Empirical methods for the estimation of the mixing probabilities for socially structured populations from a single survey sample.
ABSTRACT "The role of variability of sexual behavior in the transmission dynamics of HIV and AIDS has been illustrated, through the use of mathematical models, by several investigators.... In this paper we describe some practical methods for estimating the deviations from random mixing from a single survey sample.... We include a description of the role of the estimated mixing probabilities in models for the spread of HIV, a discussion of alternatives and possible extensions of the methods described in this article, and an outline of future directions of research." (SUMMARY IN FRE)
- SourceAvailable from: Benjamin R Morin[Show abstract] [Hide abstract]
ABSTRACT: The problem of who is mixing with whom is of great theoretical importance in the context of heterosexual mixing. In this article, we publish for the first time, data from a study carried out in 1989 that had the goal of estimating who is mixing with whom, in heterosexually active college populations in the presence of co-factors like drinking. The gathering of these data and the challenges involved in modelling the interaction between and among heterosexually active populations of individuals are highlighted in this manuscript. The modelling is based on the assumptions that at least two processes are involved: individual affinities or preferences determine 'what we want' while mixing patterns describe 'what we get'. We revisit past results on the role of affinity/preference on observed mixing patterns in one- and two-sex mixing populations. Some new results for homosexually active populations are presented. The study of mixing is but the means to an end and consequently, we also look at the role of affinity on epidemics as filtered by observed mixing patterns. It would not be surprising to observe that highly distinct preference or mixing structures may actually lead to quite similar epidemic patterns.Journal of Biological Dynamics 09/2010; 4(5):456-77.
- [Show abstract] [Hide abstract]
ABSTRACT: To study the spread of sexually transmitted diseases (STDs) using social/sexual mixing models, one must have quantitative information about sexual mixing. An unavoidable complication in gathering such information by survey is that members of the surveyed population will almost certainly have sexual contacts outside that population. The number of these outsiders may be substantial and, hence, important for the modelling process. In this paper, we develop a mark-recapture model for estimating the size of the population at risk for contracting a STD due to direct sexual contact with a specified population targeted by a survey. This mark-recapture methodology provides a reliable method of estimating the number of outsiders. Because not everyone in the targeted population may be sexually active, the size of the sexually active subset, used as the number marked in our tag-recapture formulation, must be estimated, which introduces extra variability. We derive an estimator of the variance of the estimated total number at risk that accounts for this extra variability and an expression for the bias of that estimator. We extend the methodology to stratified surveys and illustrate its use with data collected from a population of university undergraduates to estimate sexual mixing parameters of a deterministic model of the spread of STDs.Statistics in Medicine 12/1991; 11(12):1533 - 1549. · 2.04 Impact Factor