Table 4.1.
Summary of applications and extensions using weighted pseudo-likelihood approach
| Author (s) | Data application | Survey type | Statistical method | |
|---|---|---|---|---|
| Sexual health | Bastos et al. (2018) * | Prevalence of HIV, Hepatitis C, Hepatitis B, syphilis among transgender women in Brazil | Divas Research: Respodent-driven sampling in 12 Brazilian cities | Simple logistic regression model with weights formed from degree in respondent-driven sample |
| Kunihama et al. (2016) | Total number of sexual partners among adolescents in the US | National Longitudinal Study of Adolescent Health: stratified sampling design | Dirichlet process mixture model with a rounded kernel method for latent continuous variables | |
| Kunihama et al. (2019) ** | Trajectories of variables for adolescent sexual development in the US | National Longitudinal Study of Adolescent Health: stratified sampling design | Trajectories of associations between mixed scale longitudinal responses | |
| Envir health | Vedensky et al. (2022) | Lead concentrations in moss in Galicia, Spain | Heavy metal biomonitoring data from Diggle et al. (2010): preferential, lattice sampling | Posterior mean predicted surfaces for geostatistical data |
| Physical and mental health | Alivertl & Russo (2022)* | Evolution of behaviors in compliance with COVID-19 preventive measures during lockdown in Italy | Imperial College London YouGov Covid 19 Behaviour Tracker Data Hub: repeated cross-sectional sampling | Dynamic latent-class regression model for longitudinal multivariate categorical data |
| Parker et al. (2022) ** | Health insurance rates by county in the US | American Commuity Survey: modified to informative sampling | Small area estimation of binary and count data, accounting for spatial dependence | |
| Parker & Holan (2022) ** | Association between physical activity and five-year mortality in the US | National Health and Nutrition Examination Survey: stratified, multistage design | Functional regression model with binomial or multinomial outcomes | |
| Trendtel & Robltzsch (2021) ** | Influence of test item position on test performance among students in New | Programme for International Student Assessment: balanced incomplete block | Probit mixed effects regression model. Balanced repeated replication for variance | |
| Williams & Savitksy (2021) ** | Association between smoking and depression for adults in the US | National Survey on Drug Use and Health: stratified, multistage design | Simple logistic regression model | |
| Poverty & housing | Fourrier-Nicolai & Lubrano (2020) | Child poverty incidence, intensity, and inequality in Germany | German Socio-Economic Panel: stratified sample with weights | Zero-inflated log-normal mixture model estimating the Three I’s of Poverty (TIP) curve |
| Gunawan et al. (2020) | Distribution of household disposable income in Australia | Household Income and Labour Dynamics in Australia: multistage sampling design | Rnite gamma mixture model | |
| Employement | Savitsky & Srivastava (2018) | Factors influencing number of employees at business establishments in California | Current Employment Statistics: informative, stratified sampling | Count regression model for large-scale data with unit missingness. Involves sample splitting |
| Savitsky & Toth (2016) * | Factors influencing number of job hires and separations in the US | Job Openings and Labor Turnover Survey: informative sampling | Multivariate count regression model with over-dispersion | |
| Savitsky et al. (2022) | Preserving differential privacy guards for family income in the US | Current Employment Statistics: informative, stratified sampling | Generation of synthetic databases that downweight high-risk records | |
| Sun et al. (2022) ** | Prevalence estimates for state-level employment and housing outcomes in the US | Household Pulse Survey: complex, multistage design | Small area estimation of binomial or multinomial outcome. Spatially correlated random effects |
*:
Software referenced.
**:
Software referenced and code available