how-to · 5 min read
Significant figures and scientific notation: reporting numbers honestly
Significant figures are the digits a measured value honestly supports. Scientific notation writes very large or very small values without a trail of zeros, and together the two conventions decide whether a reported result looks careful or claims precision that was never measured.
This guide walks through the counting rules, rounding without inventing precision, converting between decimal and scientific form, and switching between Celsius, Fahrenheit and Kelvin. The aim is a result you can defend line by line, with the unit and the uncertainty visible where they matter.
The rules that decide which digits count
Three rules cover almost every case. Non-zero digits always count, wherever they sit. Zeros between non-zero digits count because they were written between measured digits. Leading zeros never count, because they only place the decimal point. Trailing zeros are the tricky ones: they count when the number has an explicit decimal point, and are not counted when the number ends in zeros with no decimal point.
Written precision is part of the value, not decoration. 0.00450 has three significant figures, 1020 has three, 1200 has two, 1200. has four, and 6.022e23 has four. The difference between 1200 and 1200. is small in value and large in meaning: one claims two measured digits, the other four.
Round once, at the end
Rounding to a number of significant figures starts at the most significant digit and keeps that many digits. The next digit decides: 5 or more rounds the last kept digit up, less than 5 leaves it. A round-up can carry, as 9.99 to two significant figures becomes 10, written 1.0 × 10^1 to keep two digits visible. Round the final answer, not each intermediate step, or small errors accumulate.
Rounding can only lose precision, never add it. If a value is written with two significant figures, asking for four would invent digits; report it with two, or measure again with a finer instrument. When the rounded form needs trailing zeros to show its precision, write it in scientific notation.
Scientific notation without losing digits
Scientific notation writes a value as a coefficient between 1 and 10 multiplied by a power of ten. The exponent counts how far the decimal point moved, and its sign says which way: positive for large values, negative for small ones. 0.00045 becomes 4.5 × 10^-4, and 6.022e23 is the same notation typed without the multiplication sign.
Only the coefficient's digits count as significant figures, so converting into notation does not change the precision, and it should not be used to hide or add digits. Moving between forms is a rewrite: the same number with a different arrangement of the decimal point. For very large or very small values the plain decimal form can run to hundreds of digits, which is exactly why the notation exists.
Temperature scales are not simple ratios
Temperature scales have different zero points as well as different step sizes, so a ratio alone is wrong. Celsius to Fahrenheit is F = C × 9/5 + 32; Celsius to Kelvin is K = C + 273.15. The reverse directions use C = (F − 32) × 5/9 and C = K − 273.15. Absolute zero is -273.15 °C, -459.67 °F or 0 K, and -40 is the single point where the Celsius and Fahrenheit readings agree.
For gas-law and thermodynamics work, convert to kelvin before substituting into a formula, since those equations use absolute temperature. Report a converted temperature to the same number of significant figures as the reading that produced it: an instrument reading of 21.5 °C has three significant figures, and 70.7 °F keeps three rather than claiming more. Temperature differences use only the 5/9 ratio, not the offset.
Reporting a measurement honestly
Significant figures describe written precision; they are not an uncertainty budget. A careful report records the raw reading at the instrument's precision, carries extra digits through the calculation, rounds once at the end, and states the final value with its unit and uncertainty. For safety-critical or regulated work, follow the procedure your lab or course specifies rather than a general rule.
A quick check catches most errors. The exponent's sign should match the size of the number, the rounded value should be close to the original, and the number of significant figures should never grow without a reason. If a conversion produces extra precision, round it back to what the original measurement supports.
- Record the raw reading and the instrument's smallest division.
- Carry one or two extra digits through intermediate calculations.
- Round once, at the end, to the figures the measurement supports.
- Write very large or very small values in scientific notation.
- State units, and uncertainty where the procedure requires it.