Statistical Test for Differences

This app supports common tests of difference, structured data entry, guided test selection, basic assumption checks, plots, and exportable result summaries. It is designed for fast laboratory or classroom analysis, not as a replacement for full statistical review.

Supported Tests

Independent t-test: two independent groups under approximate normality.

Paired t-test: two related measurements from the same or matched units.

One-way ANOVA: three or more independent groups in a parametric design.

Two-way ANOVA: two independent factors in a parametric design.

Repeated Measures ANOVA: the same subjects measured across three or more conditions.

Mann-Whitney U: non-parametric alternative to the independent t-test.

Wilcoxon Signed-Rank: non-parametric alternative to the paired t-test.

Kruskal-Wallis: non-parametric alternative to one-way ANOVA.

Friedman Test: non-parametric alternative to repeated-measures ANOVA.

Results Sections

Test Statistics: the main inferential result, including statistic, degrees of freedom when applicable, p-value, and effect size when available.

Data Summary: descriptive statistics for each group.

Data Visualization: descriptive and box-and-whisker plots.

Assumptions: normality and homogeneity checks where appropriate.

Q-Q Plot: a visual normality check for the same scope used in the assumptions section.

Post Hoc Analysis: pairwise follow-up comparisons after significant omnibus tests where implemented.

Post Hoc Methods

One-way ANOVA: Tukey HSD or Tukey-Kramer for unequal group sizes.

Repeated Measures ANOVA: paired t-tests with Bonferroni adjustment.

Kruskal-Wallis: Dunn's test with Holm correction.

Two-way ANOVA: automated simple-effects post hoc testing is not implemented.

Important Notes

Normality: the app reports an approximate Shapiro-Wilk procedure based on simulation rather than a dedicated exact implementation.

Homogeneity: Levene's test is used for group variance homogeneity when relevant.

Interpretation: a small p-value suggests evidence against the null hypothesis, but scientific importance still depends on effect size, design quality, and context.

Exports: tables can be copied as TSV and plots or sections can be exported as PNG.

Citation Recommendation

If you use this web application in academic work:

Olarve, J. S. L. (2026). Statistical Test for Differences. i-Nano Research Facility, De La Salle University Manila. Retrieved from https://www.inanolab.com/statistics.html

Updated citation will be provided once DOI is assigned.

Disclaimer

This application is intended for instructional, screening, and routine laboratory use only. Although reasonable statistical procedures are implemented, the software does not replace expert statistical judgment, study-specific methodological review, or independent verification of results. Users remain responsible for confirming test assumptions, data quality, design appropriateness, and the scientific interpretation of all outputs before using them in reports, publications, theses, or formal decision-making.

Workflow