Scientific Calculator: 3 Mistakes That Give You Wrong Answers
Degree vs radian mode, the −2² trap, and what the memory keys are actually for. How to type expressions a scientific calculator can't misread.
Almost every “the calculator gave me the wrong answer” moment traces back to one of three things. None of them are the calculator’s fault, which is the annoying part.
Degrees or radians: decide before you press sin
Type sin(30) in degree mode and you get 0.5. Flip to radians and the same keystrokes return −0.988. Both answers are correct. Only one is the one you meant.
What makes this bite is that the mode is sticky and lives in a tiny corner of the display. Calculus and physics assume radians. A trigonometry worksheet almost always means degrees. If you switch contexts mid-session and forget to switch the mode, nothing warns you — the number just quietly comes out wrong.
The Scientific Calculator puts DEG and RAD as two buttons right above the keypad, and the answer recomputes the instant you toggle. Checking both takes about a second.
−2² is −4, and that’s not a bug
Exponents bind tighter than a leading minus sign, so -2^2 parses as -(2^2) and gives −4. If you want 4, you have to type (-2)^2. This is the standard convention in mathematics and in every calculator I’ve tested.
Excel is the famous exception. Put =-2^2 in a cell and it hands you 4, because Excel gives unary minus higher priority than exponentiation. Microsoft has documented this for years and won’t change it — too many spreadsheets depend on the old behavior.
Two more precedence facts worth internalising:
- Powers are right-associative.
2^3^2is2^(3^2)= 512, not(2^3)^2= 64. - Implied multiplication ranks the same as
×. So6/2(1+3)comes out as 12, not 1.
That last one is genuinely contested. The TI-82 gave implied multiplication higher priority than division; TI removed that behavior in the TI-83, and old textbooks still disagree with each other. There’s no secret correct answer to find. There’s just the boring fix: add parentheses and the ambiguity disappears.
Memory keys, which nobody explains
MS stores the current result. MR pastes it back into the expression. M+ adds the current result to what’s in memory, and M− subtracts it.
The everyday use is a running total across separate calculations. Say you’re pricing three line items with different tax rates: compute the first, hit M+, compute the second, hit M+, and so on. Memory holds the sum while your display stays free for the next line. It saves you from writing intermediate numbers on the back of an envelope, then squinting at your own handwriting.
Why 0.1 + 0.2 doesn’t show 0.30000000000000004
Because computers store decimals in binary, 0.1 has no exact representation — it’s a repeating fraction, the same way 1/3 is in decimal. Add it to 0.2 and the tiny rounding errors surface in the seventeenth digit.
Every calculator on your phone and laptop has this problem. They just round the display before showing you the result. This one rounds to 12 significant digits, which is enough to hide the noise and still show real precision when you need it.
What you can actually type
The input field is plain text, so keyboard and keypad work together. It accepts sin, cos, tan and their inverses, sinh/cosh/tanh, ln, log (base 10), log2, sqrt, cbrt, abs, exp, round, floor, ceil, factorial with !, remainder with mod, and the constants pi, e, tau, phi. Implicit multiplication works, so 2pi and 3(4+5) are valid. Enter records the result to history, Escape clears the line, and clicking any history entry loads it back for editing.
Everything runs in your browser. Nothing you type is uploaded, and the page keeps working offline once it has loaded. Open the Scientific Calculator and try 2^3^2 — if you expected 64, you now know why you didn’t get it.