a result, but that is only a correct presentation of data if the insignificant figure is identified in some clear manner. For example, the EIA table could have presented the 2011 data entry as 7,300 where red indicated the first insignificant figure. If the data were presented as 100 btu/$ GDP then the number would have been 73. (or 73.0 if the first insignificant figure were included).
Why is it worthwhile to include the first insignificant figure? This has value when the data is to be used in further calculations because it reduces the build-up of rounding errors. Consider five data points, each with two significant figures: 73., 7.3, 2.1, 8.6 and 21. Multiply the five numbers to get 202,107.654, which rounds to 200,000 (two significant figures which is not specified until we mark the first insignificant figure with 202,000).
Next rewrite the five numbers and include the first insignificant figure for each: 73.0, 7.30, 2.14, 8.63 and 21.4 Multiply these to get: 210,612.4609, which rounds to 211,000 with the third digit in the result marked as insignificant. By including the random distribution of first insignificant figures in each of the component data points and rounding to the correct significance at the end we have reduced accumulated rounding error. 211,000 is a more accurate result than 200,000
Final note: Significant figures are no stringer than the weakest link. If you have an experimental measurement that determined the average weight of an egg from a certain chicken farm is 1.012 ounces (three significant figures) and you have counted the eggs produced in a year to be 412,473 eggs (five significant figures), how many pounds of eggs are produced in a year? The product of 2.012 x 412,473 is 829,895.676 ounces. There are exactly 16.0000.... ounces per pound so the farm has produced 51,868.47975 pounds of eggs. The significant figures in a result is the smallest number of significant figures in the input, so the accurate answer is 51,870 pounds of eggs. If you wanted to give the tons per year the result would be 25.93 tons of eggs per year. The number of ounces in a pound and the number of pounds in a ton are exactly precise numbers with no uncertainty and cannot influence the accuracy of the result.
This type of exercise is commonly presented at the beginning of most introductory level science courses and students are expected to observe precision and accuracy rules in all their work.
Confronted by the failure of its principles, Europe insists it has no others. The US, on the other hand, has invented new ones as it goes along. Even the partially sighted can see the difference.
Read also If Europe insists on sticking to rules, recovery will be a dream (Wolfgang Munchau, Financial Times)
"The tapes reveal a basic cultural issue: The Fed can identify problems, but can't bring itself to make the banks fix those problems. We need congressional hearings to dig into what's gone wrong at the Fed, and we need to do it now because our whole economy is riding on the Fed's ability to supervise the biggest banks."
for the last two decades or so, not having a substantive conflict of interest policy has been Goldman’s business model. Representing both sides in mergers, betting alongside and against clients, and exploiting its informational edge wherever possible is simply how the firm makes its money. Forcing it to sharply reduce these conflicts would be potentially devastating.
Videotapes of Lectures for Economics 244: Discrete Choice Methods with Simulation (NERA Economic Consulting)
Have we done enough to regulate finance? (Simon Johnson, World Economic Forum)
After hack, bank customers need to be vigilant, experts say (Boston Globe)
Fact Checker: Has ACA raised Nevada health premiums? (Reno Gazette-Journal)
Silver's slide hurting investors, but may help shoppers (Pittsburgh Post-Gazette)
Econintersect Behind the Wall
|.... and keep up with economic news using our dynamic economic newspapers with the largest international coverage on the internet|
|Asia / Pacific|
|Middle East / Africa|
This Web Page by Steven Hansen ---- Copyright 2010 - 2017 Econintersect LLC - all rights reserved