Slashdot Mirror
'Calculators Killed the Standard Statistical Table' (sas.com)
theodp writes: In an obituary of sorts for the standard probability tables that were once ubiquitous in introductory statistics textbooks, Rick Wicklin writes: "In my first probability and statistics course, I constantly referenced the 23 statistical tables (which occupied 44 pages!) in the appendix of my undergraduate textbook. Any time I needed to compute a probability or test a hypothesis, I would flip to a table of probabilities for the normal, t, chi-square, or F distribution and use it to compute a probability (area) or quantile (critical value). If the value I needed wasn't tabulated, I had to manually perform linear interpolation from two tabulated values. I had no choice: my calculator did not have support for these advanced functions. In contrast, kids today have it easy! When my son took AP statistics in high school, his handheld calculator (a TI-84, which costs about $100) could compute the PDF, CDF, and quantiles of all the important probability distributions. Consequently, his textbook did not include an appendix of statistical tables."
I was in one of the last years my high school taught to use sliderules. Fancy ones already had trig scales. Didn't need the trig tables anymore.
I just don't see this as a problem. At some point, you have to consider whether NOT walking to school in 12 feet of snow up hill both ways somehow contributed to a better education that allows us to do the amazing things we do these days. Some things simply harder, without being better.
Which has more power: the hammer, or the anvil?
...is TI-84s still cost $100.
There used to be books of nothing but tables of logarithms and other mathematical tables, like trig functions. You used them when you needed more significant places in your answer than a slide rule could give you. I still have the one my dad used in college. They don't make those any more.
From the end of TFA:
It might be bad luck to speak ill of the dead, but I say, "good riddance"; I never liked using those tables anyway.
They're not saying it was good to have them. That's mostly implied from what the bit above, or from the submitter.
Slide rules, log tables, trig tables, knowledge on how to interpolate for intermediate values in those tables. In fact, much of the non-pictorial content in the math version of the CRC. Then graphing calculators killed off that part too. All that, and the monumental amount of work that it once took to compute those tables by hand, with occasional errors, before computing devices did them. One thing I don't lament is that those tables typically listed function values to a set number of decimal places, not significant figures. Or hauling around those big books.
When you run a statistical test they automatically run sanity checks which warn you if the data doesn't look like it should be used with that particular test.
This is a huge advance over the way statistics was done when I was a college student in the 80s, where it was common to collect the data and then go hunting for exotic tests that would give you a significant result because nobody had the time to check. Although I'm sure that's still done, it's a lot easier to double check someone's significance claims.
You spend a lot more time thinking about what data means these days, and that's a good thing.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
Today, those dollars are inflated. So the cost is in fact lower in terms of earnings.
Luxury! We had to calculate 50 digit figures in our heads from memorized tables for 23 hours a day.
Support Right To Repair Legislation.
I'm all for various forms of retro tech but, seriously, statistical tables? I don't want them back.
I like things such as slide rules, pocket calculators, and even statistical graph paper but tables don't aid my intuition one little bit. If I were to need a statistical table, I would calculate it with a spreadsheet or R or whatever tool handy.
First. World. Problems: We no longer waste paper to print archaic Mathematical tables /sarcasm OH NOES!
You know what else is "dead" ?
* Slide rule
* Tables of common Logarithms
* Tables of Trigonometric functions
Guess what, nobody is stopping you from buying those tables from old CRC Handbook of Chemistry and Physics which have them.
Apparently you didn't get the memo that a cheap calculator is "good enough."
What's next?
Whining that we don't have rotary telephones? Black and White televisions?
Like several of the posters above, I am of the age where trig, log (base 10 and e), statistics values were all found at the end of the textbook (I was also the last year at my high school to use a slide rule). What nobody has noted is what a pain in the ass it was to look up values in a table. Tables of these values aren't big, nicely spaced and easy to read like you see in a modern power point presentation, there were 100 or so rows per page (which means a smaller font) with 10 columns and maybe extra space after every five lines to make following them a bit easier.
If the author really wanted to lament something more significant, maybe he should have written about the loss of jobs for humans who made the calculations for these tables (they were called "computers") when machines started doing it for them. This may sound trivial but a couple of hundred years ago, ship navigators had to keep their books of trig values current because they were always finding errors in them - they way they found the errors was when other navigators, using the values ended up missing their destinations or running aground.
Mimetics Inc. Twitter
Now we're getting to the point where: "A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects."
Mimetics Inc. Twitter
They have it more convenient. There is nothing much difficult in consulting precomputed tables. With a calculator one achieves the same thing, only far more conveniently. You are not advocating to go back to using log and trig tables as well, are you? Or slide rules?
When you can buy a very capable calculator for $100 but a mathematics textbook is $200 you know something is backwards. And that $200's doesn't even get you any practical tables.
“Common sense is not so common.” — Voltaire
Yes they do. Stegun and Abramowitz ("Handbook of Mathematical Functions," Dover Books, originally NIST) is what I use and it's still in print. Cost is about $30 from Amazon. If you want tables of integrals, Gradshtyen and Ryzhik is also available, but I can't tell if it's still being printed or that's just new-old-stock copies they're selling. Cover is the same as my 25-year-old copy.
-JS
Vanity of vanities, all is vanity...
... to fall back on when you want to get up close to papyrus.. https://www.amazon.com/Million...
Hell, I took statistics more than 25 years ago and I'm pretty sure my old HP calculator could do all those functions. Certainly, we could do them with Quattro Pro. I just had to pull out my Statistics textbook ('91) to verify that it actually had those tables in it. I certainly don't recall using them even in the early 90s.
- Necron69
The thing I really miss is a good RPN calculator. Yeah, I could get an HP emulator, but it's just not the same thing.
I'd hang on to that book. It could actually be handy when EMP from the nuclear war knocks out stuff. Even the solar calculators may not work in nuclear winter. But your book will continue to work.
I don't think I'll be looking up chi square values after the bomb drops.
"That's the way to do it" - Punch
Just a few days ago I was watching an old documentary which stated that the impetus for Charles Babbage creating the Difference Engine was his frustration with dealing with inaccurate mathmatical tables. I was skeptical of this claim, but at least Wikipedia backs it up:
https://en.wikipedia.org/wiki/...
If you post as Anonymous Coward, don't expect a reply.
They still do.
CRC Handbook of Standard Mathematical Tables
https://www.amazon.com/Standar...
There was a short story (Asimov IIRC) about a future world where everyone had calculators, and then a man reverse engineered one and figured out that he could "simulate" its operations on paper. He finally convinced a skeptical general that it worked, who set in motion a plan to build a manned missile (suicide weapon) because a human pilot was cheaper than a computer.
The truly inventive part of this story is the new problem it found. Anyone could think of "No one knows how to use a table of logarithms anymore, or how to create one," after, e.g., an EMP war.
(Disclaimer: I'm an HP-41 fan myself, but I could create my own trig tables if I had to.)
'Calculators Killed the Standard Statistical Table'
That's the name of my favorite Buggles song!
I think.
- First they ignore you, then they laugh at you, then ???, then profit.
According to Wikipedia (https://en.wikipedia.org/wiki/Abramowitz_and_Stegun) "Because the Handbook is the work of U.S. federal government employees acting in their official capacity, it is not protected by copyright in the United States." At the end of the article are links to places you can download it. The successor to A&S, the NIST Handbook of Mathematical Functions (which has an online companion, https://dlmf.nist.gov/) omits the tables (referring to A&S or to other papers.)
How about using log tables to multiply 1.27546 by 10,271,398?
If and when that happens, the world would have far greater problems than dealing w/ multiplication or trigonometry. Going from an electrified society to one where the day lasts only as long as one can see is asking for major adjustments
"Good riddance." That's literally the conclusion in the article.
Basically, a newsletter produced by a company which makes statistical analysis software published a fluff piece laying some history on the kids and musing, "Aren't you glad we have software now?" The article is mildly nostalgic for the pre-calculator crowd and mildly interesting for the post-calculator crowd. It's not meant to be a controversial think-piece.
Chelloveck
I give up on debugging. From now on, SIGSEGV is a feature.
Yes and no. The survivors will initially have more to worry about than computing a sinc value of finding a F value, but if they want to rebuild to a technological civilization these tables will be amongst the most valuable artefacts on the planet. The ability to do complex calculations will cut decades off the recovery time for getting out technology back.
Just because you are paranoid does not mean that no-one is out to get you.
Each has its advantages. A calculator obviously can get you very precise numbers for critical values, probabilities, and the like. For those learning statistics though, teaching how to use a table helps to slow students down, reinforce their understanding of what they are actually doing, and facilitates a conceptual connection to the underlying probability distribution. It is very easy to mistype the wrong number into a calculator, or use the wrong function, or fail to calculate the probability of the appropriate tail(s). So by forcing a student to use a table, it helps them to think about what they're actually looking up. It's not quite the same as using a table of trig functions or logarithms, because those are single-argument functions, whereas a statistical distribution can have a density, cumulative probability, quantile, and/or degrees of freedom. And that ties into another reason to teach tables: unless everyone agrees to use the same calculator syntax, teaching the computation with a specific calculator will make students dependent on that particular syntax. R, SAS, SPSS, Excel, Mathematica--each one uses different syntax for the same computation. Now, tables are laid out differently too, but for the most part, they are consistent and not difficult to understand. Once you've seen one, you know how to use others.
So, I teach both, and I show how you get the same (or similar) results either way.
I wouldn't touch APK's software with a ten foot pole. He's not even providing digital signatures on his software.
You must assume it is infected with malware at any given time. Even if a third party tested APK Hosts File Engine you have no idea if what was tested is what is on his website.
Compounded on this is the requirement for running as root on Linux/BSD and the lack of good malware scanning on Linux.
Total garbage software from human garbage. APK can't code and he knows jack shit about security best practices.
Peeko
I teach college statistics, and I can assure everyone that standard statistical tables are alive and well. They are in the back of all introductory college statistics textbooks, and students are taught how to use them.
Chalk on buffalos? Pure luxury. Back when I learnt maths buffalo hadn't evolved and sedimentary rock hadn't yet formed. We used to use write in the ash from an erupting volcano while dodging their lava flows and running Fortran code in our heads and, if we were lucky, the volcano would explode and kill us all before the calculation was done.
....but you tell that to kids nowadays and they just don't believe you!
...but if they want to rebuild to a technological civilization these tables will be amongst the most valuable artefacts on the planet.
Hardly. It might take a bit of boring effort but it is not hard to calculate all the values in these tables from scratch by hand.
Back when the earth was still cooling, I bought a HP 48GX graphic calculator and spent so much time programming it, I actually failed calculus the first time. I had to take it again in summer school. Ah the good old days...
That's ok, I gave up my star charts and sextant/compass for GPS too. This is how modern technology works. There is a reason we don't teach people how to churn butter or plow a field either.
I was mourning my slide rules drop in popularity after calculators a long time before I would be troubled by standard statistical tables!
Yes, elementary function libraries are normally highly table driven and may only need a single linear interpolation at the cost of bigger tables, all for speed. The slickest method I've worked with was the accurate tables method, which gets better-than-expected accuracy by pre-computing the table values, not at a fixed interval, but at nearby points that just happen to have better accuracy than the type's precision (because the values are chosen such that, if they're represented in extended precision, they have extra binary zeros). They're expensive to find and compute, but you do that once per function to build the table that's then built into the library.
But to my original point, it's the specialists who do these today, not essentially every engineer using a printed table.
And I remember learning about the more accurate manual interpolation methods too, such as taking several table values and doing a curve fit. Ugh!
Ever tried it? There is a reason that Mr Babbage tried to build an analytic engine. Old artillery tables tended to be riddled with errors. Universities and research labs used to have hundreds of people devoting their lives to "computing" with pencil and paper. A book of statistical tables or logs used to take years to prepare. I don't know about you, but if I am trying to rebuild civilization I want to be able to do it before my children die of old age.
Just because you are paranoid does not mean that no-one is out to get you.
A book of statistical tables or logs used to take years to prepare.
True but this is just a matter of time: if you know what sine, logs etc are then you know enough to be able to calculate them by hand even if it will take years to make accurate tables. Now consider something like steel. Simply knowing that steel exists is not even close to enough information to be able to produce it. The same is true for the vast majority of materials around us. If we lose that information it will take many, many more years to recreate it.