Features/Analysis Recommendation
Analysis Recommendation

Not sure which analysis? Start here.

Upload your data, describe your research question, and the AI recommends the right method from a catalog of 60+ analyses — then hands you off with everything pre-filled.

Upload & Auto-detectAI recommends methodsVariables pre-mappedChat about results
Analysis Recommendation — chat with AI to find the right method
60+
Methods in Catalog
AI
Powered by Gemini
0
Manual Setup
Q&A
Result Chat

01 · UPLOAD

Start with your data.

Drop your file. AI reads the structure.

Upload a CSV or Excel file and the AI immediately scans column names, data types, and sample rows — no configuration needed. It understands what kind of data you have before you say anything.

Supported Formats
CSVExcel (.xlsx)JSON
Auto-detection
Column typesSample rowsVariable rolesData structure

Analysis Recommendation

Drop your data file here

CSV · Excel · JSON

survey_data.csv · 847 rows detected

ColumnTypeRole
ageNumericVariable
genderCategoricalGroup var
satisfactionNumericDependent
departmentCategoricalGroup var
tenure_yearsNumericVariable

02 · RECOMMEND

Describe your goal. Get the right method.

AI picks from 60+ methods — matched to your data.

Tell the AI what you're trying to find out. It scans the full analysis catalog and recommends every method that fits — based on your variable types, sample size, and research question. No wrong tests.

Recommendation basis
Variable typesSample sizeResearch questionData structure
Catalog coverage
Statistical methodsML modelsCustomer analyticsOR models

AI · Analysis Recommendation

I can see your dataset has 847 rows, 2 categorical group variables (gender, department), and a numeric outcome (satisfaction). What are you trying to find out?

Does job satisfaction differ by department?

Based on your data (1 categorical group, 1 numeric outcome, n=847), here are the best fits:

One-Way ANOVABest fit

4 departments · continuous outcome · n>30

Kruskal-WallisIf non-normal

Non-parametric alternative to ANOVA

Welch's ANOVAUnequal variance

Robust when group sizes differ

03 · JUMP IN

Pick a method. Run it immediately.

One click from recommendation to result.

Select any recommended method and jump straight into the analysis — variables are pre-mapped, settings are pre-filled. No setup from scratch. The AI hands you off to the right lab with everything ready.

Pre-filled setup
Variable mappingParameter defaultsAssumption checks
Connected labs
Statistical LabModel LabCustomer Insight LabOR Lab

One-Way ANOVA · Pre-filled Setup

Ready to run

Variables · Auto-mapped from your data

Group Variable

department

Categorical

Dependent Variable

satisfaction

Numeric

Assumption Checks

Normality (Shapiro-Wilk)✓ Pass
Homogeneity (Levene's test)✓ Pass
Sample size (n ≥ 30 / group)✓ Pass

04 · CHAT ABOUT RESULTS

Run the analysis. Then ask questions.

AI explains your results in plain language.

After any analysis runs, open the AI chat and ask anything — "what does this p-value mean?", "why is this coefficient negative?", "how do I report this in APA?". The AI has full context of your results and answers in plain language.

Result Q&A
Interpret coefficientsExplain p-valuesEffect size guidanceAPA reporting
Context-aware
Full analysis contextYour data valuesConversation historyFollow-up questions

Chat about ANOVA results

F(3,843) = 12.4, p < .001

📊 One-Way ANOVA result loaded as context

What does F(3,843) = 12.4 mean?

F(3, 843) = 12.4 means the ANOVA compared 3 degrees of freedom between groups (your 4 departments) against 843 within-group df (847 rows minus 4 groups). The ratio of 12.4 is large — it says group differences are about 12× bigger than random variation within groups.

How do I report this in APA format?

APA 7th edition format for your result:

A one-way ANOVA revealed a significant effect of department on job satisfaction, F(3, 843) = 12.41, p < .001, η² = .042.

Start with a question, not a method.

Upload your data, describe what you want to find out, and let the AI do the rest.

No credit card required
60+ methods in catalog