Science
Science calculators deal with uncertainty: summarizing measurements, standardizing scores, testing hypotheses and quantifying how likely an outcome is. They are the inferential counterpart to the descriptive tools in the math category.
Every result depends on assumptions about the data — independence, distribution, sample size — so the calculators are best used to interpret evidence rather than to produce certainty.
Browse science calculators
Inference & testing
Describing data
Probability & sampling
Which science calculator do you need?
Assess whether an observed difference is statistically significant under a chosen significance level.
Put a plausible range around an estimate from a sample.
Standardize a value relative to a distribution and read it against a normal curve.
Explore quartiles and the interquartile range of a data set.
Work out the probability of one or more events, including combinations.
Draw a reproducible random sample or simulate a random outcome.
Significance, intervals and standardized scores
A p-value is the probability of seeing data at least as extreme as yours if the null hypothesis were true. It is not the probability that the null hypothesis is true, and a small p-value is not a measure of effect size.
A confidence interval expresses the same evidence as a range: it shows the precision of an estimate rather than a yes/no verdict. A z-score re-expresses a value as a distance from the mean in standard deviations, which makes different distributions comparable.
Probability and random draws
Independent events multiply, mutually exclusive events add. Most probability errors come from losing track of whether events can occur together.
Random number generators are pseudorandom: with a fixed seed the same sequence can be reproduced, which is useful for reproducible sampling but unsuitable for security.
Where statistical results stop
- Statistical significance is not practical importance — always consider the size of the effect.
- Tests assume independence and a distributional shape; violations weaken the conclusion.
- Pseudorandom output is for simulation and sampling, never for passwords, keys or security decisions.