Pub Date : 2020-08-07DOI: 10.1080/24709360.2020.1796176
H. Katki, R. Dey, P. Saha-Chaudhuri
Risk stratification is the ability of a test or model to separate those at high vs. low risk of disease. There is no risk stratification metric that is in terms of the number of people requiring testing, which would help with considering the benefits, harms, and costs associated with the test and interventions. We introduce the expected number needed to test (NNtest) to identify one more disease case than by randomly selecting people for disease ascertainment. We show that NNtest measures risk stratification, allowing us to decompose NNtest into components that contrast the increase in risk upon testing positive (‘concern’) versus the decrease in risk upon testing negative (‘reassurance’). A graph of the reciprocals of concern vs. reassurance have linear contours of constant NNtest, visualizing the relative importance and tradeoff of each component to better understand the properties of risk thresholds with equal NNtest. We apply NNtest to the controversy over the risk threshold for who should get testing for BRCA1/2 mutations that cause high risks of breast and ovarian cancers. We show that risk thresholds between 0.78% and 5% optimize NNtest. At these thresholds, people will require risk-model evaluation to find one more mutation-carrier. However, these thresholds of equal NNtest provide very different concern and reassurance, with 0.78% providing much more reassurance (and thus much less concern) than 5%. Given that genetic testing costs are declining rapidly, the greater reassurance provided by the 0.78% threshold might be deemed more important than the greater concern provided by the 5% threshold.
{"title":"Number needed to test: quantifying risk stratification provided by diagnostic tests and risk predictions","authors":"H. Katki, R. Dey, P. Saha-Chaudhuri","doi":"10.1080/24709360.2020.1796176","DOIUrl":"https://doi.org/10.1080/24709360.2020.1796176","url":null,"abstract":"Risk stratification is the ability of a test or model to separate those at high vs. low risk of disease. There is no risk stratification metric that is in terms of the number of people requiring testing, which would help with considering the benefits, harms, and costs associated with the test and interventions. We introduce the expected number needed to test (NNtest) to identify one more disease case than by randomly selecting people for disease ascertainment. We show that NNtest measures risk stratification, allowing us to decompose NNtest into components that contrast the increase in risk upon testing positive (‘concern’) versus the decrease in risk upon testing negative (‘reassurance’). A graph of the reciprocals of concern vs. reassurance have linear contours of constant NNtest, visualizing the relative importance and tradeoff of each component to better understand the properties of risk thresholds with equal NNtest. We apply NNtest to the controversy over the risk threshold for who should get testing for BRCA1/2 mutations that cause high risks of breast and ovarian cancers. We show that risk thresholds between 0.78% and 5% optimize NNtest. At these thresholds, people will require risk-model evaluation to find one more mutation-carrier. However, these thresholds of equal NNtest provide very different concern and reassurance, with 0.78% providing much more reassurance (and thus much less concern) than 5%. Given that genetic testing costs are declining rapidly, the greater reassurance provided by the 0.78% threshold might be deemed more important than the greater concern provided by the 5% threshold.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"5 1","pages":"134 - 148"},"PeriodicalIF":0.0,"publicationDate":"2020-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1796176","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42483253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-07-02DOI: 10.1080/24709360.2021.1936421
Shasha Han, D. Rubin
Basic propensity score methodology is designed to balance the distributions of multivariate pre-treatment covariates when comparing one active treatment with one control treatment. However, practical settings often involve comparing more than two treatments, where more complicated contrasts than the basic treatment-control one, , are relevant. Here, we propose the use of contrast-specific propensity scores (CSPS), which allows the creation of treatment groups of units that are balanced with respect to bifurcations of the specified contrasts and the multivariate space spanned by these bifurcations.
{"title":"Contrast-specific propensity scores","authors":"Shasha Han, D. Rubin","doi":"10.1080/24709360.2021.1936421","DOIUrl":"https://doi.org/10.1080/24709360.2021.1936421","url":null,"abstract":"Basic propensity score methodology is designed to balance the distributions of multivariate pre-treatment covariates when comparing one active treatment with one control treatment. However, practical settings often involve comparing more than two treatments, where more complicated contrasts than the basic treatment-control one, , are relevant. Here, we propose the use of contrast-specific propensity scores (CSPS), which allows the creation of treatment groups of units that are balanced with respect to bifurcations of the specified contrasts and the multivariate space spanned by these bifurcations.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"5 1","pages":"1 - 8"},"PeriodicalIF":0.0,"publicationDate":"2020-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2021.1936421","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46223106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-12DOI: 10.1080/24709360.2021.1978270
D. Sisti, E. Rocchi, S. Peluso, S. Amatori, M. Carletti
The novel coronavirus SARS-CoV-2 was first identified in China in December 2019. In just over five months, the virus affected over 4 million people and caused about 300,000 deaths. This study aimed to model new COVID-19 cases in Italian regions using a new curve. A new empirical curve is proposed to model the number of new cases of COVID-19. It resembles a known exponential growth curve, which has a straight line as an exponent, but in the growth curve proposed, the exponent is a logistic curve multiplied for a straight line. This curve shows an initial phase, the expected exponential growth, then rises to the maximum value and finally reaches zero. We characterized the epidemic growth patterns for the entire Italian nation and each of the 20 Italian regions. The estimated growth curve has been used to calculate the expected time of the beginning, the time related to peak, and the end of the epidemics. Our analysis explores the development of the outbreaks in Italy and the impact of the containment measures. Data obtained are useful to forecast future scenarios and the possible end of the epidemic.
{"title":"A new regression model for the forecasting of COVID-19 outbreak evolution: an application to Italian data","authors":"D. Sisti, E. Rocchi, S. Peluso, S. Amatori, M. Carletti","doi":"10.1080/24709360.2021.1978270","DOIUrl":"https://doi.org/10.1080/24709360.2021.1978270","url":null,"abstract":"The novel coronavirus SARS-CoV-2 was first identified in China in December 2019. In just over five months, the virus affected over 4 million people and caused about 300,000 deaths. This study aimed to model new COVID-19 cases in Italian regions using a new curve. A new empirical curve is proposed to model the number of new cases of COVID-19. It resembles a known exponential growth curve, which has a straight line as an exponent, but in the growth curve proposed, the exponent is a logistic curve multiplied for a straight line. This curve shows an initial phase, the expected exponential growth, then rises to the maximum value and finally reaches zero. We characterized the epidemic growth patterns for the entire Italian nation and each of the 20 Italian regions. The estimated growth curve has been used to calculate the expected time of the beginning, the time related to peak, and the end of the epidemics. Our analysis explores the development of the outbreaks in Italy and the impact of the containment measures. Data obtained are useful to forecast future scenarios and the possible end of the epidemic.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"6 1","pages":"48 - 56"},"PeriodicalIF":0.0,"publicationDate":"2020-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44476898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2020.1794705
Hua Shen, R. Cook
ABSTRACT We consider the setting in which a categorical exposure variable of interest can only be measured subject to misclassification via surrogate variables. These surrogate variables may represent the classification of an individual via imperfect diagnostic tests. In such settings, a random number of diagnostic tests may be ordered at the discretion of a treating physician with the decision to order further tests made in a sequential fashion based on the results of preliminary test results. Because the underlying latent status is not ascertainable these cheaper but imperfect surrogate test results are used in lieu of the definitive classification in a model for a long-term outcome. Naive use of a single surrogate or functions of the available surrogates can lead to biased estimators of the association and invalid inference. We propose a likelihood-based approach for modeling the effect of the latent variable in the absence of validation data with estimation based on an expectation–maximization (EM) algorithm. The method yields consistent and efficient estimates and is shown to out-perform several common alternative approaches. The performance of the proposed method is demonstrated in simulation studies and its utility is illustrated by applying the proposed method to the stimulating study on breast cancer.
{"title":"Regression with incomplete multivariate surrogate responses for a latent covariate","authors":"Hua Shen, R. Cook","doi":"10.1080/24709360.2020.1794705","DOIUrl":"https://doi.org/10.1080/24709360.2020.1794705","url":null,"abstract":"ABSTRACT We consider the setting in which a categorical exposure variable of interest can only be measured subject to misclassification via surrogate variables. These surrogate variables may represent the classification of an individual via imperfect diagnostic tests. In such settings, a random number of diagnostic tests may be ordered at the discretion of a treating physician with the decision to order further tests made in a sequential fashion based on the results of preliminary test results. Because the underlying latent status is not ascertainable these cheaper but imperfect surrogate test results are used in lieu of the definitive classification in a model for a long-term outcome. Naive use of a single surrogate or functions of the available surrogates can lead to biased estimators of the association and invalid inference. We propose a likelihood-based approach for modeling the effect of the latent variable in the absence of validation data with estimation based on an expectation–maximization (EM) algorithm. The method yields consistent and efficient estimates and is shown to out-perform several common alternative approaches. The performance of the proposed method is demonstrated in simulation studies and its utility is illustrated by applying the proposed method to the stimulating study on breast cancer.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"247 - 264"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1794705","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45075884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2018.1469810
Xiao‐Hua Zhou
In biomedical research, missing data are a common problem. The statistical literature to solve this problem is well developed but overly technical and complicated for health science researchers who are not experts in statistics or methodology. In this paper, we review available statistical methods for handling missing data and provide health science researchers with the means of understanding the importance of missing data in their own personal research, and the ability to use these methods given the available software.
{"title":"Challenges and strategies in analysis of missing data","authors":"Xiao‐Hua Zhou","doi":"10.1080/24709360.2018.1469810","DOIUrl":"https://doi.org/10.1080/24709360.2018.1469810","url":null,"abstract":"In biomedical research, missing data are a common problem. The statistical literature to solve this problem is well developed but overly technical and complicated for health science researchers who are not experts in statistics or methodology. In this paper, we review available statistical methods for handling missing data and provide health science researchers with the means of understanding the importance of missing data in their own personal research, and the ability to use these methods given the available software.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"15 - 23"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2018.1469810","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45344516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2020.1734391
S. Vilakati, G. Cortese
The focus on two-stage randomization designs with survival end points is on estimating and comparing survival distributions for the different treatment policies. The objective is to identify the treatment policy which prolongs survival. In this paper, a method for comparing two treatment policies is proposed. These treatment policies may be shared path or independent path treatment policies. Simulation studies are performed to evaluate the performance of the new approach. The simulation studies reveal that the new method has better statistical power in cases where the survival curves cross. The new method is applied to a clinical trial dataset for leukemia.
{"title":"Weighted Lin and Xu test for two-stage randomization designs","authors":"S. Vilakati, G. Cortese","doi":"10.1080/24709360.2020.1734391","DOIUrl":"https://doi.org/10.1080/24709360.2020.1734391","url":null,"abstract":"The focus on two-stage randomization designs with survival end points is on estimating and comparing survival distributions for the different treatment policies. The objective is to identify the treatment policy which prolongs survival. In this paper, a method for comparing two treatment policies is proposed. These treatment policies may be shared path or independent path treatment policies. Simulation studies are performed to evaluate the performance of the new approach. The simulation studies reveal that the new method has better statistical power in cases where the survival curves cross. The new method is applied to a clinical trial dataset for leukemia.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"221 - 237"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1734391","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49638481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01Epub Date: 2019-02-23DOI: 10.1080/24709360.2019.1572344
Chunhua Weng
Clinical data is the staple of modern learning health systems. It promises to accelerate biomedical discovery and improves the efficiency of clinical and translational research but is also fraught with significant data quality issues. This paper aims to provide a life cycle perspective of clinical data quality issues along with recommendations for establishing appropriate expectations for research based on real-world clinical data and best practices for reusing clinical data as a secondary data source.
{"title":"Clinical data quality: a data life cycle perspective.","authors":"Chunhua Weng","doi":"10.1080/24709360.2019.1572344","DOIUrl":"https://doi.org/10.1080/24709360.2019.1572344","url":null,"abstract":"<p><p>Clinical data is the staple of modern learning health systems. It promises to accelerate biomedical discovery and improves the efficiency of clinical and translational research but is also fraught with significant data quality issues. This paper aims to provide a life cycle perspective of clinical data quality issues along with recommendations for establishing appropriate expectations for research based on real-world clinical data and best practices for reusing clinical data as a secondary data source.</p>","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"6-14"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2019.1572344","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"37810160","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2020.1842048
J. R. Khan, Jahida Gulshan
Earlier studies to assess the effects of risk factors on child nutritional status in Bangladesh have used conventional regression models that are inadequate to capture a complete scenario of effects. Therefore, this study aimed to evaluate the heterogeneous effects of factors at different points of conditional height-for-age Z-score (HAZ) distribution accounting for cluster-level variation using linear quantile mixed model (LQMM) and to compare them with a linear mixed model (LMM). In addition, an unconditional quantile model (UQM) was used to measure the effect of factors on the unconditional (marginal) HAZ distribution. A total of 6340 children aged 0–59 months extracted from the 2014 Bangladesh Demographic and Health Survey. Different factors – maternal characteristics (age, occupation, nutritional status, parity, birth interval), parental education, child age, breastfeeding status, and morbidity had significant heterogeneous effects on HAZ distribution. For example, secondary or higher educated parents had substantial differential impacts on the lower tail and upper tail of the child HAZ distribution, which was masked by LMM estimate. Moreover, significant cluster-level variations found across all quantiles of child HAZ. During intervention design, heterogeneous effects of factors and cluster variation ought to consider addressing the undernutrition problem in Bangladesh.
{"title":"Heterogeneous effects of factors on child nutritional status in Bangladesh using linear quantile mixed model","authors":"J. R. Khan, Jahida Gulshan","doi":"10.1080/24709360.2020.1842048","DOIUrl":"https://doi.org/10.1080/24709360.2020.1842048","url":null,"abstract":"Earlier studies to assess the effects of risk factors on child nutritional status in Bangladesh have used conventional regression models that are inadequate to capture a complete scenario of effects. Therefore, this study aimed to evaluate the heterogeneous effects of factors at different points of conditional height-for-age Z-score (HAZ) distribution accounting for cluster-level variation using linear quantile mixed model (LQMM) and to compare them with a linear mixed model (LMM). In addition, an unconditional quantile model (UQM) was used to measure the effect of factors on the unconditional (marginal) HAZ distribution. A total of 6340 children aged 0–59 months extracted from the 2014 Bangladesh Demographic and Health Survey. Different factors – maternal characteristics (age, occupation, nutritional status, parity, birth interval), parental education, child age, breastfeeding status, and morbidity had significant heterogeneous effects on HAZ distribution. For example, secondary or higher educated parents had substantial differential impacts on the lower tail and upper tail of the child HAZ distribution, which was masked by LMM estimate. Moreover, significant cluster-level variations found across all quantiles of child HAZ. During intervention design, heterogeneous effects of factors and cluster variation ought to consider addressing the undernutrition problem in Bangladesh.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"265 - 281"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1842048","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43047827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2020.1719690
Lucy Beggs, R. Briscoe, C. Griffiths, G. Ellison, M. Gilthorpe
Background: Intervention differential effects (IDEs) occur where changes in an outcome depend upon the initial values of that outcome. Although methods to identify IDEs are well documented, there remains a lack of understanding about the circumstances under which these methods are robust. One context that has not been explored is the identification of intervention differential effect in studies where sample selection is based on the initial value of the outcome being evaluated. We hypothesise that, in such settings, established methods for detecting IDEs will struggle to discriminate these from regression to the mean. Methods: Using simulated datasets of weight-loss intervention programmes that recruit according to initial body mass index, we explore the reliability of Oldham's method and multilevel modelling (MLM) to detect IDEs. Results: In datasets simulated with no IDE, Oldham's method and MLM yield Type I error rates >90%, confirming that threshold selection/truncation leads to bias due to regression to the mean. Type I error rates return close to 5% for both methods when a control group is introduced. Conclusions: Oldham's method and MLM can robustly detect IDEs in this setting, but only if analyses incorporate a control group for comparison.
{"title":"Intervention differential effects and regression to the mean in studies where sample selection is based on the initial value of the outcome variable: an evaluation of methods illustrated in weight-management studies","authors":"Lucy Beggs, R. Briscoe, C. Griffiths, G. Ellison, M. Gilthorpe","doi":"10.1080/24709360.2020.1719690","DOIUrl":"https://doi.org/10.1080/24709360.2020.1719690","url":null,"abstract":"Background: Intervention differential effects (IDEs) occur where changes in an outcome depend upon the initial values of that outcome. Although methods to identify IDEs are well documented, there remains a lack of understanding about the circumstances under which these methods are robust. One context that has not been explored is the identification of intervention differential effect in studies where sample selection is based on the initial value of the outcome being evaluated. We hypothesise that, in such settings, established methods for detecting IDEs will struggle to discriminate these from regression to the mean. Methods: Using simulated datasets of weight-loss intervention programmes that recruit according to initial body mass index, we explore the reliability of Oldham's method and multilevel modelling (MLM) to detect IDEs. Results: In datasets simulated with no IDE, Oldham's method and MLM yield Type I error rates >90%, confirming that threshold selection/truncation leads to bias due to regression to the mean. Type I error rates return close to 5% for both methods when a control group is introduced. Conclusions: Oldham's method and MLM can robustly detect IDEs in this setting, but only if analyses incorporate a control group for comparison.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"172 - 188"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1719690","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41500269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-01-01DOI: 10.1080/24709360.2020.1721965
O. Marrero
ABSTRACT We present a detailed exposition of the development and application of a likelihood-ratio test for seasonality. It is well known that likelihood-ratio tests have optimal power properties. We assess the test's performance by means of a simulation study. The test's application is illustrated with three examples that have different alternative hypotheses, thus extending the original presentation of the test. These examples are not artificial or contrived, but they come from actual, real applications. As far as we know, these are the only completely worked-out examples of this test's application that are available in the literature. Thus, our exposition can serve as a tutorial on the test's application. Our presentation is detailed so as to facilitate further extension and application of the test to other alternative hypotheses. We supply pertinent R computer code in an appendix. For those who teach maximum-likelihood estimation, our examples provide interesting, real-life cases that may be used in teaching.
{"title":"Application and extension of a likelihood-ratio test for seasonality in epidemiological data","authors":"O. Marrero","doi":"10.1080/24709360.2020.1721965","DOIUrl":"https://doi.org/10.1080/24709360.2020.1721965","url":null,"abstract":"ABSTRACT We present a detailed exposition of the development and application of a likelihood-ratio test for seasonality. It is well known that likelihood-ratio tests have optimal power properties. We assess the test's performance by means of a simulation study. The test's application is illustrated with three examples that have different alternative hypotheses, thus extending the original presentation of the test. These examples are not artificial or contrived, but they come from actual, real applications. As far as we know, these are the only completely worked-out examples of this test's application that are available in the literature. Thus, our exposition can serve as a tutorial on the test's application. Our presentation is detailed so as to facilitate further extension and application of the test to other alternative hypotheses. We supply pertinent R computer code in an appendix. For those who teach maximum-likelihood estimation, our examples provide interesting, real-life cases that may be used in teaching.","PeriodicalId":37240,"journal":{"name":"Biostatistics and Epidemiology","volume":"4 1","pages":"189 - 220"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/24709360.2020.1721965","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44181921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：