Sandro Leidi

Sandro worked for the University of Reading, UK for over 20 years, providing consultancy and CPD courses for both the public and private sectors. During this time he accumulated a wealth of experience in many practical applications of statistical modelling.

He is an experienced user of both linear and generalised linear mixed models for complex designs such as cross-overs with repeated measures. In particular, he specialises in the analysis of discrete responses: counts, proportions and ordered categories (from surveys of customers' degree of satisfaction for example), as well as modelling of multivariate responses which are clustered but measure different attributes, hence require separate supporting distributions. He also conducts joint modelling of longitudinal (repeated measures) data and time-to-event outcomes.

Sandro also applies survival analysis (time-to-event) techniques, such as the Cox and Weibull models with frailties, and is experienced in using the counting process format for modelling with time-varying covariates.

He has an excellent knowledge of mainstream statistics packages such as Genstat, Minitab, R, SAS, SPSS and Stata, as well as specialised packages for power studies such as PASS. As for the Bayesian flavour of statistical modelling, he uses WinBUGS, OpenBUGS, JAGS from within R and the bayes prefix in Stata.

Notably, he is a power user of Excel, versed in the use of array functions.

For a personally tailored consultancy discussion, feel free to contact Sandro in English, French, Spanish, Italian or Catalan.

Latest publication:

Post your comment

Timberlake Consultants