# How Calculators Arrive at the Answer: The Math Behind the Button

How a calculator works out an answer: order of operations, the % key, square roots by repeated guessing, rounding, and why two calculators can disagree.

---

- **Canonical URL:** https://dothecalculation.com/blog/math/how-calculators-arrive-at-the-answer
- **Category:** Math
- **Author:** Do The Calculation Team
- **Published:** 2026-10-07
- **Reading time:** 12 min read
- **Publisher:** Do The Calculation (https://dothecalculation.com)
- **Methodology:** https://dothecalculation.com/methodology

---

You press a few keys and a number appears. It looks instant, but the calculator has followed a fixed set of rules to get there: it has decided which operation to do first, turned your decimal numbers into a form its circuits can handle, run a short algorithm, and rounded the result to fit the screen.

Knowing those rules is useful. It lets you check an answer that looks wrong, explains why two calculators sometimes disagree, and shows you how to work the same problem by hand.

In this guide:

- What happens between pressing the keys and seeing the answer
- Order of operations (PEMDAS) and why basic calculators sometimes ignore it
- How the % key and square roots really work
- Why rounding can make a calculator look wrong
- Quick ways to check any calculator answer, with practice problems

## Quick Answer

A calculator reads your keys as an expression, applies the order of operations (if it is a scientific calculator), converts the numbers into binary or binary-coded decimal, and runs a fixed procedure for each operation. Addition and subtraction work digit by digit with carries, as you learned at school. Multiplication is done by shifting and adding, division by shifting and subtracting (a form of long division), and square roots by repeated improving guesses. The result is then rounded to the number of digits the screen can show.

_[Figure: What happens when you press "="]_

**What the calculator does for each operation**
| Operation | What happens inside | How you can do it by hand |
| --- | --- | --- |
| Addition | Adds digit by digit, carrying into the next place | Column addition |
| Subtraction | Subtracts digit by digit, borrowing (or adds a complement) | Column subtraction |
| Multiplication | Shift-and-add: one partial product per digit, then add them | Long multiplication |
| Division | Shift-and-subtract: repeated subtraction, place by place | Long division |
| Percent | Divides by 100, then multiplies (some models also add or subtract) | Move the decimal point two places |
| Square root | Makes a guess and improves it until it stops changing | Estimate, then refine |

## How Calculators Do the Four Basic Operations

### Addition and subtraction

For 18 + 34 the calculator adds the ones (8 + 4 = 12, write 2, carry 1), then the tens (1 + 3 + 1 = 5), giving 52. That is the same column method you use on paper. Subtraction is handled the same way with borrowing, or by adding a complement, which lets one circuit do both jobs.

### Multiplication: shift and add

A calculator does not add 7 to itself 4 times to get 28; that would be far too slow for 4,816 × 3,927. Instead it makes one partial product for each digit of the second number, shifts each one into place and adds them. In decimal, 47 × 23 = (47 × 20) + (47 × 3) = 940 + 141 = 1,081. The chip does the same thing in binary, where every partial product is either the number itself or zero, so each step is a shift and an add.

### Division: shift and subtract

Division works the other way round: the calculator subtracts the divisor, shifted to the right place value, as many times as it fits, then moves one place to the right and repeats. That is long division. For 56 ÷ 8 it finds that 8 fits into 56 seven times with nothing left. If you want to see each subtraction written out, the [long division calculator](/calculators/long-division-calculator) shows the full working, and the [long division step-by-step guide](/blog/math/long-division-examples-step-by-step-guide) explains the layout.

> **Repeated addition is the idea, not the method**: Multiplication means repeated addition and division means repeated subtraction. A calculator uses faster versions of both, working one place value at a time, so a 10-digit multiplication takes about 10 steps rather than billions.

## Order of Operations: PEMDAS

When an expression has more than one operation, a scientific calculator follows the order of operations, often remembered as PEMDAS (BODMAS or BIDMAS in the UK).

**PEMDAS**
| Step | Letter | Operation |
| --- | --- | --- |
| 1 | P | Parentheses (brackets): work inside them first, innermost first |
| 2 | E | Exponents and roots |
| 3 | M and D | Multiplication and division, from left to right |
| 4 | A and S | Addition and subtraction, from left to right |

> **M does not come before D**: Multiplication and division have the same rank, and so do addition and subtraction. Work each pair from left to right: 12 ÷ 3 × 2 = 4 × 2 = 8, not 12 ÷ 6 = 2.

### Worked example 1: 6 × 5 + 3 ÷ 2 − 6

A scientific calculator turns this into a tree. The operations lower in the tree are done first, and each result feeds the operation above it.

_[Figure: How a scientific calculator reads 6 × 5 + 3 ÷ 2 − 6 - Multiplication and division sit lowest in the tree, so they are worked out first. The subtraction at the top is done last.]_

### Worked example 2: 3 × (4 + 6 × (8 + 2))

- Innermost brackets: 8 + 2 = 10.
- Next brackets: 4 + 6 × 10 = 4 + 60 = 64 (multiply before adding).
- Outside: 3 × 64 = 192.

### Worked example 3: 4 ÷ 3 × (4 × 10¹⁵)

- Brackets: 4 × 10¹⁵ = 4,000,000,000,000,000.
- Left to right: 4 ÷ 3 = 1.3333…, then 1.3333… × 4 × 10¹⁵.
- Answer: about 5.333 × 10¹⁵ (5,333,333,333,333,333). A 10-digit screen shows it in scientific notation, as 5.333333333 × 10¹⁵ or "5.333333333E15".

If the "E15" on the screen is unfamiliar, the [scientific notation guide](/blog/math/scientific-notation-basics) explains how to read it.

## Why Two Calculators Can Give Different Answers

Type 2 + 3 × 4 into a basic four-function calculator and many will show 20. A scientific calculator shows 14. Neither is broken. They use different input logic.

_[Figure: The same keys, 2 + 3 × 4 =, on two kinds of calculator]_

**Other reasons answers differ**
| Cause | Example | What to do |
| --- | --- | --- |
| Input logic | 2 + 3 × 4 gives 20 or 14 | Use brackets or a scientific calculator |
| Implied multiplication | 6 ÷ 2(1 + 2) gives 9 on some models and 1 on others | Write it with brackets: (6 ÷ 2) × (1 + 2) = 9 |
| Degrees vs radians | sin(30) gives 0.5 in degrees, −0.988 in radians | Check the DEG/RAD indicator |
| Rounding | 1 ÷ 3 × 3 can show 0.9999999 or 1 | Round the final answer, not the steps |
| The % key | 100 + 8% gives 108 on many models, 100.08 on others | Type 100 × 1.08 instead |

## How Calculators Handle Percentages

A percent is a number out of 100, so the calculator divides by 100 and multiplies. 5% of 40 is 0.05 × 40 = 2. On most basic calculators, typing 40 × 5 % gives 2 directly.

Adding a percentage is where calculators differ. Many basic models treat 100 + 8 % as "add 8% of 100" and show 108. Others read 8% simply as 0.08 and show 100.08, and some scientific calculators have no % key at all. The method that works on every calculator is to multiply by 1 plus the rate: 100 × 1.08 = 108. That is exactly how sales tax is added to a price, as the [sales tax guide for online sellers](/blog/business/sales-tax-guide-for-shopify-and-etsy-sellers) shows, and the [percentage calculator](/calculators/percentage-calculator) handles every form of the question.

**Percent, three ways**

```
x% of N = N × x ÷ 100    ·    N increased by x% = N × (1 + x ÷ 100)    ·    N decreased by x% = N × (1 − x ÷ 100)
```
- Example: $100 plus 8% = 100 × 1.08 = $108. $100 minus 8% = 100 × 0.92 = $92.

## How Calculators Find Powers and Square Roots

Whole-number powers are repeated multiplication: 5³ = 5 × 5 × 5 = 125. For large powers a calculator saves time by squaring: 2¹⁶ is 2 squared four times (2, 4, 16, 256, 65,536) rather than fifteen separate multiplications.

Square roots of perfect squares are easy to check by hand: √144 = 12 because 12 × 12 = 144. For a number like 50, the calculator makes a guess and improves it. One common method, Newton's method, replaces a guess g with the average of g and 50 ÷ g. Starting from 7 (because 7² = 49):

**Finding √50 by improving a guess**
| Step | Guess | Guess² | Correct digits |
| --- | --- | --- | --- |
| Start | 7 | 49 | 2 |
| 1 | 7.0714286 | 50.0001276 | 4 |
| 2 | 7.0710678211 | 50.0000000001 | 8 |
| 3 | 7.0710678118655 | 50.0000000000000 | 15 or more |

_[Figure: Correct digits of √50 after each step - Each round of Newton's method roughly doubles the number of correct digits, so three steps fill a calculator screen.]_

Powers that are not whole numbers, such as 1.08^7.5, are worked out with logarithms behind the scenes. You can try any of these with the [exponent calculator](/calculators/exponent-calculator) or the [scientific calculator](/calculators/scientific-calculator).

## Rounding: Why the Last Digit Can Look Wrong

A calculator screen holds a fixed number of digits, often 8, 10 or 12, and it stores only a few more internally. Numbers like 1 ÷ 3 = 0.3333… never end, so the calculator must cut them off. Multiply that stored value by 3 and some calculators show 0.9999999 while others show 1, depending on how many hidden digits they keep and how they round.

Computers have a related quirk. Most programming languages store numbers in binary floating point, where 0.1 cannot be held exactly, so 0.1 + 0.2 comes out as 0.30000000000000004. Spreadsheets and phone calculators usually hide this by rounding the display. The practical rule is the same everywhere: keep full precision through the working and round only the final answer. The [rounding rules guide](/blog/math/rounding-rules-complete-guide) covers how to round correctly, and the [rounding calculator](/calculators/rounding-calculator) does it for you.

## How to Check a Calculator Answer

**Five quick checks**
| Check | How | Example |
| --- | --- | --- |
| Estimate first | Round the numbers and work it out in your head | 49 × 21 is about 50 × 20 = 1,000, so 1,029 is reasonable |
| Reverse the operation | Multiply to check a division, add to check a subtraction | 48 ÷ 6 = 8 because 8 × 6 = 48 |
| Check the decimal point | Count decimal places, or compare with your estimate | 0.4 × 0.3 = 0.12, not 1.2 |
| Check the sign | Two negatives multiplied give a positive | −4 × −5 = 20 |
| Use a second method | Work it by hand or on a different calculator | Long division to confirm 851 ÷ 3 = 283 R2 |

## Practice Problems (With Answers)

**Work these out, then check**
| Problem | Working | Answer |
| --- | --- | --- |
| 1. 6 × 5 + 3 ÷ 2 − 6 | 30 + 1.5 − 6 | 25.5 |
| 2. 3 × (4 + 6 × (8 + 2)) | 3 × (4 + 60) = 3 × 64 | 192 |
| 3. 5% of 80 | 0.05 × 80 | 4 |
| 4. √169 | 13 × 13 = 169 | 13 |
| 5. 2⁴ | 2 × 2 × 2 × 2 | 16 |
| 6. 15 + 20 × 3 − 10 | 15 + 60 − 10 | 65 |
| 7. 12 ÷ 3 × 2 | Left to right: 4 × 2 | 8 |
| 8. $250 increased by 6% | 250 × 1.06 | $265 |

## Common Calculator Mistakes

**Mistakes and fixes**
| Mistake | Fix |
| --- | --- |
| Typing a long expression into a basic calculator | Use brackets on a scientific calculator, or press = after each part |
| Doing multiplication before division in every case | M and D share a rank: go left to right |
| Rounding in the middle of a problem | Keep full precision, round the final answer only |
| Typing 5 instead of 0.05 for 5% | Divide the percent by 100, or use the % key |
| Forgetting to clear | Press AC (all clear) before a new problem; C or CE only clears the last entry on many models |
| Trusting the screen without an estimate | Estimate first so a misplaced decimal point stands out |

## Frequently Asked Questions

**How does a calculator work?**

It reads your keys as an expression, decides the order of operations, converts the numbers to binary or binary-coded decimal, and runs a fixed procedure for each operation: carrying for addition, shift-and-add for multiplication, shift-and-subtract for division, and repeated guessing for roots. It then rounds the result to fit the screen.

**Do all calculators follow PEMDAS?**

Scientific and graphing calculators do. Many basic four-function calculators do not: they work each operation as you type it, so 2 + 3 × 4 gives 20 instead of 14.

**How do calculators find square roots?**

They start with a guess and improve it with a short repeated calculation, such as Newton's method or a shift-and-add method. Each step roughly doubles the correct digits, so only a few steps are needed.

**Why does my calculator give a different answer from my friend's?**

The usual causes are different input logic (immediate execution vs PEMDAS), how implied multiplication is treated, degree vs radian mode, rounding, and how the % key works on each model.

**What does the % key do?**

It divides by 100. On many basic calculators it also adds or subtracts a percentage of the previous number, so 100 + 8 % = 108, but this varies by model. Typing 100 × 1.08 gives 108 on any calculator.

**What is the difference between C, CE and AC?**

On most calculators CE (clear entry) erases only the number you are typing, while AC (all clear) erases the whole calculation. On some models a single C key does both: press once to clear the entry, twice to clear everything.

## Final Summary

- Calculators follow fixed procedures: carry for addition, shift-and-add for multiplication, shift-and-subtract (long division) for division.
- Scientific calculators apply PEMDAS; many basic ones work left to right as you type.
- Square roots come from a guess that is improved until it stops changing.
- The screen shows a rounded value, so round only the final answer.
- Estimate first and reverse the operation to check any result.

To see the shift-and-subtract method the way it is taught in school, read the [4th grade long division guide](/blog/math/4th-grade-long-division-visual-models), which uses area models and the step-by-step layout. And for a real-world use of powers, the guide to [which investment has the highest long-term ROI](/blog/finance/which-investment-has-the-highest-long-term-roi) shows how (1 + r)ⁿ turns a steady return into large numbers over decades.

---

_Source: [Do The Calculation](https://dothecalculation.com/blog/math/how-calculators-arrive-at-the-answer). Quote freely with attribution and a link to this page._