Did I miss something? Did the mods move this into IMHO without me noticing? Did I seriously miss the posts where people gave actual evidence or arguments of any of these positions?
Here. I assert that for non-trivial problems, calculators are measurably faster than computation mentally, and there are precious few situations where solving on paper is more certain or quick then solving on a calculator.
I guessed .9 and a bit.
0.90909090909090909090909090909091
You know how I did that? In the course of computing grades where an assignment is 10 points and a test is 100 points, I got used to dividing numbers by 11. My estimation formula for n/11 is .((n - 1) +(x where x decreases as n increases). I worked that out simply by computing lots of n/11.
You say that we should be focused more on what math means than the answer? So, since integrals are defined as infinite Riemann sums, does that mean that we should be forced to evaluate integrals without using (x^(n+1) / n+1)?
(My calculus teacher thinks, in certain instances, yes, we should. Yes, I’m bitter.)
(x^(n+1) / n+1) and nDeriv() are both shortcuts. In both cases, you get students who will merrily plug 1/x into them, and move along. In fact, I would posit that you get more students writing down things like ((x^0)/0) then ERROR.
If you want to fight the attitude of “plug it in, get it out”, then fine (I disagree, but fine). But please distinguish between a formula memorized and a formula in a calculator.