In the last blog post we looked at some definitions of accuracy, precision, and randomness, ending by considering how accurately we need to know the value of Pi. Now, let’s consider how we measure things in the world, both scientifically and in human activity. Human activity includes the things we do in everyday life (like buying a gallon of gasoline, or is it a liter) and how we function interactively in society (like buying a gallon of gasoline).
The International System of Units (abbreviated SI) is documented in a brochure published by CGPM (Conference Generale des Poids et Measures). The US National Institute of Standards and Technology (NIST) publishes NIST SP 811 that gives general guidance for the use of SI in the US. NIST is a measurement standards lab and a non-regulatory agency under the US Department of Commerce. The United States uses “customary units” to be discussed later.
In the SI, there are seven base unist that are used for measuring everything physical. These base units are then used to define the secondary and other units of measurement. The base units include: meter (distance); kilogram (mass); second (time); Ampere (electrical current); Kelvin (temperature); mole (entities in a substance); and Candela (luminous intensity). The kilogram is the most reasonably understood because it was selected randomly from 40 prototypes of a 90 percent platinum – 10 percent iridium alloy in 1889. It is something physical that can be held and seen, albeit it is not passed around.
The other base units are more complicated to understand and group. For instance, the meter is the distance travelled by light in a vacuum in 1/299,792,458 second (with time being measured by a Cesium-133 atomic clock). And the mole is the amount on material containing number elementary entities (atoms or molecules) that equal Avogadro’s number (6.022140857 X 1023). And that’s probably more than you wanted to know.
In the United States, we do not use the metric system in totality, more commonly using the United States customary units (USCS), in place before it became an independent country. In the weight category, the Avoirdupois system is generally used with its smallest weight unit of the grain being 1/7000 lb., or 64.79897 mg. Maybe we should stick with ounces and pounds. If a candy bar weighs 1.25 ounces , how much accuracy in grains is needed?
Fluid ounces are based on the gallon, which is four quarts or 231 cubic inches. For completeness, a shot is 1.5 fluid ounces or 0.0234 gallons (rounded). Note that the gallon is specified by volume, allowing a fluid ounce to have any weight, dependent only on the density of the fluid. There are other units of volume including cubic inches, feet, and yards, based on cubing units of linear measure. In agriculture, we use bushels, defined as 2150.42 cubic inches. Next time you buy a bushel of green chile, measure the box – or not.
As previously stated, in the United States we have NIST. Its mission is to “Promote U.S. innovation and industrial competitiveness by advancing measurement science, standards, and technology in ways that enhance economic security and improve our quality of life.” In addition to the material and physical areas, NIST examines nanoscale science and technology, communications technology, engineering, information technology and neutron research.
On April 21, 2014, NIST removed a cryptographic algorithm from its draft guidance on random number generators. After much controversy, the Dual Elliptic Curve Deterministic Random Bit Ge3nerator lost any recommendation for use because it might have a back door, a means where by hackers could decode your encrypted message. In the last column, random number generators were discussed and in Cyber (Part Three) softwar (software without the “e”) was acknowledged. Here the two merge in cyberspace and secure communication. Yet again the question is asked, how much is enough and how accurate, precise, and secure do our systems have to be?