Keywords: composite likelihood, effective sample size, REML, spatial dependence Composite likelihood methods have become popular in spatial statistics. This is mainly due to the fact that large matrices need to be inverted in full maximum likelihood and this becomes computationally expensive when you have a large number of regions under consideration. We introduce restricted pairwise composite likelihood (RECL) methods for estimation of mean and covariance parameters in a spatial Gaussian random field, without resorting back to the full likelihood. A simulation study was carried out to investigate how this method works in settings of increasing domain as well as infill asymptotics, whilst varying the strength of correlation, with similar scenarios as Curriero and Lele (1999). Preliminary results showed that pairwise composite likelihoods tend to underestimate the variance parameters, especially when there is high correlation, while RECL corrects for the underestimation. Therefore, RECL is recommended if interest is in both the mean and the variance parameters. The methods are made available in the spatialRECL package and implemented in R. The methodology will be highlighted in the first part of the presentation, and some analysis will be made on a real data example of TSH levels from Galicia, Spain. References Curriero, F, and S Lele. 1999. “A Composite Likelihood Approach to Semivariogram Estimation.” J Agric Biol Envir S 4 (1): 9–28.
