For optimal use, please visit numiqo on your desktop PC!

Structural Equation Modeling
Measured metric variables

Add numeric data to the table to select measured variables.

Measured ordinal variables

Set a column to ordinal in the data table to model it with thresholds.

Estimator
Missing values
Factor scaling
Means

How-to

Path Analysis Calculator

Path analysis example data

Here you can create and estimate a path analysis online. Path analysis represents a set of directional relationships between measured variables and estimates those regression paths simultaneously. It is a special case of structural equation modeling without a latent measurement model.

How to create a path model

  1. Enter or paste the data for all variables into the table.
  2. Select the measured variables that belong in the path model.
  3. Drag from a predictor to an outcome and set the connection to Regression.
  4. Add further paths to represent the relationships specified by your model.
  5. Add covariances between exogenous predictors when the model requires them.
  6. Choose the estimator and missing-value treatment, and then select Run model.

Arrow direction matters. A regression arrow points from the predictor to the outcome. A covariance is shown with arrows at both ends and describes an undirected association. The diagram expresses the model you ask the calculator to estimate; the direction of an arrow should therefore come from the study design and substantive theory, not from the observed correlations alone.

Path analysis results

The parameter table reports each path coefficient with its standard error, z statistic, p-value, 95% confidence interval, and standardized estimate. R-squared shows the proportion of variance explained for each outcome. When the model has degrees of freedom, global fit measures such as chi-square, CFI, TLI, RMSEA, and SRMR describe how well the complete path model reproduces the observed covariance structure.

Direct, indirect, and total effects

For a chain such as X to M to Y, first draw both regression arrows. Then choose Indirect effect and select the starting and ending variables. The calculator labels the relevant paths and adds the indirect effect. If a direct path from X to Y is present, it also reports the total effect. You can use delta-method confidence intervals or request 1,000 bootstrap samples.

If the mediator or another construct is measured by several indicators, use the SEM mediation analysis calculator to combine the path model with latent measurement models. For a general model that includes both latent and measured variables, return to the structural equation modeling calculator.

Cite numiqo: numiqo Team (2026). numiqo: Online Statistics Calculator. numiqo e.U. Graz, Austria. URL https://numiqo.com