I am using a published scale — do I still need to run EFA?

Usually not. If the scale is established and you are using it in a comparable population, its factor structure is already a hypothesis, and CFA is the test of that hypothesis. Run an EFA only when the context is genuinely new (a different culture, language or population) or when the published structure clearly does not fit your data.

Which model fit indices must I report for CFA?

Report chi-square with its df and p, plus at least one index from each family: an incremental index (CFI or TLI), an absolute index (RMSEA with its 90% confidence interval), and a residual index (SRMR). Most management journals also expect chi-square/df.

Can I run EFA and CFA on the same data?

No. A CFA on the data that generated the structure is circular, and it is the most common reason scale-development papers are rejected. Use two independent samples, or randomly split one large sample and say so explicitly in your Method section.

Is my data uploaded anywhere?

No. Every calculation runs inside your own browser using JavaScript. Nothing is sent to a server and nothing is stored.

Should I use CB-SEM or PLS-SEM?

CB-SEM if you are confirming an established theory with reflective multi-item constructs and model fit is part of your contribution. PLS-SEM if your focus is prediction or theory development, or you have formative constructs.

Free research tool · runs entirely in your browser

Which analysis should I run — and what do I report?

Load your Excel or CSV file and this page walks you through it, one step at a time: it checks your variables, recommends the analysis that fits your design, runs it, and drafts the Methods and Results sections for your paper.

Your data never leaves your device Excel & CSV No sign-up Free, always

Load your data

Excel (.xlsx), CSV, or paste straight from a spreadsheet. Nothing is uploaded.

Drop your file here, or browse

.xlsx · .csv · .tsv · .txt — up to about 20 MB

Processed on your device — no server involved.

Check your variables

Confirm what each column is. This is what decides which tests are valid — worth 30 seconds.

Prepare your data — optional

Reverse-score items, build construct scores, band a variable, or filter cases. Every step is logged for your Method section.

Choose your analysis

Let the advisor work it out from your design, or pick it yourself.

Results

Assumption checks first, then the tables and charts your paper needs.

Run an analysis in Step 4 and the results appear here.

Write-up & report

APA 7 text with your real numbers, Word-ready tables, reproducible code, and the whole thing as one download.

Run an analysis and your draft appears here.

What am I supposed to report?

The reference tables researchers most often go looking for — every threshold attributed to its source, so you can cite it rather than assert it.

Go further in the live labs

Full SEM, multilevel models and PLS need a real statistics engine. These launch in one click, with the code from Step 6 ready to paste in.

RStudio in your browser

The full IDE with lavaan, psych, semTools and the tidyverse. Best for CFA, SEM, multilevel models and PLS.

Launch RStudio →

Jupyter in your browser

Python with pandas, statsmodels, semopy and factor_analyzer. Best if your workflow is already Python.

Launch Jupyter →

Lab sessions are temporary and private — save your work before closing the tab.

Frequently asked questions

Keep exploring

P-value calculator Sample size calculator Live Analytics Lab Journal Finder Correlation vs regression Levels of measurement Free statistics books

A word of caution. This is a decision-support tool for researchers, not a replacement for your statistics faculty, your methodologist, or the primary literature in your field. Thresholds are widely used conventions, not laws — always check what your target journal and your committee actually expect. Every calculation runs in your browser and is validated against scipy, statsmodels and factor_analyzer, but the responsibility for what you report remains yours.