ABSTRACT
Introduction: Multilevel models have gained immense popularity across almost every discipline due to the presence of hierarchy in most data and phenomena. In this paper, we present a systematic review on the adoption and application of multilevel models and the important information reported on the results generated from the use of these models. Methods: The review was performed by searching Google Scholar for original research articles on the application of multilevel models published between 2010 and 2020. The search strategy involved topics such as "multilevel models," "hierarchical linear models," and "mixed models with hierarchy." The search placed more emphasis on the application of hierarchical models in any discipline but excluded software methodological development and related articles. Results: A total of 121 articles were initially obtained from the search results. However, 65 articles met the inclusion criteria for the review. Out of the 65 articles reviewed, 46.2% were related to health/epidemiology, 15.4% to education and psychology, and 16.9% to social life. The majority of the articles (78.5%) were two-level models, and most of these studies modelled univariate responses. However, the few that modelled more than one response modelled them separately. Moreover, 83.1% were cross-sectional design, and 9.2% and 6.2% were longitudinal and repeated measures, respectively. Moreover, a little over half (55.4%) of articles reported on the intraclass correlation measure, and all articles indicated the response variable distribution where most (47.7%) were normally distributed. Only 58.5% of articles reported on the estimation methods used as Bayesian (20%) and MLE (18.5%). Again, model validation measures and statistical software were reported in 70.8% and 90.8% articles, respectively. Conclusion: There is an increase in the utilization of multilevel modelling in the last decade, which could be attributed to the presence of clustered and hierarchically correlated data structures. There is a need for improvement in the area of measurement and reporting on the intraclass correlation, parameter estimation, and variable selection measures to further improve the quality of the application of multilevel models. The integration of spatial effects into multilevel models is very limited and needs to be explored in the future.
ABSTRACT
Discrete count time series data with an excessive number of zeros have warranted the development of zero-inflated time series models to incorporate the inflation of zeros and the overdispersion that comes with it. In this paper, we investigated the characteristics of the trend of daily count of COVID-19 deaths in Ghana using zero-inflated models. We envisaged that the trend of COVID-19 deaths per day in Ghana portrays a general increase from the onset of the pandemic in the country to about day 160 after which there is a general decrease onward. We fitted a zero-inflated Poisson autoregressive model and zero-inflated negative binomial autoregressive model to the data in the partial-likelihood framework. The zero-inflated negative binomial autoregressive model outperformed the zero-inflated Poisson autoregressive model. On the other hand, the dynamic zero-inflated Poisson autoregressive model performed better than the dynamic negative binomial autoregressive model. The predicted new death based on the zero-inflated negative binomial autoregressive model indicated that Ghana's COVID-19 death per day will rise sharply few days after 30th November 2020 and drastically fall just as in the observed data.
