Partial Eta Squared Calculator
Calculates partial eta squared for an effect relative to that effect’s error term. The form displays SS_effect/(SS_effect+SS_error) beside partial eta squared, using a worked condition that can be recalculated with the labeled inputs.
Set the labeled inputs
Partial Eta Squared
Scope of the partial eta squared method
The partial eta squared page calculates partial eta squared for an effect relative to that effect’s error term.
Partial Eta Squared is limited to the statistical quantity named by the result panel. The partial eta squared calculation does not silently add a population, time horizon, causal direction, or decision threshold that is absent from the fields.
Inputs that define partial eta squared
- Effect sum of squares: For partial eta squared, the worked value for effect sum of squares is 60 squared units. Treat the effect sum of squares entry (60 squared units) explicitly as a count, proportion, rate, estimate, or model parameter before comparing partial eta squared conditions. The form enforces minimum 1e-06.
- Error sum of squares: For partial eta squared, the worked value for error sum of squares is 240 squared units. Treat the error sum of squares entry (240 squared units) explicitly as a count, proportion, rate, estimate, or model parameter before comparing partial eta squared conditions. The form enforces minimum 1e-06.
The entries used for partial eta squared must refer to one coherent analysis condition. Combining incompatible populations, periods, or measurement definitions can produce valid partial eta squared arithmetic for a nonexistent study.
From inputs to partial eta squared
For partial eta squared, match every symbol in the relationship to a labeled field before substituting numbers. Partial Eta Squared is reported in the scale implied by the inputs and formula.
While checking partial eta squared, inspect every denominator in SS_effect/(SS_effect+SS_error). For partial eta squared, a zero or near-zero denominator can make partial eta squared undefined or unstable.
When the analysis changes, compare omega squared.
A reproducible partial eta squared case
The default partial eta squared condition is Effect sum of squares = 60 squared units, Error sum of squares = 240 squared units.
Effect SS 60 and error SS 240 give partial eta squared .20.
The live calculator reports Partial eta squared 0.2. Repeating one intermediate step from SS_effect/(SS_effect+SS_error) provides a fixed partial eta squared reference check for later code changes.
Statistical context for partial eta squared
Partial eta squared is not interchangeable with eta squared when several effects share a model.
For partial eta squared, power and design calculations are prospective scenarios, not guarantees. For partial eta squared, their answer changes when the planned effect, variance, allocation, alpha, or attrition assumption changes.
How to interpret the partial eta squared output
When interpreting partial eta squared, report the design inputs as assumptions and compare at least one plausible alternative before committing resources to the plan.
As a second check for partial eta squared, reversing the numerator and denominator answers a different question, so retain the direction printed in SS_effect/(SS_effect+SS_error).
Varying a single partial eta squared input at a time
Change effect sum of squares while holding the remaining entries fixed, then state why the direction and size of the partial eta squared change are plausible from SS_effect/(SS_effect+SS_error).
Repeat the partial eta squared exercise with error sum of squares. If a modest defensible change materially alters the interpretation, report both conditions rather than presenting that partial eta squared scenario as exact.
Documenting the partial eta squared result
Report partial eta squared using SS_effect/(SS_effect+SS_error), followed by the entered values, units, exclusions, and analysis date. Name the partial eta squared population or dataset boundary instead of leaving it implicit.
Keep the full calculator output with the record, including Partial eta squared 0.2. A later partial eta squared review can then distinguish a changed input from a different convention or software implementation.
Questions about partial eta squared
What does partial eta squared represent on this page?
It is the quantity produced by SS_effect/(SS_effect+SS_error) from the displayed effect sum of squares, error sum of squares. This page calculates partial eta squared for an effect relative to that effect’s error term.
What should be saved with partial eta squared?
Save the entered values and units for effect sum of squares, error sum of squares, along with the analysis date, exclusions, software or formula version, and the relationship SS_effect/(SS_effect+SS_error). That record is sufficient to rebuild this specific partial eta squared calculation.
Does partial eta squared establish a causal or population conclusion?
No. The displayed partial eta squared value is conditional on the entered data and named method. The partial eta squared design, measurement process, and assumptions determine what can be concluded beyond those values.
How should partial eta squared be rounded?
Keep the unrounded partial eta squared for subsequent arithmetic, then report only the precision supported by the source measurements and the decision context. Extra digits in partial eta squared do not correct sampling or model error.
Which input deserves the closest boundary check?
For partial eta squared, start with error sum of squares and then effect sum of squares. Confirm the partial eta squared units and allowed domain because a valid-looking entry can still describe the wrong statistical setup.
Why could another program report a different partial eta squared?
A different convention for rounding, tails, ties, interpolation, parameterization, or missing values can change partial eta squared. Compare the printed partial eta squared formula and its input definitions before treating either output as wrong.