The PRO-Online Project
The PRO-Online project aims at proposing to students or researchers a way to perform psychometrical analyses of their own data, in a user-friendly manner, without complex handling of the data.
A wide variety of psychometric analyses, based either on models from classical test theory, Rasch measurement theory or non-parametric item response theory, can be performed directly through a web browser.
The various modules of PRO-online are provided with manuals explaining the underlying methods in detail, and figures and examples to guide interpretation of the results.
Item selection features programs for Mokken scale procedure (MSP), clustering of variables around latent components (CLV) and hierarchical clusters analysis of variables (HCAvar). Validity and reliability of a multidimensional measurement scale using CTT methods can be assessed with Validscale.
Parameters estimation of a Rasch model for dichotomous items and fit tests are performed in the Raschtest module. Marginal and conditional maximum likelihood estimations are implemented.The PCM module performs parameters estimation of a Partial Credit or a Rating scale Model with marginal maximum likelihood estimation for cross-sectional or longitudinal data.
Item responses from a dichotomous (Rasch, OPLM, Birnbaum, 3PLM, 4PLM, 5PAM) or polytomous (PCM, RSM) IRT or Rasch model can be simulated. Longitudinal or multigroup data simulation are also handled. The SIMIRT module also provides item characteristic curves.
The power of studies comparing the latent variable mean between two groups of patients or two visits can be estimated using models from the Rasch family (Raschpower).
Lack of measurement invariance (DIF, response shift) can be assessed using SEM (Rsoort) or Rasch models (ROSALI).
The different programs constitute an extension of the Anaqol project (http://www.anaqol.org) that includes SAS macro-programs or Stata modules. These developments are supported by the team of psychometrics of the SPHERE unit (https://www.sphere-inserm.fr). This unit is part of the INSERM, of the Nantes University and of the University of Tours (France).
