Statistical Test Decision Checker for Student Projects
Choose what your research question actually asks. This decision aid identifies a candidate family of methods and the assumptions you must inspect. It never analyzes uploaded data or declares your study valid.
What this tool actually checks
The program uses your answers to distinguish independent groups from repeated or paired observations; numeric outcomes from categorical counts; and linear association questions from group comparisons. It is not connected to your data, a statistical database or an AI inference API. Even when you mark assumptions as reviewed, the status describes your declared screening progress, not independent verification or statistical validity.
Read the accompanying step-by-step guide to checking AI-suggested statistical tests before using a candidate test. When a method and its assumptions really fit, continue to the statistical calculators with the actual course methodology. Do not present the tool's candidate choice as a p-value or conclusion.
Worked classification example
Twelve students take an assessment before and after the same course. Each before-score belongs to the same student as an after-score, so there are 12 paired differences, not 24 independent participants. Choose the paired-change branch. The candidate is a paired t-test, subject to inspecting the distribution of those differences, missing pairs and unusual observations. If those same 24 scores came from two unrelated groups instead, choosing a paired test merely because the numbers look similar would be a design error.
For two categorical variables, expected counts are not the observed cell totals: each expected count equals (row total × column total) ÷ grand total under the independence model. Small cells warrant additional review; this checker stops rather than automatically endorsing a chi-square approximation.
Useful methodology references: Penn State STAT ONLINE on independence and expected counts, NIST on one-way ANOVA assumptions, and NIST on residual diagnostics.