Precision Math: Significant Figures, Rounding, Ratios, and Roots Explained

You measure a table as 4.5 feet long and 3.25 feet wide. You calculate the area: 4.5 x 3.25 = 14.625 square feet. But your calculator shows 14.625, and you write that down. The problem is that 4.5 has only two significant figures. Reporting 14.625 implies you know the area to the nearest thousandth of a square foot. You do not. The correct answer is 15 square feet.

Precision matters in science, engineering, and medicine. Reporting too many digits implies false accuracy. Reporting too few discards real information. Significant figures are the language of precision. They tell the reader exactly how much you know and how much you do not.

This guide covers four precision math concepts: significant figures, rounding methods, ratios and proportions, and square roots. Each has specific rules that scientists, engineers, and students must follow.

Significant Figures: The 6 Rules

Significant figures (sig figs) are the meaningful digits in a measured or calculated number. They include all certain digits plus one estimated digit. The rules for counting them are precise.

Rule 1: Non-Zero Digits Are Always Significant

Every digit from 1 through 9 counts as significant.

• 4892 has 4 sig figs
• 3.14 has 3 sig figs
• 99 has 2 sig figs

Rule 2: Captive Zeros Are Significant

Zeros sandwiched between non-zero digits are significant because they represent measured values.

• 1005 has 4 sig figs
• 3.02 has 3 sig figs
• 10101 has 5 sig figs

Rule 3: Leading Zeros Are Not Significant

Zeros at the beginning of a number are placeholders. They do not count.

• 0.0052 has 2 sig figs
• 0.0003 has 1 sig fig
• 007 has 1 sig fig (leading zeros)

Rule 4: Trailing Zeros in Decimals Are Significant

Zeros after the decimal point at the end of a number indicate precision. They count.

• 15.00 has 4 sig figs
• 3.50 has 3 sig figs
• 0.0200 has 3 sig figs (leading zeros do not count, trailing zeros do)

Rule 5: Trailing Zeros in Whole Numbers Are Ambiguous

Without a decimal point, trailing zeros in a whole number may or may not be significant. The number 1500 could have 2, 3, or 4 sig figs. Scientific notation removes the ambiguity.

• 1.5 x 10^3 = 2 sig figs
• 1.50 x 10^3 = 3 sig figs
• 1.500 x 10^3 = 4 sig figs

Rule 6: Exact Numbers Have Infinite Sig Figs

Numbers from definitions (1 meter = 100 centimeters), counting (12 eggs), and mathematical constants do not limit calculation precision.

• 1 km = 1000 m (exact, infinite sig figs)
• 12 objects (exact count, infinite sig figs)
• Pi (mathematical constant, infinite sig figs)

Use our Significant Figures Calculator to count sig figs, round numbers, and perform calculations with proper precision.

Calculations with Significant Figures

The rules differ for multiplication/division versus addition/subtraction. Using the wrong rule is the most common mistake in chemistry and physics classes.

Multiplication and Division: Match Fewest Sig Figs

The result must have the same number of significant figures as the input with the fewest sig figs.

Example: 4.56 (3 sig figs) x 1.4 (2 sig figs) = 6.384, rounded to 6.4 (2 sig figs)

Example: 10.25 (4 sig figs) / 2.0 (2 sig figs) = 5.125, rounded to 5.1 (2 sig figs)

Addition and Subtraction: Match Fewest Decimal Places

The result must have the same number of decimal places as the input with the fewest decimal places.

Example: 12.52 (2 decimals) + 349.0 (1 decimal) + 8.24 (2 decimals) = 369.76, rounded to 369.8 (1 decimal)

Example: 1500 (ambiguous) + 25.3 (1 decimal) = 1525.3, rounded to 1525 (nearest whole number, assuming 1500 has no decimal places)

Mixed Operations

For calculations with both multiplication/division and addition/subtraction, track sig figs through each step. Keep one extra "guard digit" during intermediate steps and round only the final answer. NIST guidance explicitly states: do not round intermediate calculations, as this introduces cumulative errors.

Use our Rounding Calculator to round numbers to any precision.

Rounding Methods

Rounding seems simple until you encounter the tie-breaking case. Different methods handle it differently, and choosing the wrong one can bias your results.

Standard Rounding (Round Half Up)

If the digit to drop is 5 or greater, round up. Otherwise, round down.

• 4.35 rounded to 2 sig figs = 4.4
• 4.34 rounded to 2 sig figs = 4.3
• 4.449 rounded to 2 sig figs = 4.4

This is the method most people learn in school. It introduces a slight upward bias because 5 always rounds up.

Banker's Rounding (Round Half to Even)

When the digit to drop is exactly 5 (with nothing after it), round to the nearest even digit. This reduces systematic bias over many calculations.

• 4.35 rounds to 4.4 (4 is even)
• 4.25 rounds to 4.2 (2 is even)
• 4.45 rounds to 4.4 (4 is even)
• 4.55 rounds to 4.6 (6 is even)

NIST recommends banker's rounding in its calibration guidelines (GLP 9) because it avoids the asymmetry of standard rounding. In a dataset of many numbers ending in 5, standard rounding biases the mean upward. Banker's rounding eliminates this bias.

Rounding Up (Ceiling)

Always round up regardless of the digit. Used for conservative estimates in engineering and construction.

• 4.1 rounds to 5
• 4.9 rounds to 5

Rounding Down (Floor)

Always round down regardless of the digit. Used when you need to ensure you do not overestimate.

• 4.9 rounds to 4
• 4.1 rounds to 4

NIST Rounding Guidance

NIST Special Publication 811, Appendix B provides three rounding options for reporting measurement uncertainties:

1. Even/Odd Method: Banker's rounding. Preferred for statistical work to avoid bias.
2. Standard Spreadsheet Rounding: Round half up. Used in Excel and most software.
3. Round Uncertainties Up: Always round uncertainties up. Conservative approach for compliance testing.

The key NIST rule: never round intermediate calculations. Round only the final result.

Ratios and Proportions

A ratio compares two quantities. A proportion states that two ratios are equal.

Ratio Format

Ratios can be written as:

• a:b (colon notation)
• a/b (fraction notation)
• a to b (word notation)

Proportion Formula

If a/b = c/d, then a x d = b x c (cross multiplication).

Example: Recipe Scaling

A recipe uses 2 cups of flour for 12 cookies. How much flour for 30 cookies?

2/12 = x/30 12x = 2 x 30 12x = 60 x = 5 cups

Example: Map Scale

A map has a scale of 1:50,000. Two points are 4 cm apart on the map. What is the real distance?

1/50000 = 4/x x = 4 x 50000 = 200,000 cm = 2,000 m = 2 km

Aspect Ratio

Aspect ratio compares width to height. Common examples:

Format	Ratio	Application
16:9	1.78:1	HDTV, YouTube
4:3	1.33:1	Old TV, iPad
21:9	2.33:1	Ultrawide monitors
3:2	1.5:1	Photography (35mm)
1:1	1:1	Instagram posts

Use our Ratio Calculator to simplify ratios, solve proportions, and compare quantities.

Square Roots and Cube Roots

Square Root

The square root of a number is the value that, when multiplied by itself, gives the original number.

sqrt(25) = 5 because 5 x 5 = 25 sqrt(144) = 12 because 12 x 12 = 144

Estimating Square Roots

For non-perfect squares, estimate using nearby perfect squares:

sqrt(50) is between sqrt(49) = 7 and sqrt(64) = 8. Closer to 7 because 50 is closer to 49. Actual value: 7.07.

Cube Root

The cube root of a number is the value that, when cubed, gives the original number.

cbrt(27) = 3 because 3 x 3 x 3 = 27 cbrt(125) = 5 because 5 x 5 x 5 = 125

Nth Root

The nth root of a number is the value that, when raised to the nth power, gives the original number.

nth_root(x, n) = x^(1/n)

4th root of 81 = 81^(1/4) = 3 because 3^4 = 81.

Use our Root Calculator to calculate square roots, cube roots, and any nth root.

Real-World Scenarios

Scenario 1: Lab Report Precision

You measure the mass of a sample as 2.35 g on a balance that reads to 0.01 g. You measure the volume as 5.1 mL using a graduated cylinder that reads to 0.1 mL.

Density = 2.35 / 5.1 = 0.460784... g/mL

Sig figs: 2.35 has 3 sig figs. 5.1 has 2 sig figs. The result must have 2 sig figs.

Reported density: 0.46 g/mL

Reporting 0.460784 would imply you know the density to 6 significant figures, which is false. Your volume measurement only supports 2.

Scenario 2: Construction Material Estimation

You need concrete for a pad that is 4.2 meters x 3.1 meters x 0.1 meters.

Volume = 4.2 x 3.1 x 0.1 = 1.302 cubic meters

For construction, round up to ensure you have enough: order 1.4 cubic meters (rounding up with 10% waste factor).

If you used sig fig rules instead: 4.2 (2 sig figs) x 3.1 (2 sig figs) x 0.1 (1 sig fig) = 1 cubic meter. But in construction, precision rules do not apply. Practical estimation does.

Scenario 3: Pharmaceutical Dosing

A medication dosing chart says 5 mg per kg of body weight. A patient weighs 72.4 kg.

Dose = 5 x 72.4 = 362 mg

The "5 mg/kg" is an exact dosing instruction (infinite sig figs). The weight measurement (72.4 kg) has 3 sig figs. The result has 3 sig figs: 362 mg.

But in practice, medications come in specific doses (e.g., 50 mg tablets). The practical dose is 360 mg (7.2 tablets, rounded to 7 tablets = 350 mg) or 350 mg as prescribed by the physician.

Scenario 4: Financial Ratio Analysis

A company has revenue of $4,523,000 and expenses of $3,891,000.

Profit margin ratio = (4,523,000 - 3,891,000) / 4,523,000 = 632,000 / 4,523,000 = 0.1397 = 13.97%

In financial reporting, this is typically rounded to 1 decimal place: 14.0%. The precision of the input (thousands) supports this level of reporting.

Common Mistakes

1. Using multiplication sig fig rules for addition. 12.1 + 1.15 should be rounded to 1 decimal place (13.3), not 3 sig figs. Addition uses decimal place rules, not sig fig counting.

2. Reporting calculator output without rounding. Your calculator shows 12 digits. If your measurements have 3 sig figs, your answer should have 3 sig figs. Reporting 3.14159265359 when your measurement supports 3 sig figs is wrong.

3. Rounding intermediate steps. NIST explicitly warns against this. Rounding at each step introduces cumulative errors. Keep guard digits and round only the final answer.

4. Confusing standard rounding with banker's rounding. Standard rounding (round half up) biases results upward. Banker's rounding (round half to even) eliminates this bias. Use banker's rounding for statistical work and NIST-compliant reporting.

5. Treating exact numbers as measured. A dosing instruction of "5 mg/kg" is exact. It does not limit sig figs. Only measured quantities (weight, volume, temperature) limit precision.

External Research and Resources

• NIST Special Publication 811: Guide for the Use of the International System of Units (SI) provides the official U.S. government guidance on significant figures, rounding methods, and unit conversion in scientific and technical publications. Appendix B.7 covers rounding rules.
• NIST GLP 9: Rounding Expanded Uncertainties and Calibration Values details three rounding options (even/odd, standard, round up) for calibration laboratories and explicitly states that intermediate calculations should not be rounded.
• NIST: Unit Conversion provides dimensional analysis methodology and emphasizes using exact conversion factors and rounding only the final result.

People Also Ask

What are the rules for significant figures?

There are 6 rules: (1) non-zero digits are always significant, (2) zeros between non-zero digits are significant, (3) leading zeros are not significant, (4) trailing zeros in decimals are significant, (5) trailing zeros in whole numbers are ambiguous (use scientific notation), (6) exact numbers have infinite sig figs. Use our Significant Figures Calculator to count and round.

How do I round numbers correctly?

For standard rounding: if the digit to drop is 5 or greater, round up. For banker's rounding (NIST preferred): if the digit is exactly 5, round to the nearest even digit. Always round only the final answer, never intermediate calculations. Use our Rounding Calculator for instant results.

What is the difference between multiplication and addition sig fig rules?

For multiplication and division, the result has the same number of sig figs as the input with the fewest sig figs. For addition and subtraction, the result has the same number of decimal places as the input with the fewest decimal places. Using the wrong rule is the most common mistake in chemistry and physics.

How do I calculate a ratio or proportion?

A ratio compares two quantities (a:b). A proportion states two ratios are equal (a/b = c/d). Cross multiply to solve: a x d = b x c. For example, if 2 cups of flour makes 12 cookies, then 2/12 = x/30, so x = 5 cups for 30 cookies. Use our Ratio Calculator for instant results.

Related Calculators

• Significant Figures Calculator - Count and round sig figs
• Rounding Calculator - Round numbers to any precision
• Ratio Calculator - Simplify ratios and solve proportions
• Root Calculator - Calculate square, cube, and nth roots
• Square Root Calculator - Calculate square roots
• Scientific Notation Calculator - Convert to scientific notation
• Scientific Calculator - Advanced mathematical calculations
• Percentage Calculator - Calculate percentages
• Factorial Calculator - Calculate factorials
• Number Rounder - Round to any decimal place