Modeling Error in Quantitative Macro-Comparative Research

Much quantitative macro-comparative research (QMCR) relies on a common set of published data sources to answer similar research questions using a limited number of statistical tools. Since all researchers have access to much the same data, one might expect quick convergence of opinion on most topics. In reality, of course, differences of opinion abound and persist. Many of these differences can be traced, implicitly or explicitly, to the different ways researchers choose to model error in their analyses. Much careful attention has been paid in the political science literature to the error structures characteristic of time series cross-sectional (TSCE) data, but much less attention has been paid to the modeling of error in broadly cross-national research involving large panels of countries observed at limited numbers of time points. Here, and especially in the sociology literature, multilevel modeling has become a hegemonic but often poorly understood research tool. I argue that widely-used types of multilevel models, commonly known as fixed effects models (FEMs) and random effects models (REMs), can produce wildly spurious results when applied to trended data due to mis-specification of error. I suggest that in most commonly-encountered scenarios, difference models are more appropriate for use in QMC.

Download Full-text

Getting the Within Estimator of Cross-Level Interactions in Multilevel Models with Pooled Cross-Sections: Why Country Dummies (Sometimes) Do Not Do the Job

Sociological Methodology ◽

10.1177/0081175018809150 ◽

2018 ◽

Vol 49 (1) ◽

pp. 190-219 ◽

Cited By ~ 4

Author(s):

Marco Giesselmann ◽

Alexander W. Schmidt-Catran

Keyword(s):

Cross Sections ◽

Fixed Effects ◽

Multilevel Models ◽

Unobserved Heterogeneity ◽

Level Structure ◽

Cross Sectional Survey ◽

Social Scientists ◽

Cross Sectional ◽

Country Level ◽

Fixed Effects Estimator

Multilevel models with persons nested in countries are increasingly popular in cross-country research. Recently, social scientists have started to analyze data with a three-level structure: persons at level 1, nested in year-specific country samples at level 2, nested in countries at level 3. By using a country fixed-effects estimator, or an alternative equivalent specification in a random-effects framework, this structure is increasingly used to estimate within-country effects in order to control for unobserved heterogeneity. For the main effects of country-level characteristics, such estimators have been shown to have desirable statistical properties. However, estimators of cross-level interactions in these models are not exhibiting these attractive properties: as algebraic transformations show, they are not independent of between-country variation and thus carry country-specific heterogeneity. Monte Carlo experiments consistently reveal the standard approaches to within estimation to provide biased estimates of cross-level interactions in the presence of an unobserved correlated moderator at the country level. To obtain an unbiased within-country estimator of a cross-level interaction, effect heterogeneity must be systematically controlled. By replicating a published analysis, we demonstrate the relevance of this extended country fixed-effects estimator in research practice. The intent of this article is to provide advice for multilevel practitioners, who will be increasingly confronted with the availability of pooled cross-sectional survey data.

Download Full-text

Study protocol for a systematic review and meta-analysis of comorbidities and stroke characteristics associated with troponin elevation after acute stroke

BMJ Open ◽

10.1136/bmjopen-2020-043613 ◽

2021 ◽

Vol 11 (1) ◽

pp. e043613

Author(s):

Amado Jimenez-Ruiz ◽

Sebastian Fridman ◽

Luciano A Sposato

Keyword(s):

Systematic Review ◽

Renal Failure ◽

Fixed Effects ◽

Meta Analysis ◽

Sensitivity Analyses ◽

Stroke Recurrence ◽

Published Data ◽

Stroke Severity ◽

Cross Sectional ◽

Cardiovascular Comorbidities

IntroductionIt is unknown which comorbidities and stroke characteristics are associated with elevated cardiac troponin (cTn) levels after stroke. The main objective of this systematic review and meta-analysis is to assess the association of elevated cTn with preexisting cardiovascular comorbidities (eg, coronary artery disease, heart failure and structural heart disease), specific stroke characteristics (eg, infarct/haemorrhage size, stroke severity, insular cortex involvement) and renal failure after ischaemic stroke (IS) or intracranial haemorrhage (ICH). The secondary objective is to evaluate the association of elevated cTn with stroke recurrence and death.Methods and analysisWe will include all cross-sectional, case–control, cohort studies and clinical trials involving IS and ICH adult patients (≥18 years), published between 1 January 1990 and 31 December 2020 in English or Spanish, reporting the proportion with elevated cTn. We will search PubMed, EMBASE and Web of Science by applying predefined search terms. Two reviewers will independently screen titles and abstracts, retrieve full texts, extract the data in a predesigned form, and assess the risk of bias. We will apply random-effects or fixed-effects meta-analyses to estimate the association between cardiovascular comorbidities, stroke characteristics and renal failure with cTn elevation. We will report results as risk ratios or ORs. We will perform sensitivity analyses for subtypes of cTn (cTn-I and cTn-T), regular versus high-sensitivity assays, and type of stroke (IS vs ICH). We will estimate heterogeneity by using t2Q and I2 measures. We will use funnel plots, Rosenthal’s Fail-Safe N, Duval and Tweedie’s trim and fill procedure, and Egger’s regression intercept to assess publication bias.Ethics and disseminationThis review will be based on published data and does therefore not require ethical clearance. The results will be published in peer-reviewed journals.PROSPERO registration numberCRD42020203126.

Download Full-text

Hypertensive disorders of pregnancy and risk of asthma in offspring: protocol for a systematic review and meta-analysis

BMJ Open ◽

10.1136/bmjopen-2019-035145 ◽

2020 ◽

Vol 10 (4) ◽

pp. e035145 ◽

Cited By ~ 1

Author(s):

Ping Li ◽

Tao Xiong ◽

Yong Hu

Keyword(s):

Systematic Review ◽

English Language ◽

Data Extraction ◽

Meta Analysis ◽

Published Data ◽

Hypertensive Disorders Of Pregnancy ◽

Cross Sectional ◽

Random Effects Models ◽

Hypertensive Disorders ◽

Increased Risk

IntroductionHypertensive disorders of pregnancy (HDP), one of the most common obstetrical complications, has been reported to have a controversial relationship with the increased risk of asthma in offspring. No systematic review of this topic has been performed. The aim of this systematic review will be to summarise the available evidence examining the association between HDP and the risk of asthma in offspring.Methods and analysisWe will follow the Preferred Reporting Items for Systematic Reviews and Meta-Analyses and Meta-analysis of Observational Studies in Epidemiology guidelines. A systematic search of the PubMed, Embase, Cochrane and Web of Science databases will be performed using a detailed search strategy from database inception through 31 December 2019. Cohort, case-control and cross-sectional studies that report a diagnosis of maternal HDP and asthma in offspring will be included. Studies will be limited to the English language and include only human participants. Two independent reviewers will conduct the study selection, data extraction and risk of bias assessments using a standardised data extraction form. A meta-analysis will be performed to calculate overall pooled estimates using the generic inverse variance method. The data will be synthesised by either fixed-effect or random effects models according to heterogeneity tests. All analyses will be performed in Stata 14 and RevMan 5.3. High-quality evidence of the relationship between HDP and the risk of asthma in exposed offspring will be identified through the synthesis of current studies. In addition, the results of subgroup analyses and related secondary outcomes will be reported. The following will be concluded: (i) whether HDP increases the risk of asthma in offspring, (ii) whether HDP affects the severity of asthma in exposed offspring and (iii) whether possible differences in the risk of asthma among different HDP subgroups exist.Ethics and disseminationThere is no requirement for ethics approval because the meta-analysis and systematic review will be based on published data. It is anticipated that the dissemination of results will take place at conferences and through publication in a peer-reviewed journal.

Download Full-text

Assessment of Knowledge and Practice of Various Eye Diseases in Sulaimani: A Descriptive Study

Kurdistan Journal of Applied Research ◽

10.24017/science.2018.2.1 ◽

2018 ◽

pp. 1-6

Author(s):

Abdulrahman M. Ibrahem ◽

Salah Q. Mahmood ◽

Muhammed Babakir-Mina ◽

Salar Ibrahim Ali ◽

Bakhtyar Kamal Talabany

Keyword(s):

Vision Loss ◽

Good Practice ◽

Study Data ◽

Eye Diseases ◽

Vision Impairment ◽

P Value ◽

Published Data ◽

Knowledge Level ◽

Cross Sectional ◽

Knowledge And Practice

Knowledge and practice of public, especially patients about eye diseases are important to reduce magnitude of human blindness. Vision and sight are very essential because they allow us to connect to each other’s. In accordance to the recently published data; the estimation of 253 million people lives with vision impairment, 36 million are blind and 217 million suffer from moderate to severe vision impairment. A descriptive cross-sectional study was conducted at Shahid Dr. Aso Hospital in Sulaimani city-Iraq, from April to August 2017 by face-to-face interview through close ended questionnaire for data collecting. All data were analyzed by Statistical Package for Social Sciences version 22.0 software. P-value of < 0.05 was considered as a statistically significant. A total of 430 patients were randomly chosen to participate in the study. They were 254 (59.1%) males and 176 (40.9%) females. 76.7% of respondents was worrying about vision loss, 0.7% was worrying about hair loss. Of the participants, 32.8% was with a good knowledge level and 40.5% was with a poor knowledge level, as well as 3.1% was in a good practice and 58.8% was in a poor practice level. Female knowledge mean score was 9.53±4.96 and male knowledge mean score was 8.42±5.45, the practice mean score of males was 4.33±1.96 and mean practice score of females was 4.13±1.93. The study data indicate the worrying of participates about vision loss is in the highest proportion and the awareness and practice of patients about eye diseases is unsatisfactory. Health education campaigns are needed to improve personal awareness about vision related problems and for better eye health.

Download Full-text

Extent Pursuit for Cross-Sectional Dependence in Large Panels

SSRN Electronic Journal ◽

10.2139/ssrn.3371722 ◽

2019 ◽

Author(s):

Jiti Gao ◽

Guangming Pan ◽

Yanrong Yang ◽

Bo Zhang

Keyword(s):

Cross Sectional ◽

Large Panels ◽

Cross Sectional Dependence

Download Full-text

Understanding, Choosing, and Unifying Multilevel and Fixed Effect Approaches

Political Analysis ◽

10.1017/pan.2020.41 ◽

2020 ◽

pp. 1-20

Author(s):

Chad Hazlett ◽

Leonard Wainstein

Keyword(s):

Random Effects ◽

Fixed Effects ◽

Multilevel Models ◽

Standard Errors ◽

Point Estimate ◽

Fe Model ◽

Grouped Data ◽

Coefficient Estimates ◽

Political Science Education ◽

Robust Standard Errors

Abstract When working with grouped data, investigators may choose between “fixed effects” models (FE) with specialized (e.g., cluster-robust) standard errors, or “multilevel models” (MLMs) employing “random effects.” We review the claims given in published works regarding this choice, then clarify how these approaches work and compare by showing that: (i) random effects employed in MLMs are simply “regularized” fixed effects; (ii) unmodified MLMs are consequently susceptible to bias—but there is a longstanding remedy; and (iii) the “default” MLM standard errors rely on narrow assumptions that can lead to undercoverage in many settings. Our review of over 100 papers using MLM in political science, education, and sociology show that these “known” concerns have been widely ignored in practice. We describe how to debias MLM’s coefficient estimates, and provide an option to more flexibly estimate their standard errors. Most illuminating, once MLMs are adjusted in these two ways the point estimate and standard error for the target coefficient are exactly equal to those of the analogous FE model with cluster-robust standard errors. For investigators working with observational data and who are interested only in inference on the target coefficient, either approach is equally appropriate and preferable to uncorrected MLM.

Download Full-text

To which extent are per-and poly-fluorinated substances associated to metabolic syndrome?

Reviews on Environmental Health ◽

10.1515/reveh-2020-0144 ◽

2021 ◽

Vol 0 (0) ◽

Author(s):

Maryam Zare Jeddi ◽

Rozita Soltanmohammadi ◽

Giulia Barbieri ◽

Aline S. C. Fabricio ◽

Gisella Pitter ◽

...

Keyword(s):

Metabolic Syndrome ◽

General Population ◽

Meta Analysis ◽

Small Body ◽

Epidemiological Studies ◽

Cross Sectional ◽

Random Effects Models ◽

Epidemiologic Evidence ◽

Adults And Children ◽

Meta Analyses

Abstract Exposure to per- and polyfluoroalkyl substances (PFAS), ubiquitous persistent environmental contaminants, has led to substantial global concern due to their potential environmental and human health effects. Several epidemiological studies have assessed the possible association between PFAS exposure and risk of metabolic syndrome (MetS), however, the results are ambiguous. The aim of this study was to assess the current human epidemiologic evidence on the association between exposure to PFAS and MetS. We performed a systematic search strategy using three electronic databases (PubMed, Scopus, and Web of Science) for relevant studies concerning the associations of PFAS with MetS and its clinical relevance from inception until January 2021. We undertook meta-analyses where there were five or more studies with exposure and outcomes assessments that were reasonably comparable. The pooled odd ratios (ORs) were calculated using random effects models and heterogeneity among studies was assessed by I2 index and Q test. A total of 12 cross-sectional studies (10 studies on the general population and two studies in the occupational settings) investigated the association between PFAS exposure and MetS. We pooled data from seven studies on the general population for perfluorooctanoic acid (PFOA) and perfluorooctanesulfonate (PFOS) and five studies for perfluorohexanesulfonate (PFHxS) and perfluorononanoic acid (PFNA). Predominately, most studies reported no statistically significant association between concentrations of PFAS and MetS. In the meta-analysis, the overall measure of effect was not statistically significant, showing no evidence of an association between concentrations of PFOA, PFOS, PFNA, and PFHxS and the risk of MetS. Based on the results of the meta-analysis, current small body of evidence does not support association between PFAS and MetS. However, due to limited number of studies and substantial heterogeneity, results should be interpreted with caution. Further scrutinizing cohort studies are needed to evaluate the association between various and less well-known PFAS substances and their mixture with MetS and its components in both adults and children in different settings.

Download Full-text

Association between participation in the government subsidy programme for domestic travel and symptoms indicative of COVID-19 infection in Japan: cross-sectional study

BMJ Open ◽

10.1136/bmjopen-2021-049069 ◽

2021 ◽

Vol 11 (4) ◽

pp. e049069

Author(s):

Atsushi Miyawaki ◽

Takahiro Tabuchi ◽

Yasutake Tomata ◽

Yusuke Tsugawa

Keyword(s):

Sore Throat ◽

Fixed Effects ◽

Cross Sectional Study ◽

High Fever ◽

Internet Survey ◽

Government Subsidy ◽

Sampling Weights ◽

Cross Sectional ◽

Taste Disorder ◽

The Government

ObjectiveTo investigate the association between participation in government subsidies for domestic travel (subsidise up to 50% of all travel expenses) introduced nationally in Japan on 22 July 2020 and the incidence of symptoms indicative of COVID-19 infections.DesignCross-sectional analysis of nationally representative survey data.SettingInternet survey conducted between 25 August and 30 September 2020 in Japan. Sampling weights were used to calculate national estimates.Participants25 482 survey respondents (50.3% (12 809) women; mean (SD) age, 48.8 (17.4) years).Main outcome measuresIncidence rate of five symptoms indicative of the COVID-19 infection (high fever, sore throat, cough, headache, and smell and taste disorder) within the past month of the survey, after adjustment for characteristics of individuals and prefecture fixed effects (effectively comparing individuals living in the same prefecture).ResultsAt the time of the survey, 3289 (12.9%) participated in the subsidy programme. After adjusting for potential confounders, we found that participants in the subsidy programme exhibited higher incidence of high fever (adjusted rate, 4.7% for participants vs 3.7% for non-participants; adjusted OR (aOR) 1.83; 95% CI 1.34 to 2.48; p<0.001), sore throat (19.8% vs 11.3%; aOR 2.09; 95% CI 1.37 to 3.19; p=0.002), cough (19.0% vs 11.3%; aOR 1.96; 95% CI 1.26 to 3.01; p=0.008), headache (29.2% vs 25.5%; aOR 1.24; 95% CI 1.08 to 1.44; p=0.006) and smell and taste disorder (2.6% vs 1.8%; aOR 1.98; 95% CI 1.15 to 3.40; p=0.01) compared with non-participants. These findings remained qualitatively unaffected by additional adjustment for the use of 17 preventative measures (eg, social distancing, wearing masks and handwashing) and fear against the COVID-19 infection.ConclusionsThe participation of the government subsidy programme for domestic travel was associated with a higher probability of exhibiting symptoms indicative of the COVID-19 infection.

Download Full-text

Association of bisphenol A with puberty timing: a meta-analysis

Reviews on Environmental Health ◽

10.1515/reveh-2020-0091 ◽

2020 ◽

Vol 0 (0) ◽

Author(s):

Hui Meng ◽

Yunping Zhou ◽

Yunxia Jiang

Keyword(s):

Bisphenol A ◽

Confidence Intervals ◽

Random Effects ◽

Strong Correlation ◽

Fixed Effects ◽

Meta Analysis ◽

Random Effects Models ◽

Study Heterogeneity ◽

The Relationship ◽

Prevalence Ratios

AbstractObjectivesThe results of existing studies on bisphenol A (BPA) and puberty timing did not reach a consensus. Thereby we performed this meta-analytic study to explore the association between BPA exposure in urine and puberty timing.MethodsMeta-analysis of the pooled odds ratios (OR), prevalence ratios (PR) or hazards ratios (HR) with 95% confidence intervals (CI) were calculated and estimated using fixed-effects or random-effects models based on between-study heterogeneity.ResultsA total of 10 studies involving 5621 subjects were finally included. The meta-analysis showed that BPA exposure was weakly associated with thelarche (PR: 0.96, 95% CI: 0.93–0.99), while no association was found between BPA exposure and menarche (HR: 0.99, 95% CI: 0.89–1.12; OR: 1.02, 95% CI: 0.73–1.43), and pubarche (OR: 1.00, 95% CI: 0.79–1.26; PR: 1.00, 95% CI: 0.95–1.05).ConclusionsThere was no strong correlation between BPA exposure and puberty timing. Further studies with large sample sizes are needed to verify the relationship between BPA and puberty timing.

Download Full-text

Fueling the Covid-19 pandemic: summer school holidays and incidence rates in German districts

Journal of Public Health ◽

10.1093/pubmed/fdab080 ◽

2021 ◽

Author(s):

Thomas Plümper ◽

Eric Neumayer

Keyword(s):

Public Health ◽

Fixed Effects ◽

Summer School ◽

Incidence Rates ◽

Contact Tracing ◽

Robert Koch Institute ◽

Cross Sectional ◽

Holiday Period ◽

Health Authorities ◽

Public Health Authorities

AbstractBackgroundThe Robert-Koch-Institute reports that during the summer holiday period a foreign country is stated as the most likely place of infection for an average of 27 and a maximum of 49% of new SARS-CoV-2 infections in Germany.MethodsCross-sectional study on observational data. In Germany, summer school holidays are coordinated between states and spread out over 13 weeks. Employing a dynamic model with district fixed effects, we analyze the association between these holidays and weekly incidence rates across 401 German districts.ResultsWe find effects of the holiday period of around 45% of the average district incidence rates in Germany during their respective final week of holidays and the 2 weeks after holidays end. Western states tend to experience stronger effects than Eastern states. We also find statistically significant interaction effects of school holidays with per capita taxable income and the share of foreign residents in a district’s population.ConclusionsOur results suggest that changed behavior during the holiday season accelerated the pandemic and made it considerably more difficult for public health authorities to contain the spread of the virus by means of contact tracing. Germany’s public health authorities did not prepare adequately for this acceleration.

Download Full-text