study_results.csv23 cols · 1,847 rows
01
Upload
Drop a CSV or Excel file. Decisionly profiles your columns, suggests variable types, and flags issues before you proceed.
DATAINSIGHTDECISION
Free during betaStatistical analysis,
Statistical analysis,
without a statistician.
Upload a dataset. Decisionly picks the right test from your data, runs it, and writes the result in plain language. Built for researchers, students, engineers, and analysts who want correct answers — not statistics homework.
study_results.csv23 cols · 1,847 rows
Group comparison·Welch’s t-testsignificant
Group B scores are higher than Group A,
with a large effect.
- p-value
- 0.0023
- effect (d)
- 0.84
- n
- 142, 138
distribution
→ selected by rule: scale × nominal, unequal variancePDF ready
A complete statistical analysis, in three guided steps.
Decisionly handles the work a careful analyst would do — column profiling, assumption checks, the right test, a written interpretation — in the order they should happen.
Compare
Relate
Predict
02
Pick a goal
Compare groups, find a relationship, or predict an outcome. The engine picks the test that fits your data.
decisionly.dev1 / 4
03
Get the report
Read a one-paragraph interpretation, view the key numbers, and download a clean multipage PDF.
A statistical engine that knows when to use what.
Tests are chosen from rules grounded in the data’s shape — sample size, normality, variance — not by whichever was clicked first.
20+ statistical tests built in
Classical hypothesis tests, non-parametric alternatives, and regression — chosen by data shape, sample size, and assumption checks. Post-hoc tests, effect sizes, and confidence intervals come automatic.
20+tests built in
- t-testWelch / Student
- ANOVA+ Tukey, Games-Howell post-hoc
- Pearson & Spearmancorrelation
- Linear & logisticregression
- Chi-square+ non-parametric alternatives
+ assumption checks+ effect sizes+ confidence intervals
Picks the right test
Variable types, sample sizes, normality (Shapiro / Anderson-Darling), and homogeneity (Levene) decide the test. The engine flags assumption violations — it doesn't paper over them.
example2 groups · scale outcome · normal
Plain-language interpretation
Every result comes with a paragraph written in plain English — what was found, how confident you can be, what it means in practice, and what to be careful about. P-values, effect sizes, and confidence intervals are translated into outcomes you can act on. No jargon-only statslish, no second-guessing what a number means for your decision.
Studentsusingthenewtutoringmethodscored12 pointshigheronthefinalexamthanthestandardmethod,onaverage.
Thedifferenceislarge(Cohen'sd = 0.84)andstatisticallyreliable(p < 0.001)—unlikelyduetochance.
Professional PDF export
Cited tests, effect sizes, assumption checks, and visualizations — formatted to share, attach to a paper, or hand to a reviewer. Multipage with table of contents.
Decisionly Report
study_results.pdfpage 1
page 2
page 3
page 4
page 1
page 2
page 3
page 4
Welch’s t · group comparison1 / 4
Downloaded
Run your first analysis in under two minutes.
No credit card. Sample data included.