Description
Calculate an equal-size-model cluster design-effect term.
Statistical record calculations depend on risk-set boundaries, cluster assumptions, missingness codes and exposure definitions. Entry can be excluded while exit inclusion is a policy choice, and one design-effect formula assumes equal cluster sizes. A local flag or contribution does not fit a model or establish a clinical result.
Calculate an equal-size-model cluster design-effect term.
When to use this tool
Use Trial Cluster Design Effect Input when you already have meanClusterSize, intraclusterCorrelation and need to calculate an equal-size-model cluster design-effect term.
How to use the tool and how it works
- Enter meanClusterSize, intraclusterCorrelation. Use the units and conventions below.
- Run the tool. Correct any format, range, or cross-field input error before interpreting the output.
- Read designEffect. Use the output convention below.
Require meanClusterSize>=1 and intraclusterCorrelation in [0,1]; return 1+(meanSize-1)*correlation.1
Variation in cluster sizes can change design effects and trial power. This equal-size expression does not include that variation or determine a complete sample-size requirement.2
| Input field | Meaning and units | Accepted input |
|---|---|---|
| meanClusterSize | Mean Cluster Size | Required. Finite number; minimum 1. |
| intraclusterCorrelation | Intracluster Correlation | Required. Finite number; minimum 0; maximum 1. |
| Output field | Interpretation | Returned shape |
|---|---|---|
| designEffect | 1+(meanClusterSize-1)*intraclusterCorrelation under an equal-cluster-size model; no unequal-size correction. | Required. Finite number. |
{
"input": {
"meanClusterSize": 20,
"intraclusterCorrelation": 0.05
},
"output": {
"designEffect": 1.9500000000000002
}
}Limitations and assumptions
- Mean size may be fractional. This simplified model omits cluster-size variability and does not estimate correlation, sample size, power, analysis validity or appropriate trial design.
- Inputs must satisfy both the field types and the cross-field conditions above. Invalid input causes an error; a successful false flag or documented placeholder is a different result.
- Numeric calculations use binary floating point unless exact decimal-integer arithmetic is explicitly stated. No hidden measurement tolerance or decimal rounding is added; nonfinite or out-of-contract results are rejected. Numeric signed-zero results are exported as zero.
Alternative or Complementary methods
Use the complete study design, data and statistical model for inference. Check censoring, clustering, visit windows, exposure and missingness conventions before combining these terms, and assess uncertainty with the full analysis.
References
-
Calculation and input contract: biostatistics/trial-cluster-design-effect-input — tool.io
-
Sample size for cluster randomized trials: effect of coefficient of variation of cluster size and analysis method (2006) — Sandra M. Eldridge, Deborah Ashby and Sally Kerry
Similar or alternative tools
- Matched Set Case Control Count
Calculate supplied matched-set size and controls-per-case ratio.
- Follow Up Visit Window Offset
Check a supplied visit day within asymmetric inclusive allowances.