Expand description
Generalizability theory for all-random, crossed person × item × rater designs.
Supply estimated variance components and project reliability to a new number of exchangeable items and raters. This release does not fit raw data.
use crolib::CrossedComponents;
let c = CrossedComponents { person: 12.0, item: 6.0, rater: 4.0,
person_item: 8.0, person_rater: 10.0, item_rater: 2.0, residual: 24.0 };
let result = c.project(4, 2)?;
assert!((result.generalizability - 6.0 / 11.0).abs() < 1e-14);Structs§
- Crossed
Components - Seven variance components for an all-random crossed design.
- Reliability
- Relative and absolute reliability for a proposed measurement design.
Enums§
- Projection
Error - Invalid or undefined D-study projection.