All Free Statistics Tools

Choose by the question you need to answer

DistriScope separates probability models, model comparison, convergence, data fitting, and hypothesis tests because those tasks answer different questions. Start with Distribution Explorer when you already have a named probability distribution and want to understand its parameters or calculate a point, cumulative, or tail probability. Open Family Comparator when the goal is to compare how several configured curves differ in support, shape, and spread. Use Distribution Convergence when the question is why one reference distribution can approximate another as a parameter grows.

Data Fitting begins with observed numerical values. It can rank the supported candidate families and show graphical context, but its output is an analytical aid rather than automatic proof that a model generated the data. Hypothesis Test Calculator begins with a population question and a design: one mean, two independent means, paired differences, three group means, a 2 by 2 categorical table, or two variances. Its module guides help you choose among the eight implemented procedures before entering values.

The 20 module pages below are stable, shareable landing pages. Each one defines the method in server-rendered English, states when it is and is not appropriate, explains inputs and formulas, includes a worked example, and loads the matching compact tool without asking you to select the main module again. Parameters, tail choices, and other controls that belong to the calculation remain available. Every module also links to related concepts and the methodology page so a number is never presented without its assumptions and limits.

For a responsible workflow, write the random variable or hypothesis first, identify the observational unit, preserve measurement units, and check whether observations are independent. Use the interactive output to support that written question, then verify consequential calculations with an independent implementation. A small p-value is not the probability that a null hypothesis is true, a close curve overlay is not a goodness-of-fit test, and a density height is not an exact-value probability for a continuous variable.