Multilevel Regression Approach Research Articles

The connection between social capital and Chinese public health: an analysis of six Chinese general social surveys

This study examines the connection between social capital and the perceived health status for the Chinese public. Extending from previous research that uses unrepresentative samples, the current project analyzes nationally representative samples of approximately 60,000 respondents. The data are drawn from six waves of the Chinese General Social Survey conducted in 2010, 2012, 2013, 2015, 2017, and 2018, respectively. The multilevel regression approach is used for statistical estimation. Findings show a consistent pattern that social capital, encompassing social network and generalized trust, is positively related to one’s perceived health status across these years. In addition, the moderating impact of personal capital on the link between social capital and perceived health status is identified. Social capital has a greater health impact on those with lower personal capital than those with higher personal capital. This study complements current literature and has implications for public health policies.

From policies to outcomes: multi-level analysis of the influence of regional planning on land take

ABSTRACT Urban expansion poses significant challenges to sustainability and undermines efforts for efficient spatial development. This study examines the relationship between regional planning regulations, interest groups and contextual factors affecting residential land take in Germany. The study employs a multi-level regression approach, analysing 4032 municipalities nested in 91 planning regions and utilizes a nationwide dataset on changes in residential and mixed land use. The overall goal of the study is to examine the effect of regional planning regulations and local interest groups, within varying contextual factors, on residential land take. Results show a dampening effect of regional regulatory intensity on residential land take at the local level. However, the effect becomes non-significant when considering local and regional covariates, revealing that regional planning alone may not be sufficient to reduce residential land take effectively. Findings underscore the complexity of regional planning and the importance of understanding its impacts on the local and regional level. As this study focuses on residential land take, further research is required to analyse the effectiveness of regional planning in managing land uses like trade, industry and infrastructure, concentrating urban development in central places, promoting transit-oriented development and supporting higher residential densities.

Seroprevalence of human herpes viruses in France, 2018–2022: a multilevel regression and poststratification approach

Background Information related to herpes simplex virus 1 and 2 (HSV-1 and 2), varicella-zoster virus (VZV), Epstein-Barr virus (EBV), and cytomegalovirus (CMV) seroprevalence in France is either lacking, incomplete, or outdated, despite their public health burden. Method We used routinely collected serological data between 2018 and 2022 to estimate HSV-1, HSV-2, VZV, EBV, and CMV seroprevalence in France. To account for demographic differences between our analytic samples and the French population and get estimates for sparsely sampled districts and age classes, we used a multilevel regression and poststratification approach combined with Bayesian model averaging via stacking weights. Results The observed seroprevalence (number of positive tests/number of tests) were 64.6% (93,294/144,424), 16.9% (24,316/144,159), 93.0% (141,419/152,084), 83.4% (63,199/75, 781), and 49.0% (23,276/47,525), respectively, for HSV-1, HSV-2, VZV, EBV, and CMV. Between 2018 and 2022, France had a model-based average (equal-tailed interval at 95%) expected seroprevalence equal to 61.1% (60.7,61.5), 14.5% (14.2,14.81), 89.5% (89.3,89.8), 85.6% (85.2,86.0), and 50.5% (49.3,51.7), respectively, for HSV-1, HSV-2, VZV, EBV, and CMV infections. We found an almost certain lower expected seroprevalence in Metropolitan France than in overseas territories for all viruses but VZV, for which it was almost certainly greater. The expected seroprevalences were likely greater among females for all viruses. Limitations Our results relied on the assumption that individuals were sampled at random conditionally to variables used to build the poststratification table. Implications The analysis highlights spatial and demographic patterns in seroprevalence that should be considered for designing tailored public health policies.

Demand Deposit Contracts and Probability of Bank Runs in Nigeria

The problem of bank runs is a global phenomenon and often damaging to the banking systems both in developed and developing countries. Monetary authorities always find it very difficult to resolve such a crisis; hence, it became imperative to examine this scenario in the case of Nigeria. Therefore, we examined the effect of demand deposit contracts on the probability of bank runs in Nigeria. This is considered a major contribution to the literature on finance from the perspective of an emerging market economy (EME). In this study, a multilevel Tobit regression approach was employed. This study identified statistically significant and positive effects of liquid assets (total assets) and total loans (total deposits) on bank runs in Nigeria. Based on the results, the study concluded that bank-specific factors have an effect on bank runs, and from these variables, liquid assets/total assets and total loans/total deposits have significant effects on bank runs. As a sequel to the findings and conclusion of this study, it was recommended that to avoid such a banking crisis, there is a need for deposit money banks in Nigeria to maintain adequate liquid assets and ameliorate the high level of deterioration in the quality of risk assets as well as the cost of funds.

Lack of leisure time physical activity and variations in cardiovascular mortality across US communities: a comprehensive county-level analysis (2011–2019)

ObjectivesTo investigate the associations between county-level proportions of adults not engaging in leisure-time physical activity (no LTPA) and age-adjusted cardiovascular mortality (AACVM) rates in the overall US population and across...

Are Law Schools Cream-Skimming to Bolster Their Bar Exam Pass Rates? A Multilevel Regression Approach to Estimate How Attrition and Transfer Rates Affect Bar Passage

Law schools are held accountable on many fronts to achieve and maintain high bar passage rates. While the course of legal education itself, along with various interventions, is a key driver of bar exam performance, Bahadur et al. (2021) suggests that obscure institutional practices might be inflating institutional bar passage performance. Such practices could include recruitment and admission of transfer students and academic attrition. We examine this hypothesis to assess the influence of both attrition and transfer on law schools’ bar passage differential using fixed-effects and between-within models. We also utilize Poisson regression to explore the effect of geographical proximity to other law schools on transfer rates. We find that, on average, neither attrition nor transfer activity substantively affect bar passage performance, and that although geographic proximity to other schools is related to transfer rates, this relationship does not result in notable differences in bar passage.

A novel Bayesian multilevel regression approach to the reconstruction of an eastern Mediterranean temperature record for the last 10,000 years

Climate reconstructions derived from proxy records for individual sites often fail to incorporate existing regional information, which may help to determine uncertainties and express variability within specific reconstructions or commonalities between datasets. Such an understanding is crucial when examining past human-environment interactions. Taking the eastern Mediterranean as our case study, we present here a new air temperature model for the last 10,000 years that utilises data from 33 previously published proxy-based independent reconstructions, using a novel fully Bayesian approach that applies multilevel regression models of individual temperature datasets grouped together into 300 year long consecutive sub-intervals. A Bayesian multilevel approach allows the model to share information between each regression model from the individual datasets and the 300 year grouped regression models. The results demonstrate commonalities between individual datasets derived from different sources, and embed the uncertainties within the model. Our results establish that the eastern Mediterranean region was consistently warmer than the 20th century, except for two short intervals at the end of the Early Holocene (between 8400 and 8250 yrs BP) and the start of the mid-Holocene (between 7800 and 7650 yrs BP). We also identify changes within temperature associated with both the 8.2 ka and 4.2 ka BP events, however our findings identify regional warming in the eastern Mediterranean, rather than cooling often associated with the 4.2 ka BP event. Our results are comparable with previous large scale hemispheric reconstructions, demonstrating that our model is a robust candidate for temperature reconstructions within a confined region, which can range from mesoscale up to macroscale.

Bottom-up innovation adoption of green public procurement in the United States

ABSTRACTClimate change and its outcomes are complex issues facing communities worldwide. Among policy options to address such issues, existing scholarship focuses mainly on regulatory policies. An innovative market-based policy option is green public procurement (GPP), which may induce private companies to voluntarily switch to more environmentally friendly products. The study examines the adoption and diffusion of GPP practices among local governments under the influence of state governments which may have distinct environmental policies and efforts. By drawing on data from an original national survey and supplemental sources and by employing a multilevel regression approach, this paper highlights GPP as an innovative, yet often overlooked, environmental policy option and strategic management tool. Results suggest a bottom-up, grassroots GPP adoption mechanism, emphasising local initiatives, strategic planning, and vision in sustainability advancement.

Prior Choice for the Variance Parameter in the Multilevel Regression and Poststratification Approach for Highly Selective Data. A Monte Carlo Simulation Study.

The multilevel and poststratification approach is commonly used to draw valid inference from (non-probabilistic) surveys. This Bayesian approach includes varying regression coefficients for which prior distributions of their variance parameter must be specified. The choice of the distribution is far from being trivial and many contradicting recommendations exist in the literature. The prior choice may be even more challenging when data results from a highly selective inclusion mechanism, such as applied by volunteer panels. We conduct a Monte Carlo simulation study to evaluate the effect of different distribution choices on bias in the estimation of a proportion based on a sample that is subject to a highly selective inclusion mechanism.

My Voters Should See This! What News Items Are Shared by Politicians on Facebook?

Political actors play an increasingly important role in the dissemination of political information on social media. However, relatively little is known about the mechanisms why specific news items are shared with the support base instead of others. For a timespan between December 2017 and the end of 2018, we combine the analysis of Facebook content from 1,022 politicians associated with 20 political parties from Germany, Spain, and the UK, with an automated content analysis of media coverage from 22 major online news outlets, and survey data in a multilevel binomial regression approach. By comparing news items that have been shared by one or several political parties with news items that have not been shared by any of them, we overcome the selection biases of previous studies in the field of news dissemination. Findings show that a news item's likelihood to be shared by a politician increases (1) if that politician's party is mentioned in the news item, (2) the more salient their party's owned issues are in the news item, and (3) the more party supporters tend to read the news outlet in which the news item is published. We contextualize these findings in light of political actors’ multi-faceted motivations for news sharing on social media and discuss how this process potentially reinforces an information bias that may contribute to the polarization and fragmentation of audiences.

Socio-economic inequalities in health service utilization among Chinese rural migrant workers with New Cooperative Medical Scheme: a multilevel regression approach

BackgroundWhile reducing inequity in health service utilization is an important goal of China’s health system, it has been widely acknowledged that a huge number of rural migrant workers cannot be effectually protected against risks with the New Rural Cooperative Medical Insurance (NCMS).MethodData of the 2016 China Labor-force Dynamic Survey and the Chinese Urban Statistical Yearbook were used. The multilevel regression approach was implemented with a nationally representative sample of rural migrant workers with NCMS. Our study adopted the concentration index and its decomposition method to quantify the inequality of their health service utilization.ResultThe multilevel model analysis indicated that impact variables for health service utilization were not concentrated, especially the contextual and individual characteristics. The concentration indices of the probability of two weeks outpatient and the probability of inpatient were -0.168 (95%CI:-0.236,-0.092) and -0.072 (95%CI:-1.085,-0.060), respectively. The horizontal inequality indices for the probability of two-week outpatient and the probability of inpatient were -0.012 and 0.053, respectively.ConclusionThe health service utilization of rural migrant workers with NCMS is insufficient. Our study highlighted that substantial inequalities in their health service utilization did exist. In addition, their need of health service utilization increased the pro-poor inequality. Based on the findings, our study offered notable implications on compensation policies and benefit packages to improve the equality among rural migrant workers with NCMS.

Do stop-signs improve the safety for all road users? A before-after study of stop-controlled intersections using video-based trajectories and surrogate measures of safety

Converting minor-approach-only stop (MAS) intersections to all-way-stop (AWS) intersections is a prevailing safety countermeasure in North American urban areas. Although the general population positively perceives the installation of stop-signs in residential areas, little research has investigated the impact of AWS on road safety and road user behaviour. This paper investigated the safety effectiveness of converting MAS to AWS intersections using an observational before and after approach and surrogate measures of safety. More specifically, the safety impacts of AWS conversion were investigated using multiple indicators, including vehicle speed measures, vehicle–pedestrian, vehicle-cyclist, vehicles-vehicle interactions as well as yielding rates before and after the treatment implementation. A multi-level regression approach was adopted to determine the effect of stop signs controlling for built environments, traffic exposure, and intersection geometry factors as well as site-specific unobserved heterogeneity.A unique sample of 31 intersections were used in this before-after study. From this sample, video data were collected before and after implementing AWS. In total, 245 h of video were automatically processed and corrected using a specialized computer vision software. More than 68,000 (37,668 before and 31,305 after AWS treatment) road user trajectories were obtained from 104 approaches. The results show that the conversion of MAS to AWS intersections significantly decreased vehicle speed and increased post-encroachment time. This work also shows that implementing AWS significantly increased the yielding rates from 45.7% to 76.7% in MAS conditions and reduced the average speed of motor-vehicles. Using multi-level regression model, it is estimated that when the intersection was converted from MAS to AWS, the minimum speed in the major approaches was reduced by 60.0%.

Impact of wire sensor location on coronary physiological measurement

Abstract Background The location of the wire sensor to measure FFR and dPR has not been systemically studied. Therefore, we hypothesize that the coronary physiological measurements will vary with the location of the sensor. This exploratory study is to assess whole cycle pressure distal/pressure aorta (Pd/Pa) ratio, FFR, and dPR at different sensor position within the vessel to evaluate the impact of pressure drop and measurements based on sensor position. Methods 54 patients who had an indication for coronary angiography were screened and 30 consecutive patients were enrolled. The OptoWire 2 or 3 generation fiberoptic pressure wire was used to assess Pd/Pa, dPR, and FFR. Our primary goal is to test if those measurements vary with the FFR wire sensor placed at 10mm (proximal), 35–45mm (mid), and greater than or equal 60–70mm (distal) distal to the target lesion respectively. We used a multilevel linear regression approach. Results Of 30 patients enrolled, 23 (76.6%) were males, mean age was 64.7 years (± 11.0 yrs), and mean stenosis was 61.6% (±13.4%). Adjusting for age, gender, and severity of stenosis, results showed that for all three measures (Pd/Pa, DPR, and FFR) (Figure 1), pressure decreased in a linear fashion the further the sensor was from the target lesion (p values <0.001). Further, pairwise comparisons of the measurements at adjacent locations similarly showed significant declines in pressure (p values <0.001). Conclusions This is first study to demonstrate that the location of the pressure wire can impact the results of both resting and hyperemic pressures which can cause a false negative result. This is especially important where the values are near the cut-off. Funding Acknowledgement Type of funding sources: None.

The Association Between Physical Activity and Self-Rated Health Among Older Adults

Although there is substantial evidence on the association between physical activity and self-rated health among older adults globally, there is a paucity of findings on this topic for the aging Indonesian population. The present study was conducted to investigate the association between self-reported physical activity and self-rated health among older adults in Indonesia. The data used were from the fifth wave of the Indonesian Family Life Survey (IFLS). A multilevel regression approach was used for analysis. A sample of 1,813 older adults (aged 60 and over) within 311 communities was used in the study. Self-reported physical activity measured by time spent walking was the independent variable, and self-rated health was the dependent variable. The analysis was adjusted for gender, age, education level, marital status, religiosity, retirement status, neighborhood safety, extroversion, and subjective poverty. The results revealed that increased physical activity was consistently associated with better self-rated health in both unadjusted (β=0.04, p<0.05) and adjusted (β=0.06, p<0.01) models. To improve public health in Indonesia, primary health care professionals, health policymakers, and other health promoters should consider including physical activity in health policies and encourage older individuals to engage in regular physical activity.

The impact of nurse staffing levels on nursing-sensitive patient outcomes: a multilevel regression approach

The goal of this study is to provide empirical evidence of the impact of nurse staffing levels on seven nursing-sensitive patient outcomes (NSPOs) at the hospital unit level. Combining a very large set of claims data from a German health insurer with mandatory quality reports published by every hospital in Germany, our data set comprises approximately 3.2 million hospital stays in more than 900 hospitals over a period of 5 years. Accounting for the grouping structure of our data (i.e., patients grouped in unit types), we estimate cross-sectional, two-level generalized linear mixed models (GLMMs) with inpatient cases at level 1 and units types (e.g., internal medicine, geriatrics) at level 2. Our regressions yield 32 significant results in the expected direction. We find that differentiating between unit types using a multilevel regression approach and including postdischarge NSPOs adds important insights to our understanding of the relationship between nurse staffing levels and NSPOs. Extending our main model by categorizing inpatient cases according to their clinical complexity, we are able to rule out hidden effects beyond the level of unit types.

Heterogeneity in Manufacturing Growth Risk

We analyze output growth risk with respect to financial conditions across U.S. manufacturing industries. Using a multi-level quantile regression approach, we find strong heterogeneity in growth risk, particularly between the more vulnerable durable goods sector and the more resilient nondurable goods sector. Moreover, we show that industry characteristics significantly explain these differences. Large, or material intensive durable goods producing, or energy intensive nondurable goods producing industries are more vulnerable to adverse financial conditions, while industries engaging in labor hoarding, or with a high capital or overhead labor intensity are less susceptible.

Longitudinal patterns of cocaine use among patients receiving injectable hydromorphone or diacetylmorphine for the treatment of opioid use disorder: A growth curve modeling approach

Background and aimsCocaine use is prevalent among people receiving injectable opioid agonist treatment. Investigations of cocaine use in this population have been descriptive and the potential heterogeneity existing in patterns of use have not been characterized. As such, among patients receiving injectable opioid agonist treatment, this study aimed to: 1) quantify intra- and inter-individual variation in cocaine use over 24-months and; 2) determine how predictors of interest explained this variation. MethodsParticipants were patients receiving injectable opioid agonist treatment for opioid use disorder. Study visits were completed at baseline prior to receiving treatment, and 3,6,9,12,18, and 24 months after baseline. A multi-level regression approach to growth curve modeling was employed to estimate and explain intra- (within-person) and inter-individual (between-person) variation in cocaine use. ResultsSignificant intra and inter-individual variation in cocaine use was identified over 24-months. Treatment engagement was on average associated with reductions in the prior month number of days of cocaine use (range: 0–30)(Estimate (standard error): -0.05(0.02), p = 0.003). On average, men reported less cocaine use compared to women (Estimate (standard error): -5.91(1.57), p=<0.001), and participants reporting ever regularly using cocaine at baseline reported more cocaine use over 24-months compared to participants reporting never regularly using cocaine (Estimate (standard error): 4.72 (1.91), p = 0.013). ConclusionsSignificant reductions in cocaine use were observed and significant heterogeneity in patterns of cocaine use was identified. These heterogeneous cocaine use profiles suggest that an individualized approach to care will be critical in responding to patients’ cocaine use in injectable opioid agonist treatment.

Large-scale baseflow index prediction using hydrological modelling, linear and multilevel regression approaches

Baseflow is critical for water balance budget, water resources management, and environmental evaluation. Prediction of baseflow index (BFI), the ratio of baseflow to total streamflow, has a great significance in unravelling the baseflow characteristics for large scale trajectory. Therefore, this study compares BFI predictive performance derived from a new multilevel regression approach along with two other commonly used approaches: hydrological modelling (SIMHYD, a simplified version of the HYDROLOG model, and Xinanjiang model), and linear regression (traditional linear regression, and alternative traditional regression considers the second-order interaction). The multilevel regression approach does not only group the catchments into the four climate zones (arid, tropics, equiseasonal and winter rainfall), but also considers inter-catchment and inter-climate zone variances. Likewise, calibration and two regionalisation techniques namely spatial proximity and integrated similarity are used to obtain the BFI from hydrological modelling approach. Correspondingly, the traditional linear regression technique estimates BFI establishing linear regressions between catchment attributes and four climate zones. Then, all the three approaches are evaluated against combined average estimation from four well-parameterised baseflow separation methods (Lyne-Hollick (LH), United Kingdom Institute of Hydrology (UKIH), Chapman-Maxwell (CM) and Eckhardt (ECK)) at 596 catchments across Australia for 1980–2012. The findings show that the multilevel regression has greatly improved the performance of BFI prediction in comparison to other methods. In particular, the two calibrated and regionalised hydrological models perform worst in predicting BFI with a Nash-Sutcliffe Efficiency (NSE) of −8.44 and −2.58 along with an absolute percent bias (PBIAS) of 81% and 146% (overestimation of baseflow), respectively. However, the traditional linear regression remains in intermediate position with the NSE of 0.57 and bias of 25. In addition, alternative traditional regression also shows very close proximity. In contrast, the multilevel regression approach shows the best performance with the NSE of 0.75 and bias of 19%. The study also demonstrates that the multilevel regression approach can improve BFI prediction, and shows potential for being used in the prediction of other hydrological signatures in large-scale.

Comparing students' engineering and science aspirations from age 10 to 16: Investigating the role of gender, ethnicity, cultural capital, and attitudinal factors

AbstractBackgroundWomen (along with minority ethnic and low‐income communities) remain underrepresented in engineering, despite a 30‐year history of research and equality legislation. Compared with the United States and other European Union countries, this underrepresentation is particularly pronounced in the United Kingdom. While existing literature gives insights into factors shaping retention and progression in university engineering students, comparatively less is known about the development of primary and secondary school students' engineering aspirations.PurposeThis paper contrasts science and engineering analyses to explore how relationships between background and attitudinal factors and aspirations change across primary and secondary schools. We examine the relative influence of gender on aspirations in both science and engineering.Design/MethodWe drew on survey data from more than 20,000 English students from the ASPIRES projects. A multilevel regression approach is implemented to test for the effects of gender, ethnicity, and cultural capital on science and engineering aspirations.ResultsGender is the main factor related to engineering aspirations, while science aspirations are influenced by a broader range of factors. School‐level factors become increasingly important for engineering aspirations. We also report evidence of the early distinctiveness of young women who aspire to engineering in terms of their relatively high self‐concept and motivations.ConclusionsThe association of engineering with masculinity is evident in aspirations from age 10, and students aspiring to engineering are distinctive in several respects. Efforts aimed at improving participation in engineering might more usefully focus on challenging the elitist culture and practices, which may influence student perceptions, rather than focusing on changing student aspirations directly.

Using 3 Health Surveys to Compare Multilevel Models for Small Area Estimation for Chronic Diseases and Health Behaviors.

BackgroundWe used a multilevel regression and poststratification approach to generate estimates of health-related outcomes using Behavioral Risk Factor Surveillance System 2013 (BRFSS) data for the 500 US cities. We conducted an empirical study to investigate whether the approach is robust using different health surveys.MethodsWe constructed a multilevel logistic model with individual-level age, sex, and race/ethnicity as predictors (Model I), and sequentially added educational attainment (Model II) and area-level poverty (Model III) for 5 health-related outcomes using the nationwide BRFSS, the Massachusetts BRFSS 2013 (a state subset of nationwide BRFSS), and the Boston BRFSS 2010/2013 (an independent survey), respectively. We applied each model to the Boston population (2010 Census) to predict each outcome in Boston and compared each with corresponding Boston BRFSS direct estimates.ResultsUsing Model I for the nationwide BRFSS, estimates of diabetes, high blood pressure, physical inactivity, and binge drinking fell within the 95% confidence interval of corresponding Boston BRFSS direct estimates. Adding educational attainment and county-level poverty (Models II and III) further improved their accuracy, particularly for current smoking (the model-based estimate was 15.2% by Model I and 18.1% by Model II). The estimates based on state BRFSS and Boston BRFSS models were similar to those based on the nationwide BRFSS, but area-level poverty did not improve the estimates significantly.ConclusionThe estimates of health-related outcomes were similar using different health surveys. Model specification could vary by surveys with different geographic coverage.