Exploratory factor analysis (EFA) calculator
Exploratory factor analysis shows how many factors lie behind your items and which items go with each one. It is the most common test for construct validity in theses and research papers. Paste your data from Excel: one row per person and one column per item of the whole scale. If the first row has the question names, it is skipped automatically.
The calculation runs in your browser. We don’t store what you type.
How it’s calculated
The Pearson correlation matrix of the items and its eigenvalues are calculated. By default, as many factors are extracted as there are eigenvalues above 1 (Kaiser criterion), and you can change the number.
Extraction by minimum residuals (minres), which gives the same solution as iterated principal axis factoring. The initial communalities are the squared multiple correlations.
Varimax rotation with Kaiser normalization, which keeps the factors uncorrelated and makes it easier to see which items go with each one.
Each item’s communality (h²) is the sum of its squared loadings: the part of its variance the factors explain.
How to read the result
- Each item should load 0.40 or more on one factor (some authors accept 0.30) and less than 0.30 on the others.
- An item that loads high on two factors (cross-loading) or doesn’t reach 0.40 on any is a candidate to revise or remove.
- Ideally the factors match the dimensions you planned. If an item falls on another dimension, check its wording.
- In the social sciences, factors explaining 50% of the variance or more is considered adequate. Communalities below 0.30 point to items that share little with the rest.
Frequently asked questions
- What do I do before the factor analysis?
- Check the KMO and Bartlett’s test. If the KMO is below 0.60 or Bartlett’s test is not significant, the data are not suitable for factor analysis.
- How many factors do I extract?
- Start with the Kaiser criterion (eigenvalues above 1) and compare it with the dimensions in your theory. If your instrument has 3 dimensions, try 3 factors and see whether the items fall where they should.
- Why varimax and not oblimin?
- Varimax is the most used rotation in theses and the default in many textbooks. If your dimensions are strongly related, an oblique rotation (oblimin or promax) may fit better; in that case use SPSS, R or jamovi.
- Does it match SPSS?
- With the same method, yes. In SPSS choose Extraction > Unweighted least squares (equivalent to minres) or Principal axis factoring, and Rotation > Varimax. The third decimal can differ depending on how many iterations each program runs.
Know what you want to ask?
Build the survey in SurvIQ, share it by link, WhatsApp or QR code, and watch results come in live. The Free plan is free forever.
Create a free survey