My research

I am a researcher in statistical inference, with interests in cognitive science and philosophy of science, and with a background in quantum and statistical physics.

Since I am a researcher in IMT-Lucca, I am pursuing novel research lines consisting of the development of statistical inference and information-theory grounded data analysis methods (like statistical validation of clustering algorithms and null models of complex networks), with applcations to the social sciences and especially to the analysis of psychometric data.

IAS Fellowship

Psychometrics and sociological survey analysis are arguably in need of novel interpretation paradigms to overcome their current limitations regarding reductionism and circularity. Such paradigms require novel methodological tools that can assess the rigidity and descriptive power of pre-defined constructs, and disentangle the conditional dependencies between item responses from the semantic or logical constraints inherent in the questions themselves.

Together with my colleague Clélia de Mulatier (UvA), we propose principled, construct-agnostic methods grounded in information theory to select the most informative questionnaire items. These novel criteria for item informativeness go beyond classic approaches that rely solely on internal consistency. Instead, they strike a balance between randomness and redundancy, ideally excluding both uncorrelated items (which cannot be reduced to latent variables) and redundant items (which replicate information already captured by others). We address this optimization problem using methods at the intersection of information theory, statistical inference, and statistical physics.

Furthermore, this fellowship will leverage the interdisciplinary environment of the IAS and the expertise of its research groups in psychological methods. The goal is to explore the broader potential and practical applications of these novel methodologies and paradigms across psychology and the social sciences.