Description: Composite indices are increasingly recognized as a useful tool for measuring complex and multidimensional phenomena, such as development, poverty, quality of life, well-being, globalization, competitiveness, freedom, and so on. Technically, a composite index is a mathematical combination of a set of indicators representing the different dimensions of a phenomenon to be measured. The idea of summarizing a complex phenomenon into a single number is not straightforward, as it implies both theoretical and methodological assumptions that must be carefully evaluated to avoid producing results of dubious analytic rigor. One of the main criticisms of composite indices is that their simple "big picture" results may lead users to draw simplistic analytical or policy conclusions. Methods for constructing composite indices are proposed, taking into account appropriate methodologies for different areas of application.
Description: Weighting is one of the major components in survey sampling. For a given sample survey, to each unit of the selected sample is attached a weight that is used to obtain estimates of population parameters of interest (e.g. means, totals, rates). The weighting process usually involves three steps: (i) obtain the design weights, which account for sample selection; (ii) adjust these weights to compensate for nonresponse; (iii) adjust the weights so that the estimates coincide to some population figures known from external trusted sources. The principle behind estimation in a probability survey is that each sample unit represents not only itself, but also several units of the survey population. The design weight of a unit usually refers to the average number of units in the population that each sampled unit represents. This weight is determined by the sampling method and is an important part of the estimation process. While the design weights can be used for estimation, most surveys produce a set of estimation weights by adjusting the design weights to improve accuracy of the final estimates. Once the final estimation weights have been calculated, they are applied to the sample data in order to compute estimates. The ILO Department of Statistics, in collaboration with the ITCILO, is proud to offer the Online course "Weighting Methods & Strategies". This course is directly linked to the course "Sampling Design: A Practical Approach" planned to take place in the spring of 2023 as both courses complement each other. Both courses are considered a learning journey that qualifies the learner to understand comprehensively Sampling design & weighting. Hence, attending both courses is strongly recommended for a fulfilling learning journey.
Target Audience: This course requires basic knowledge of statistics and probability! - It requires basic capacity to run procedures on statistical software using syntax (e.g. Stata do files, Spss syntax files, R scripts, Sas program files, etc.), and in particular with R3. The target audience includes: - Statisticians and practitioners from national statistical offices that have a role in designing household surveys samples.
Description: This course on applied econometrics, has been prepared as an integration of mathematical economics and statistical methods with an objective of extracting the stochastic relationship among variables and their modelling.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: Modelling spatio-temporal phenomena is a key issue in today’s research. However, the extension from pure spatial to a spatio-temporal kriging approach is not trivial. We will look into a set of different perceptions of spatio-temporal dependence and the resulting covariance models (separable, product-sum, metric, …). The practical part deals with the exploration of empirical variograms and the different variogram models applied to different data sets. Finally, some computational aspects regarding R environment and specific packages available for variogram fitting and prediction purposes will be illustrated.
Target Audience: Statisticians interested in applications involving earth sciences
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: This series of webinar provides a general introduction to Bayesian modeling with a particular focus on regression and multilevel models. The use of the system R in Bayesian computation is described, including the programming of the Bayesian model and the use of different R tools to summarize the posterior. Special focus will be on the application of Markov chain Monte Carlo (MCMC) algorithms and diagnostic methods to assess convergence of the algorithms. The LearnBayes and rethinking R packages are used to illustrate MCMC fitting by the use of Gibbs sampling and Metropolis algorithms. Larger Bayesian models will be fit using JAGS and Stan and the accompanying runjags and rstan packages.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: This course is presented by the ISI Statistical Capacity Building (SCB) Committee. It is available for free to everyone. The course includes an introduction to (descriptive) statistics, and modules on sampling, probability, statistical inference, experimental design, categorical data, non-parametric methods, and linear regression. We thank Ann Maharaj (Australia), Delia North and her three collaborators (South Africa), and Edward Boone (United States) for their efforts in developing it. Capacity building is a strategic priority for ISI, and the SCB Committee has been very active in organising courses and favouring participation in events, sometimes in cooperation with other organisations. Now the SCB Committee is working to provide online courses and webinars, starting with this course. The Committee is also looking into providing sessions devoted to Questions and Answers. More information will be provided as soon as we have it available. Modules: (1) Introduction to Statistics and Descriptive Statistics; (2) Sampling; (3) Probability; (4) Statistical Inference for One Population; (5) Statistical Inference for Two Populations; (6) Experimental Design and Analysis of Variance; (7) Analysis of Categorical Data; (8) Non-Parametric Methods; (9) Simple Linear Regression; (10) Multiple Linear Regression
Description: This course on applied econometrics, has been prepared as an integration of mathematical economics and statistical methods with an objective of extracting the stochastic relationship among variables and their modelling. Learning outcomes to be covered The learning outcomes will be To get the knowledge of Linear and nonlinear modelling How to improve the estimation of regression coefficients by various restrictions. Knowledge of Restricted least squares estimation, Generalized and Weighted Least Squares Estimation. Knowledge about the difference between disturbances and measurement errors: Indicator variables versus quantitative explanatory variable Tests for Structural Change and Stability
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: This series of webinar provides a general introduction to Bayesian modeling with a particular focus on regression and multilevel models. The use of the system R in Bayesian computation is described, including the programming of the Bayesian model and the use of different R tools to summarize the posterior. Special focus will be on the application of Markov chain Monte Carlo (MCMC) algorithms and diagnostic methods to assess convergence of the algorithms. The LearnBayes and rethinking R packages are used to illustrate MCMC fitting by the use of Gibbs sampling and Metropolis algorithms. Larger Bayesian models will be fit using JAGS and Stan and the accompanying runjags and rstan packages. Part 1 (2 hours): Introduction to Bayesian Inference. Basic tenets of Bayesian thinking including construction of priors, summarization of the posterior to perform inferences, and the use of prediction distributions for prediction and model checking. Part 2 (2 hours): Bayesian Regression. Implementation of Bayesian thinking for regression models for continuous and categorical response data. Part 3 (2 hours): Bayesian Multilevel Modeling. Introduction to multilevel models as a flexible way of modeling regressions over groups.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: The fields of randomized experiments and probability sampling are traditionally two separated domains of applied statistics. Both fields share one similarity, which makes them unique from other areas of statistics: design plays a crucial role in this type of empirical research. While design of randomized experiments is traditionally focused on establishing the causality between differences in treatments and observed effects (internal validity), design of probability samples is focused on generalizing results observed in a small sample to an intended target population (external validity). Many experiments conducted with the purpose to improve survey methods are small scaled or conducted with specific groups. The value of empirical research into survey methods is strengthened as conclusions can be generalized to populations larger than the sample that is included in the experiment. This can be achieved by selecting experimental units randomly from a larger target population and naturally leads to randomized experiments embedded in probability samples. This results in experiments that potentially combine the strong internal validity from randomized experiments with the strong external validity of probability sampling. Generalizing conclusions observed in an experiment to a larger target population can be achieved with a design-based inference framework known from sampling theory. In this webinar a general design-based framework for the analysis of single factor and factorial randomized experimental designs embedded in general complex probability samples is presented. Methods are illustrated with real life applications conducted at Statistics Netherlands.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: Within the framework of its Statistical Capacity Building (StatCaB) Programme, SESRIC will organise an Online Training Course on ‘Survey Methods and Sampling’ for the benefit of National Statistical Offices (NSOs) of OIC countries on 23 – 26 May 2022. Mr. Cenker Burak Metin, Head of Survey and Sampling Design Group at the Turkish Statistical Institute (TurkStat) will conduct the course and cover the following topics: Definition of a statistical survey and basic concepts Defining the survey objectives, variables and concepts Survey planning and steps of a survey Data collection methods PAPI, CATI, CAPI and other methods Introduction to Sampling Design Implementation of Sampling Techniques Estimation theory for sample surveys The course will be conducted through a video conferencing platform by following synchronous learning and instruction approaches designed in line with the virtual training solutions undertaken by SESRIC in order to better serve the Centre’s training activities and keep participants motivated and engaged during this time of global crisis due to COVID-19. For more information on SESRIC Statistical Capacity Building (StatCaB) Programme, please visit: http://www.oicstatcom.org/statcab.php
Title in Arabic: تصميم العينات ومنهجيات المسوح في الاحصاءات الرسمية
Organizer(s): AITRS
Description: مواصلة للدورة التدريبية التي نظمها المعهد خلال الفترة 15 فبراير / شباط - 13 نيسان / ابريل 2021 حول مجال العينات بواقع 45 ساعة واعتبارا لأهمية هذا المجال لما له دور أساسي في تخطيط وبرمجة واعداد المسوح الإحصائية والمشاكل التي تواجه عملية الاستقصاء وجمع البيانات الإحصائية وأثر العينة على نتائج تلك المسوح والية اختيار المناسب منها، سينظم المعهد دورة تدريبية متقدمة حول تصميم العينات ومنهجيات المسوح في الاحصاءات الرسمية بالاضافة الى استعراض المستجدات التي طرأت في علم العينات واستخدام التقنيات الجديدة والبرمجيات الحديثة. ,تهدف هذه الدورة الى: * ترسيخ المفاهيم الضرورية لتصميم عينات المسوح الإحصائية * اعداد وتجهيز اطر المعاينة وسحب العينات منها * حساب الاوزان وتعديلها ومعايرتها * حساب التباين والدقة في التقديرات الناتجة من المسوح بالعينة * استخدام البرامج الحديثة في سحب العينات وحساب التباين * ضبط ومراقبة الجودة في كل مراحل العمل
Description: In recent years, a range of methods have been developed for handling unusual data sources – Big Data and non-probability samples –to enable the production of public and official statistics. The core methods available are quasi-randomization, superpopulation modelling and doubly-robust estimation. They rely on the use of generalized linear models and aim to produce estimates with reliability like that of estimates from traditional probability samples of similar sizes. Quasi-randomization involves using a probability sample survey as reference to estimate pseudo-weights for units in a non-probability sample or big data-type source, where coverage of the target population is insufficient or unknown. We present a brief review of the available methods and an application in which quasi-randomization was used successfully to make inference from a web-panel survey carried out by CETIC.br.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Organizer(s): ISI Baltic-Nordic-Ukrainian Network on Survey Statistics
Description: The Summer School on Survey Statistics 2021 is fully virtual and offers educational sessions in English on Friday 3, 10, 17 and 24 September at 15-18 and sessions in Russian on Saturday 4, 11, 18 and 25 September (exact times TBA). Detailed information can be accessed here. The summer school is free of participation fee and is open for anyone interested. Registration required. Information on registration and contributed paper submission can be found on the event web site. The main aim of the summer school is to promote scientific and educational cooperation in survey and official statistics between statisticians interested in new trends in the area. Educational sessions include keynote lectures, invited lectures and contributed papers. Main topics are Data integration, Machine Learning and Small area estimation. Topics related to survey and official statistics are welcome in contributed papers. The summer school is organized by the Baltic-Nordic-Ukrainian (BNU) Network on Survey Statistics. Today, the network involves partners from eight countries: Belarus, Estonia, Finland, Latvia, Lithuania, Poland, Sweden and Ukraine. The Summer School on Survey Statistics 2021 is the 24th of the scientific or educational events organized by the network since 1997. The event is sponsored by the International Association of Survey Statisticians (IASS).
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: This course will provide a comprehensive introduction to cure models, including basics of the cure models as well as many recent developments related to the methodological issues and software implementations of cure models for right-censored time-to-event data subject to non-informative censoring. The course will feature real-world examples from clinical trials and population-based studies and a detailed introduction to R packages, SAS macros, and WinBUGS programs to fit some cure models. Applications of cure models in other disciplines will be discussed. This course will be useful for statistical researchers and graduate students, and practitioners in other disciplines to have a thorough review of modern cure model methodology and to seek appropriate cure models in applications.
Target Audience: This course should appeal to a broad audience, including statisticians and graduate students in statistics/biostatistics as well as clinician-scientists, health researchers, health policymakers, and researchers in cancer research and the biopharmaceutical industry who have a good understanding of the basics of biostatistics. Graduate students in epidemiology, public health, and management who have strong training in biostatistics can also be benefited from this course.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: We begin with a graphical approach to bootstrapping and permutation testing, illuminating basic statistical concepts of standard errors, confidence intervals, p-values and significance tests. We consider a variety of statistics (mean, trimmed mean, regression, etc.), and a number of sampling situations (one-sample, two-sample, stratified, finite-population), stressing the common techniques that apply in these situations. We’ll look at applications from a variety of fields, including telecommunications, finance, and biopharm. These methods let us do confidence intervals and hypothesis tests when formulas are not available, so we can do better statistics, e.g. use robust methods like medians, trimmed means, or robust regression. They can help clients understand statistical variability. And some of the methods are more accurate than standard methods.
Target Audience: Practicing statisticians and students. This is not a highly technical course.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: Large and innovative spatial data are now ubiquitous across science and engineering ranging from the microscale properties of 3D printed materials to the exposure of populations to pollutants to the global views of our planet from satellites. The challenge to statistical science is to adapt methods from geostatistics to these new problems. Large data sets break traditional spatial methods and multivariate spatial data are not well modeled by classical approaches. This course will provide a hands-on and modern introduction to spatial data, followed by methods for large and nonstationary data and models for multivariate processes. It will be taught by active researchers in this area who have contributed to theory, new methods, and maintain software that makes spatial data analysis easy and accessible.
Target Audience: We seek to reach graduate students who are interested in challenging data problems to motivate new research topics or to enlarge their toolbox of statistical methods. This course is also a cogent overview for the faculty member who is always been curious what spatial statistics is all about. Finally, this course is for data scientists who would like to see the use of statistics for drawing inferences and quantifying uncertainty for spatial data as opposed to just forming prediction.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: This course will focus on methodological and practical issues in the scope of competing risks and multi-state models. Basic concepts, models, estimation algorithms and statistical software will be reviewed. Simulation exercises and real data analyses will be provided in order to enlighten the interpretation and to facilitate the understanding. Both nonparametric methods and semiparametric approaches will be considered. Some background in foundations of statistics and in parametric and nonparametric inference is required. Background in Survival Analysis (Kaplan-Meier estimation, Cox regression) is recommended. BSc or MSc in Statistics, Biostatistics or Mathematics is ideal.
Target Audience: PhD students and postgraduates in general.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: Finite population sampling has found numerous applications in the past century. The validity of sampling inference of real populations derives from the known probability sampling design under which the sample is selected, “irrespectively of the unknown properties of the target population studied” (Neyman, 1934). This is the key theoretical justification for its universal applicability. Valued graph is a more powerful representation, which allows one to incorporate the connections among the population units in addition to the units on their own. The underlying structure is a graph given as a finite collection of nodes (units) and edges (connections). Attaching measures to the nodes and/or edges yields a valued graph. Many technological, socio-economic, biological phenomena exhibit a graph structure that may be the central interest of study, or the edges may provide effectively access to those nodes that are the primary targets. Either way, graph sampling is a statistical approach to study real graphs. Just like finite population sampling, it is universally applicable based on exploring the variation over all possible subgraphs (i.e. sample graphs), which can be taken from the given population graph, according to a specified method of sampling. Graph sampling thus encompasses finite population sampling, because any latter situation can be represented as a special case of the former. All the so-called “unconventional” finite population sampling techniques, such as indirect, network, adaptive cluster or line-intercept sampling, can be more effectively studied as special cases of graph sampling. Whereas snowball sampling and targeted random walk sampling are probabilistic versions of breadth- or depth-first non-exhaustive search algorithms in graphs. The course provides an introduction to the central concepts of graph sampling, the most common sampling methods, and the construction of graph sampling strategy. An emphasis is the extension from the traditional sampling strategy (finite population sampling, Horvitz-Thompson estimator) to a much more general strategy consisting of bipartite incidence graph sampling (BIGS) and incidence weighting estimator (IWE). The application of the BIGS-IWE strategy will be illustrated for all the aforementioned unconventional situations of finite population sampling, as well as the more complicated graph sampling situations such as snowball sampling and targeted random walk sampling.
Target Audience: Graduate students, statisticians at national statistical offices or other organisations working with sampling methods, data scientists interested in network analysis, graph mining or compression.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
Description: Fraud has been around since the early days of commerce, continuously evolving and adapting to changing times. The fraudulent cases are seen in a wide range of domains such as finance, credit card, telecommunications, insurance and health care. Examples include but not limited to the post COVID-19 instances in financial stimulus, unemployment eligibility and health care procurement. For instance, in health care, overpayments are estimated to correspond up to ten percent of total expenditures. This short course presents the use of analytical methods for fraud assessment. Fraud data and its types will be introduced with some examples and pre-processing techniques. Next, the course will cover the use of visualization and unsupervised methods (outlier detection, clustering, topic models) to describe data and reveal hidden relationships. Whereas supervised methods such as classification and regression can be used with labeled data sets for prediction purposes. These methods will be discussed using examples from finance and health care industries. The course will conclude with an overview of applications using R. After completing the course, the attendees will have learnt various types of fraud, and the use of data and statistical methods for fraud detection.
Target Audience: Researchers in financial service companies, banks, insurance companies, government institutions, health care institutions, and consulting firms as well as fraud data analysts/scientists; consultants working in fraud detection. This course is also expected to be of interest to early career statisticians that can gain insights about how different data mining/statistical methods are applied in this emerging crucial subject domain using R.
Topics:
Original webpage was deleted, archived version from the Internet Archive (not a UN service): Link
