ClassicalTestTheory.jl is a Julia package for data analysis using Classical Test Theory.

] add https://github.com/p-gw/ClassicalTestTheory.jl.git

ClassicalTestTheory.jl provides two entry points to doing data analsis. The input data can either be a numeric Matrix or a PsychometricTest. While Matrix methods provide full functionality, PsychometricTest methods provide some additional convenience such as scale analysis. For details on how to use ClassicalTestTheory.jl with PsychometricTest see XXX.

Consider some input data x,

julia> n_persons = 100;
julia> n_items = 8;

julia> x = rand(0:100, n_persons, n_items);

we can get some descriptive analysis of the items,

itemanalysis(x)

or estimate the internal consistency (e.g. using Cronbach's alpha)

reliability(x, Alpha())

The package will automatically calculate the coefficient from the data and construct appropriate confidence intervals.

To get multiple estimates of reliability just pass a vector of methods:

coefficients = [Alpha(), GLB(), Mu(2)]
reliability(x, coefficients)