Analysis

Stats

Jalaran Stats runs descriptive statistics, hypothesis tests, correlation and regression — and reports the assumptions behind each result alongside the result.

Open Stats →Create free account

Who Stats is for

For students and researchers doing the statistics a thesis actually requires, who have been handed a p-value by software before and had no idea whether the test was appropriate. The assumption checks are the part that matters.

What Stats does

Stats runs the standard inferential toolkit and reports, with each result, the assumptions that result depends on and whether they hold. That is the difference between a number and a finding: a t-test on data that violates its assumptions produces a perfectly formatted p-value that means nothing, and most statistics software will hand it to you without comment. Input is CSV or Excel and figures export for write-up.

Descriptive, hypothesis, correlation and regression
Assumption checks reported with every result
CSV and Excel input
Figure export for write-up

How Stats works

Load your data
CSV or Excel, straight from wherever you collected it.
Start descriptive
Distributions and summaries first. Half of all statistical mistakes are visible in a histogram.
Run the test
Hypothesis tests, correlation or regression, as your design requires.
Read the assumptions
Reported with the result. A p-value from a test whose assumptions fail is not a finding.

What Stats does not do

Stats will not choose your test for you or design your study — a well-executed test on a badly designed study is still worthless, and no software fixes that. It covers classical inference: no Bayesian methods, no multilevel or mixed models, no structural equation modelling. Machine learning is Analyze and forecasting is Predict, deliberately kept apart so each stays legible.

Common questions

Will it tell me which test to use?

No. Choosing the test follows from your research design, and a tool that picked one for you would be guessing about a decision you have to justify. It does tell you whether the assumptions of the test you chose hold.

Why report assumptions so prominently?

Because a p-value from a test whose assumptions are violated looks identical to a valid one. Most software hands it over without comment, and that is where a lot of published error comes from.