Application of Hierarchical/Multilevel Models and Quality of Reporting (2010-2020): A Systematic Review.

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.

Examining the risk factors for overweight and obesity among women in Ghana: A multilevel perspective.

Overweight/obesity prevalence is on the increase in epidemic proportions across Low- and Middle-Income countries (LMICs). The public health burden associated with obesity/overweight cannot be underestimated due to its association with chronic health outcomes. This study investigated the individual- and community level risk factors for obesity/overweight among reproductive women. The data used consist of 4393 reproductive women and form part of the 2014 Ghana Demographic and Health Survey (GDHS). Information on these women are clustered within 427 communities. A 2-tier random intercept multilevel logistic model was used to assess the effect of individual- and community level factors on the likelihood of a woman to be obese/overweight. The obesity/overweight prevalence among reproductive women was estimated to be 35.5% (95% CI: 34.04, 36.90%), which significantly differed across clusters. Most at risk were women from middle-income households (aOR = 2.85; 95% CI: 2.28, 3.56), upper-income households (aOR = 5.019, 95% CI: 3.85, 6.55), attaining secondary education (aOR = 1.74; 95% CI: 1.41, 2.16), and higher (aOR = 1.63; 95% CI: 1.14, 2.33), old age (20-29 years (aOR = 4.26; 95% CI: 3.142,5.78); 30-39 years (aOR = 8.59; 95% CI: 6.15, 12.00); 40-49 years (aOR = 12.81; 95% CI: 9.10, 18.16)). Significant differences in the probability of being overweight/obese between different communities were observed (MOR = 1.39). The high prevalence of overweight/obesity requires urgent public health interventions to prevent future public health crises. Efforts to strengthen the healthcare system, encourage lifestyle modification, and public health education are needed to solidify the gains of ensuring a healthy population by 2030 (SDG 3).

Structural equation modelling of COVID-19 knowledge and attitude as determinants of preventive practices among university students in Ghana.

Coronavirus disease (COVID-19) has distorted the economic development activities of many countries across continents. This undesirable tragedy has highly affected the educational system, which majorly contributes to the wellbeing of an individual and the economy as a whole. The study aims to explore the determinants of COVID-19 preventive practices among students considering their knowledge about COVID-19 and attitudes toward the disease. The data for the study were collected through an online questionnaire survey involving university students. The relationship between students' knowledge, attitude and their preventive practices towards COVID-19 were investigated using structural equation modelling. The results indicated that most students demonstrated substantial knowledge on COVID-19, moderate to strongly agree attitude towards COVID-19, and sometimes practiced COVID-19 preventive and safety protocols. In addition, a positive relationship between knowledge and attitudes towards COVID-19 was established. Also, a positive effect was established for students' knowledge about COVID-19 and preventive practices, whilst an adverse effect was confirmed for attitudes towards COVID-19 and practices to avoid spreading the COVID-19 disease.

The relationship between driver and passenger's seatbelt use: a bivariate probit analysis.

Vehicle seatbelt has been shown to have a beneficial impact on occupants. However, some occupants do not use the seatbelt when inside a moving vehicle. Despite the numerous investigation on the risk factors associated with seatbelt use by occupants, little is known about the relationship between driver and passenger seatbelt use. This gap is analyzed with road side observational survey data on driver and front-right seat passenger's seatbelt use behaviour using bivariate probit model. The use of the bivariate probit model is based on the premise that the front-right passenger's seatbelt use is endogenously related to that of the driver. Out of the 5,433 vehicles observed, the prevalence rate of driver and front-right passengers' seatbelt use were 81% and 33%, respectively. In addition, there is a positive relationship between driver and passenger's seatbelt use with correlation coefficient of 0.53. Thus, the unobserved factors that influence the probability of a driver seatbelt use also influence their front-right passenger's seatbelt use propensity.

The effect of road and environmental characteristics on pedestrian hit-and-run accidents in Ghana.

The number of pedestrians who have died as a result of being hit by vehicles has increased in recent years, in addition to vehicle passenger deaths. Many pedestrians who were involved in road traffic accident died as a result of the driver leaving the pedestrian who was struck unattended at the scene of the accident. This paper seeks to determine the effect of road and environmental characteristics on pedestrian hit-and-run accidents in Ghana. Using pedestrian accident data extracted from the National Road Traffic Accident Database at the Building and Road Research Institute (BRRI) of the Council for Scientific and Industrial Research (CSIR), Ghana, a binary logit model was employed in the analysis. The results from the estimated model indicate that fatal accidents, unclear weather, nighttime conditions, and straight and flat road sections without medians and junctions significantly increase the likelihood that the vehicle driver will leave the scene after hitting a pedestrian. Thus, integrating median separation and speed humps into road design and construction and installing street lights will help to curb the problem of pedestrian hit-and-run accidents in Ghana.