Research
While some of my research may be considered methodological, it has always had a strong applied component, since I consider the transfer of knowledge to Society of vital importance. In this sense, my main lines of research are applied, finding in them the statistical and econometric methods that I have been developing throughout my research activity.
Applied lines
Mixed lines. RWD and RWE
Methodological lines
The main methodological lines in which I have worked are those that I expose below. The statistical and econometric methods that I have developed in each of them are applied in the lines mentioned above.
Multivariate survival analysis. Survival analysis is made up of a set of methods that analyze the time that elapses until an event of interest occurs. However, there are situations in which subjects may experience various events during the observation period (for example, recurrences of a disease). In these cases, a different methodology is required from that used in the standard survival analysis, which allows solving its problems, caused by the existence of several events.
Spatio-temporal models. Spatio-temporal models try to explore, describe, visualize and analyze the data, considering their distribution characteristics both in space, which are usually expressed through the use of geographic coordinates, and in time. These models also consider the interaction between both dimensions, that is, the temporal variation of spatial dependence.
Mixed models. In many areas, data appears quite frequently in the form of a 'cluster', that is, grouped and related. In fact, the two basic situations that produce clustered data structures are: hierarchical random processes and repeated measures or longitudinal data. Mixed models intend to study and implement a set of methods to treat and/or analyze these clusters.
Bayesian Statistics. Bayesian statistics comprises a set of techniques in which probability is considered subjective, the parameters are random variables and Bayes' theorem is used to update the uncertainty, that is, to infer the probability that a hypothesis may be true, and, therefore for decision making.