Search

BASIE is a framework for interpreting impact estimates from evaluations. It is an alternative to null hypothesis significance testing. This guide walks researchers through the key steps of applying BASIE, including selecting prior evidence, reporting impact estimates, interpreting impact estimates, and conducting sensitivity analyses. The guide also provides conceptual and technical details for evaluation methodologists.

A valuable extension of the single-rating regression discontinuity design (RDD) is a multiple-rating RDD (MRRDD). To date, four main methods have been used to estimate average treatment effects at the multiple treatment frontiers of an MRRDD: the "surface" method, the "frontier" method, the "binding-score" method, and the "fuzzy instrumental variables" method. This paper uses a series of simulations to evaluate the relative performance of each of these four methods under a variety of different data-generating models. Focusing on a two-rating RDD (2RRDD), the methods are compared in terms of their bias, precision, and mean squared error when implemented as they most likely would be in practice--using optimal bandwidth selection. The lessons learned from the simulations to a real-world example that uses data from a study of an English learner reclassification policy are also applied. Overall, this paper makes valuable contributions to the literature on MRRDDs in that it makes concrete recommendations for choosing among MRRDD estimation methods, for implementing any chosen method using local linear regression, and for providing accurate statistical inferences. Appended are: (1) Computing Average Frontier Effects for Models 3 and 4 in Our Simulations; (2) Estimation in Theory: Using Full Information; (3) Deriving Correct Model Specifications for the Fuzzy IV Method; (4) Domain or Bandwidth Selection; and (5) Simulation Results When Using Local Linear Regression, Not Including Other Rating as a Covariate.

This "What Works Clearinghouse Procedures and Standards Handbook (Version 3.0)" provides a detailed description of the standards and procedures of the What Works Clearinghouse (WWC). The remaining chapters of this Handbook are organized to take the reader through the basic steps that the WWC uses to develop a review protocol, identify the relevant literature, assess research quality, and summarize evidence of effectiveness. Organizational procedures used by the WWC to ensure an independent, systematic, and objective review are described in the appendices. Table I.1 provides a summary of the remaining chapters and associated appendices. The main differences between this version of the procedures and standards and the previous version (Version 2.1) are in clarity, detail, and scope. The organization of the Handbook, as well as all text, was reviewed and modified to support clarity; additionally, examples have been added throughout. There is more detail on the specific procedures and standards used by the WWC, including how to deal with missing data, random assignment probabilities, and cluster-level designs. Finally, whereas the previous version focused almost exclusively on intervention reports, this version provides information on other key WWC products, which include practice guides, single study reviews, and quick reviews. As the WWC continues to refine processes, develop new standards, and create new products, the "What Works Clearinghouse Procedures and Standards Handbook" will be revised to reflect these changes. Appended are: (1) Staffing, Reviewer Certification, and Quality Assurance; (2) Policies for Searching and Prioritizing Studies for Review; (3) The WWC Study Review Process; (4) Pilot Regression Discontinuity Design Standards; (5) Pilot Single-Case Design Standards; (6) Magnitude of Findings for Randomized Controlled Trials and Quasi-Experimental Designs; and (7) Statistical Significance for Randomized Controlled Trials and Quasi Experimental Designs. [See ED503772 to view previous version of this guide.]

This report examines the differences in impact estimates and standard errors that arise when these are derived using state achievement tests only (as pre-tests and post-tests), study-administered tests only, or some combination of state- and study-administered tests. State tests may yield different evaluation results relative to a test that is selected, and administered, by the research team for several reasons. For instance, (1) because state tests vary in content and emphasis, they also can vary in their coverage of the types of knowledge and skills targeted by any given intervention. In contrast, a study-administered test will correspond to the intervention being evaluated. In addition to differences in alignment with treatment, state tests may yield divergent evaluation results due to differences in (2) the stakes associated with the test, (3) missing data, (4) the timing of the tests, (5) reliability or measurement error, and (6) alignment between pre-test and post-test. Olsen, Unlu, Jaciw, and Price (2011) discuss how these six factors may differ between state- and study-administered tests to influence the findings from an impact evaluation. Specifically, Olsen et al. use data from three single-state, small-scale evaluations of reading interventions that collected outcomes data using both study-administered and state achievement tests to examine this and other issues. The authors found that (1) impact estimates based on study-administered tests had smaller standard errors than impact estimates based on state tests, (2) impacts estimates from models with "mismatched" pre-tests (e.g., a state pre-test used in combination with a study-administered post-test) had larger standard errors than impact estimates from models with matched pre-tests, and (3) impact estimates from models that included a second pre-test covariate had smaller standard errors than impact estimates from models that included a single pre-test covariate. Study authors caution that their results may not generalize to evaluations conducted in other states, with different study-administered tests, or with other student samples. Appended are: (1) Description of the Three Experiments; (2) Scatter Plots of Student Test Scores; (3) Quartiles of the Test Score Distribution; (4) Estimates from Other Evaluations; (5) Estimates from the Full Sample; (6) Hypothesis Tests and Minimum Detectable Differences; (7) Conceptual Approach to Generating Correlated Residuals for the Parametric Bootstrap; (8) Results from Bootstrapping and Hypothesis Testing; (9) Differences in Sample Size Requirements; (10) Correlations between Scores on State and Study-Administered Tests; and (11) Estimates of Key Statistical Power Parameters. (Contains 37 tables, 3 figures and 45 footnotes.)

Understanding the concept of the sampling distribution of a statistic is essential for the understanding of inferential procedures. Unfortunately, this topic proves to be a stumbling block for students in introductory statistics classes. In efforts to aid students in their understanding of this concept, alternatives to a lecture-based mode of instruction have been introduced in the literature with some of these approaches using in-class activities, simulations using statistical software, and web-based applets. In this article, the use of statistical software, for the purpose of illustrating sampling distributions, is revisited through the use of Minitab macros, an approach that has not been observed in the literature. The result is a user-friendly way for students in introductory statistics classes to explore the concept of the sampling distribution of a statistic. Although the focus of this article will be on the sampling distribution, the methods described here are applicable to instruction of other statistical concepts. including confidence intervals and power. (Contains 4 figures.)

Possible relationships between five test examiner characteristics (gender, race, tenure, experience as a test administrator, and experience as a test developer or scorer) and six student achievement scores (reading, writing, language usage, mathematics, science, and social studies) were studied at the school level in a statewide assessment. The school-level results were aggregated using meta-analysis to explore the plausibility of examiner variables as threats to test validity. Very few of the average correlations across schools were statistically significant, and for all of them, even for those that were statistically significant, confidence intervals for the correlations were extremely small at both ends. Significant heterogeneity of effect sizes was found for virtually all of the 60 analyses, suggesting that further exploration is needed. Some directions for further research are discussed. (Contains 10 figures and 10 tables.) [This research was carried out by the Maryland Assessment Research Center for Education Success.]

Bootstrap analysis, both for nonparametric statistical inference and for describing sample results stability and replicability, has been gaining prominence among quantitative researchers in educational and psychological research. Procedurally, however, it is often quite a challenge for quantitative researchers to implement bootstrap analysis in their research because bootstrap analysis is typically not an automated program option in statistical software programs. This paper uses a few heuristic analytical examples to show how bootstrap analysis can be accomplished through the use of some commonly available statistical software programs. Until bootstrap analysis becomes an automated program option in standard statistical software programs (e.g., the Statistical Package for the Social Sciences or the Statistical Analysis System), quantitative researchers may have to make do with these or other creative approaches to accomplish bootstrap analysis in their research. (Contains 4 tables, 10 figures, and 37 references.) (Author/SLD)

This collection of activities is designed to show how graphing calculators can be used to explore statistics. The activities address such topics as data representation, distributions, and statistical tests. Teaching notes and calculator instructions are included as are blackline masters. (MM)

This working paper contains the overheads used in a seminar designed to introduce some basic concepts of statistics to nonstatisticians. The seminar has been presented on several occasions. The first part of the seminar, and the first set of overheads, deals with the essentials of statistics, including: (1) population, sample, and inference; (2) standard errors and confidence intervals; (3) statistical significance; (4) correlation and linear regression; and (5) graphics. The second part of the seminar, and the second group of overheads, concerns basic principles of research design and analysis, including operationalizing terms, types of bias, confounding, and aspects of validity and reliability. (Contains 91 overheads.) (SLD)

Bayesian methods for estimating dose response curves in quantal bioassay are studied. A linearized multi-stage model is assumed for the shape of the curves. A Gibbs sampling approach with data augmentation is employed to compute the Bayes estimates. In addition, estimation of the "relative additional risk" and the "risk specific dose" is studied. Model selection based on conditional predictive ordinates from cross-validated data is developed. Contains 22 references and 8 tables.) (Author)

Describes the logic of statistical testing in the Fisher and Neyman-Pearson approaches. Reviews some common misinterpretations of basic concepts behind statistical tests. Analyzes the philosophical and psychological issues that can contribute to these misinterpretations. Suggests possible ways in which statistical education might contribute to the better understanding and application of statistical inference. (Contains 53 references.) (Author/ASK)

This advance report based on the March 1988 supplement to the Current Population Survey (CPS) presents preliminary data on selected demographic, social, and economic characteristics of the Hispanic population of the United States. The Hispanic civilian noninstitutional population in March 1988 totaled about 19.4 million, a 34% increase since 1980; the non-Hispanic population increased 7% in that period. About half of Hispanic growth resulted from net migration and half from natural increase. The proportions of Hispanics completing 4 years of high school or more, and completing 4 or more years of college reached 51% and 10%, respectively, both records. About 55% of Hispanics resided in California and Texas. Married couple families decreased by 1988 to 70% from 74% in 1982. The unemployment rate among Hispanics 16 and over was 8.5%, its lowest level since the survey of March 1983, shortly after the end of the last recession. The poverty rate was 25.8%, little changed since 1982. The origins of Hispanic Americans were 62% Mexican, 13% Puerto Rican, 5% Cuban, and 12% Central or South American. These subgroups varied considerably in educational attainment, family composition, employment, and median family income. The report includes five graphs and four tables of selected social and economic characteristics, by type of Hispanic origin and by year from 1982 to 1988. Appendices discuss CPS data source, estimation procedure, and reliability of estimates, and they contain facsimiles of March 1988 CPS questionnaires. (SV)

This report based on the March 1985 supplement to the Current Population Survey presents demographic, social, and economic data on the Hispanic population in the United States, and focuses on social and economic changes between 1982 and 1985. The Hispanic civilian noninstitutional population in March 1985 totaled about 17 million, a 17% increase since 1980; the non-Hispanic population increased 4% in that period. Between 1982 and 1985: (1) the median age of Hispanics rose from 23.9 to 25.0 years; (2) the proportion of Hispanics 25 and over with at least a high school diploma increased from 45% to 48%; (3) the proportion of married couple families fell from 74% to 72%; (4) the unemployment rate for Hispanics 16 and over fell from 13.4% to 11.3%; and (5) there was no change in the real median income of Hispanics, while that of non-Hispanics rose 3.5%. The origins of Hispanic Americans were about 61% Mexican, 15% Puerto Rican, 6% Cuban, and 10% Central or South American. These subgroups varied considerably in educational attainment, family size, family composition, employment, and median family income. The report includes 3 graphs and 27 detailed tables of data on demographic, social, and economic characteristics. Appendices cover the methodology for development of independent post-census estimates (component estimation technique), summary tables of selected characteristics for the years 1982 to 1985, definitions, and data source and reliability. (SV)

This report presents data from the March 1988 Current Population Survey on the demographic, social, and economic characteristics of the Hispanic population of the United States, including age, sex, marital status, educational attainment, school enrollment, fertility, voting and registration, employment status, family composition and size, income, and poverty status. Data on educational attainment show 11.7% of Hispanic males, age 25 and over, completing less than 5 years of school compared to 2.6% of the total population. Fifty-two percent of Hispanic males completed 4 years of high school or more compared to 76.4% of the total population. Only 12% of Hispanic males completed 4 years of college or more, while 24% of the total population attained this much education. Relative percentages for females were very similar. Tables also show years of school completed by age, sex, and type of Hispanic origin and current educational enrollment. For those of Hispanic origin of both sexes, ages 3 to 34 years, 47.8% were enrolled in school in October, 1986, compared to 48.2% for those of non-Hispanic origin. The report includes five charts and 36 tables. Appendices discuss definitions, data sources, and the accuracy of estimates. (DHP)

Discusses the extent to which the existing literature offers insights into effectively teaching statistical inference. Describes an in-class exercise that helps students understand inferential statistics. Explains that students learn more about concepts such as population distribution, sampling distributions, and standard error of the estimate. (CMK)

Presents a problem that describes a real-world situation from World War II that tries to estimate the number of German tanks, half-tracks, and aircraft, and a simulation of that problem in which the selection of a suitable estimate is less apparent. (ASK)

The study of statistics can be enhanced by using real-world data and problems obtained from the Internet. Uses data on HIV infections in Asian and Pacific Islander females to teach both descriptive and inferential statistics. (ASK)

Proposes guidelines for revising the American Psychological Association (APA) publication manual or other APA materials to clarify the application of statistics in research reports. The guidelines are intended to induce authors and editors to recognize the thoughtless application of statistical methods. Contains 54 references. (SLD)

Suggests ways for using data from championship tennis as a means for exploring probabilistic models, especially binomial probability. Examples include the probability of winning a service point and the probability of winning a service game using data from tables and graphs. (AIM)

Presents two mathematical methods for fitting the logistic curve to population data supplied by the U.S. Census Bureau utilizing computer algebra software to carry out the computations and plot graphs. (JKK)

Examines potential problems in the use of inferential statistics for single population proportions, differences between two population proportions, and quotients of two population proportions. Discusses hypothesis testing versus interval estimation. Emphasizes the importance of selecting the appropriate formula for the standard error and questioning the assumption of independent observations. Suggests cautionary procedures for educational researchers. Contains 21 references. (Author/SV)

Suggests activities using large samples to simulate quality control processes in industry (specifically, manufacturing camera film). A series of small-group activities investigate the relationship of sample size to percent defective in batch sampling. Efficiency, cost of sampling, company reputation, and public relations are business factors considered. (RC)

[Objective] The main purpose of this study was to analyze the activities related to statistical inference present in secondary education textbooks in Chile, and their relationship to the development of statistical thinking. [Methodology] For the purposes of the study, a qualitative approach was used through a content analysis of the secondary education books disseminated free of charge by the Chilean Ministry of Education during the years 2016, 2018, 2020, 2021 and 2022, which were selected using intentional non-probabilistic sampling. [Results] Among the main results obtained, it was found that most of the activities related to statistical inference in the textbooks analyzed had to do with procedures related to the calculation of confidence intervals, without addressing their interpretation in the context of the problem. In addition, it was observed that there were few activities related to informal inference. Likewise, the activities and questions presented in the textbooks do not refer to the formulation of hypotheses or conjectures about the sample data made to guide the formulation of conclusions related to statistical inference. [Conclusions] The findings indicate that despite the fact that confidence intervals are among the topics related to statistical inference in the textbooks, these textbooks are not focused on decision-making at the population level, but rather on description., [Objetivo] O objetivo principal deste estudo foi analisar as atividades relacionadas à inferência estatística, presentes textos escolares do ensino médio no Chile, e sua relação com o desenvolvimento do pensamento estatístico. [Metodologia] Para efeitos do estudo, utilizou-se uma abordagem qualitativa através de uma análise de conteúdo dos livros do ensino secundário distribuídos gratuitamente pelo Ministério da Educação do Chile durante os anos de 2016, 2018, 2020, 2021 e 2022, que foram selecionados por meio de amostragem não probabilística do tipo intencional. [Resultados] Dentre os principais resultados obtidos, a maior parte das atividades relacionadas à inferência estatística nos livros analisados são do tipo processual, relacionadas ao cálculo de intervalos de confiança, sem abordar sua interpretação no contexto da situação problema. Além disso, há uma presença escassa de atividades relacionadas à inferência informal. Da mesma forma, as atividades e questões explicadas nos livros de texto não se referem à formulação de hipóteses ou conjecturas sobre os dados amostrais, de modo que a formulação de conclusões seja orientada para a inferência estatística. [Conclusões] Os resultados fornecem evidências de que, embora os intervalos de confiança estejam dentro dos tópicos relacionados à inferência estatística nos livros de texto, eles não se concentram na tomada de decisões em nível populacional, mas são tratados a partir de um vértice descritivo., [Objetivo] El presente estudio tuvo como objetivo principal analizar las actividades relacionadas con inferencia estadística, presentes en los textos escolares de enseñanza secundaria en Chile, y su relación con el desarrollo del pensamiento estadístico. [Metodología] Para efectos del estudio se hizo uso de un enfoque cualitativo mediante un análisis de contenido de los libros de enseñanza secundaria difundidos, gratuitamente, por el Ministerio de Educación chileno durante los años 2016, 2018, 2020, 2021 y 2022, los cuales fueron seleccionados mediante un muestreo no probabilístico del tipo intencionado. [Resultados] Entre los principales resultados obtenidos, la mayoría de las actividades afines a inferencia estadística en los libros analizados, son del tipo procedimental, relacionadas con el cálculo de intervalos de confianza, sin abordar la interpretación de estos en el contexto de la situación problema. Además, se observa una escasa presencia de actividades relacionadas con inferencia informal. Asimismo, las actividades e interpelaciones explicitadas en los libros de texto no hacen referencia a la formulación de hipótesis o conjeturas acerca de los datos muestrales, de manera que se oriente a la formulación de conclusiones hacia la inferencia estadística. [Conclusiones] Los hallazgos entregan evidencias que, a pesar de que los intervalos de confianza se encuentran dentro de las temáticas afines a inferencia estadística en los libros de texto, estas no se focalizan en la toma de decisiones en el nivel poblacional, más bien son tratados desde un vértice descriptivo.

Hospital’s registry data is a widely used resource in Nordic countries to estimate parameters of interest in the public health and in epidemiology. This data allows the researcher to have an unbiased representation of the population, since it is collected for all the individuals that visit the hospitals in the country, and it is stored in databases that are available for the researchers. Despite being a powerful tool, this data has some drawbacks that have to be taken into account before making the study. This project aims to expose the problems that arise when hospital’s registers are used for estimating the incidence of a disease, and explain what it is usually done to correct (or partially correct) them. First, we provide an introduction (Chapter 1) where we discuss the importance of getting good estimations to study the incidence of mental disorders, why registers are a powerful tool to make population-wide research and the main problem we encounter in this type of data: the delayed entries. We continue with a summary on the methodologies that will be used during the rest of the project (Chapter 2). In Chapter 3 we develop further the strengths and limitations of using registration data to estimate the incidence of a disease, providing a self-derived theoretical description of the delayed-entries problematic. In this chapter we also include the methodology that it is usually applied to deal with this problematic: the washout-period method, and we relate this methodology with the provided theoretical description. We finish the chapter with an introduction to the history of the registration system in Denmark and its structure. Lastly, we provide a simulation study based on the data of women diagnosed with depression in Denmark. We have simulated a hospital register and studied the incidence in terms of the cumulative incidence function applying the washout period method.

Parameter estimation is a powerful and adaptable framework that addresses the inherent complexities and uncertainties of financial data. We provide an overview of likelihood functions, and likelihood estimations, as well as the essential numerical approximations and techniques. In the financial domain, where unpredictable and non-stationary market dynamics prevail, parameter estimations of relevant SDE models prove highly relevant. We delve into practical applications, showcasing how SDEs can effectively capture the inherent uncertainties and dynamics of financial models related to time evolution of interest rates. We work with the Vašíček model and Cox-Ingersoll-Ross (CIR) model which describes the dynamics of interest rates over time. We incorporate the Maximum likelihood and Quasi-maximum likelihood estimation methods in estimating the parameters of our models.

Likelihood-based approaches are central in statistics and its applications, yet often challenging since likelihoods can be intractable. Many methods such as the EM algorithm have been developed to alleviate this. We present a new likelihood inequality involving posterior distributions of a latent variable that holds under conditions similar to the EM algorithm. Potential scopes of the inequality includes maximum-likelihood estimation, likelihood ratios tests and model selection. We demonstrate the latter by performing selection in a non-linear mixed-model using MCMC.

[ESP] El Mar Menor es la laguna costera hipersalina más grande de Europa. La eutrofización es el proceso de contaminación más importante de las aguas de lagos, estanques, ríos y embalses. Este proceso es causado por el exceso de nutrientes en el agua, principalmente nitrógeno y fósforo, provenientes principalmente de la actividad humana. En este trabajo presentamos un análisis cuantitativo de contaminantes en las aguas del Mar Menor en diferentes playas de la laguna. Las muestras se analizaron con el sistema de cromatografía iónica. [ENG] The Mar Menor is the largest hypersaline coastal lagoon in Europe. Eutrophication is the most important contamination process of the waters of lakes, ponds, rivers, and reservoirs. This process is caused by the excess of nutrients in the water, mainly nitrogen and phosphorus, coming mainly from human activity. In this work we present a quantitative analysis of contaminants in the waters of the Mar Menor in different beaches of the lagoon. Samples were analyzed on the Metrohm 861 ion chromatography system.

This thesis focuses on predicting trends in Commodity Trading Advisors (CTAs), also known as trend-following hedge funds. The paper applies a Hidden Markov Model (HMM) for classifying trends. Additionally, by incorporating additional features, a regularized logistic regression model is used to enhance prediction capability. The model demonstrates success in identifying positive trends in CTA funds, with particular emphasis on precision and risk-adjusted return metrics. In the context of regularized regression models, techniques for statistical inference such as bootstrap resampling and Markov Chain Monte Carlo are applied to estimate the distribution of parameters. The findings suggest the model's effectiveness in predicting favorable CTA performance and mitigating equity market drawdowns. For future research, it is recommended to explore alternative classification models and extend the methodology to different markets and datasets.

Statistics serves as the grammar of all science, and central to the goal of science is understanding cause-effect relationships. Scientists rely on research methodology and statistical tools to uncover causal relationships, and engineers rely on statistical methods to create artificial assistants to aid daily life. Neither statistical learning nor next-word-prediction (used to train artificial general intelligence) are consistent with rational causal learning and reasoning in humans. The present thesis examines the fundamental goals and assumptions made in dominant statistical methods and discusses their implications for statistical inference and commonsense reasoning in artificial general intelligence (AGI). The first section introduces and evaluates a causal alternative to logistic regression, which estimates the causal power (from the causal invariance framework) of treatments among covariates. Causal invariance is defined as the influence of a candidate cause (elemental or conjunctive) that is independent of background causes, with the aspiration of acquiring knowledge that’s useable, in the minimalist sense being able to generalize from a learning context to an application context. The second and final section investigates current benchmark tasks used to evaluate causal reasoning in large language models (e.g., GPT-3, GPT-4), and introduces a stricter test informed by psychological literature on human causal cognition under the causal invariance framework.

Modern radar systems must operate in a wide variety of time-varying conditions. These include various types of interference from neighboring systems, self-interference or clutter, and targets with fluctuating responses. It has been well-established that the quality and nature of radar measurements depend heavily on the choice of signal transmitted by the radar. In this dissertation, we discuss techniques which may be used to adapt the radar's waveform on-the-fly while making very few a priori assumptions about the physical environment. By employing tools from reinforcement learning and online learning, we present a variety of algorithms which handle practical issues of the waveform selection problem that have been left open by previous works. In general, we focus on two key challenges inherent to the waveform selection problem, sample-efficiency and universality. Sample-efficiency corresponds to the number of experiences a learning algorithm requires to achieve desirable performance. Universality refers to the learning algorithm's ability to achieve desirable performance across a wide range of physical environments. Specifically, we develop a contextual bandit-based approach to vastly improve the sample-efficiency of learning compared to previous works. We then improve the generalization performance of this model by developing a Bayesian meta-learning technique. To handle the problem of universality, we develop a learning algorithm which is asymptotically optimal in any Markov environment having finite memory length. Finally, we compare the performance of learning-based waveform selection to fixed rule-based waveform selection strategies for the scenarios of dynamic spectrum access and multiple-target tracking. We draw conclusions as to when learning-based approaches are expected to significantly outperform rule-based strategies, as well as the converse.

In the January 2022 issue of Perspectives, Götz et al. argued that small effects are “the indispensable foundation for a cumulative psychological science.” They supported their argument by claiming that (a) psychology, like genetics, consists of complex phenomena explained by additive small effects; (b) psychological-research culture rewards large effects, which means small effects are being ignored; and (c) small effects become meaningful at scale and over time. We rebut these claims with three objections: First, the analogy between genetics and psychology is misleading; second, p values are the main currency for publication in psychology, meaning that any biases in the literature are (currently) caused by pressure to publish statistically significant results and not large effects; and third, claims regarding small effects as important and consequential must be supported by empirical evidence or, at least, a falsifiable line of reasoning. If accepted uncritically, we believe the arguments of Götz et al. could be used as a blanket justification for the importance of any and all “small” effects, thereby undermining best practices in effect-size interpretation. We end with guidance on evaluating effect sizes in relative, not absolute, terms.

The Hepatitis C virus (HCV) envelope glycoprotein E1 forms a non-covalent heterodimer with E2, the main target of neutralizing antibodies. How E1-E2 interactions influence viral fitness and contribute to resistance to E2-specific antibodies remain largely unknown. We investigate this problem using a combination of fitness landscape and evolutionary modeling. Our analysis indicates that E1 and E2 proteins collectively mediate viral fitness and suggests that fitness-compensating E1 mutations may accelerate escape from E2-targeting antibodies. Our analysis also identifies a set of E2-specific human monoclonal antibodies that are predicted to be especially resilient to escape via genetic variation in both E1 and E2, providing directions for robust HCV vaccine development. © The Author(s) 2023. Published by Oxford University Press.

This paper comprises the results of a design study that aims at developing a theoretically and empirically based learning trajectory on statistical inference for 9th-grade students. Based on theories of informal statistical inference, an 8-step learning trajectory was designed. The trajectory consisted of two similar four step sequences: (1) experimenting with a physical black box, (2) visualizing distributions, (3) examining sampling distributions using simulation software, and (4) interpreting sampling distributions to make inferences in real -life contexts. Sequence I included only categorical data and Sequence II regarded numerical data. The learning trajectory was implemented in an intervention among 267 students. To examine the effects of the trajectory on students’ understanding of statistical inference, we analyzed their posttest results after the intervention. To investigate how the stepwise trajectory fostered the learning process, students’ worksheets during each learning step were analyzed. The posttest results showed that students who followed the learning trajectory scored significantly higher on statistical inference and on concepts related to each step than students of a comparison group (n = 217) who followed the regular curriculum. Worksheet analysis demonstrated that the 8-step trajectory was beneficial to students’ learning processes. We conclude that ideas of repeated sampling with a black box and statistical modeling seem fruitful for introducing statistical inference. Both ideas invite more advanced follow-up activities, such as hypothesis testing and comparing groups. This suggests that statistics curricula with a descriptive focus can be transformed to a more inferential focus, to anticipate on subsequent steps in students’ statistics education.

We introduce a new rating system for tracking the development of parameters based on a stream of observations that can be viewed as paired comparisons. Rating systems are applied in competitive games, adaptive learning systems and platforms for product and service reviews. We model each observation as an outcome of a game of chance that depends on the parameters of interest (e.g. the outcome of a chess game depends on the abilities of the two players). Determining the probabilities of the different game outcomes is conceptualized as an urn problem, where a rating is represented by a probability (i.e. proportion of balls in the urn). This setup allows for evaluating the standard errors of the ratings and performing statistical inferences about the development of, and relations between, parameters. Theoretical properties of the system in terms of the invariant distributions of the ratings and their convergence are derived. The properties of the rating system are illustrated with simulated examples and its potential for answering research questions is illustrated using data from competitive chess, a movie review system, and an adaptive learning system for math.

Recently, researchers have encouraged the teaching of statistical inference to students at all levels. However, what constitutes pre-service teachers’ pedagogical content knowledge for statistical inference has not yet been given specific attention in research. This paper presents a qualitative study of pre-service teachers participating in a collaborative learning setup in a mathematics course to be prepared for teaching statistics in primary school aged 6–10 years. The study reported here is the first cycle of a design research project. This first-phase study explores how pre-service teachers’ pedagogical content knowledge for statistical inference can be developed during their mathematics course. The findings show that pre-service teachers’ learning opportunities regarding pedagogical content knowledge for statistical inference are insufficient. Based on the initial results, an initial conjecture map was constructed that guides the forthcoming design cycle.

Background: The aim was to investigate inferences of statistically significant test results among persons with more or less statistical education and research experience. Methods: A total of 75 doctoral students and 64 statisticians/epidemiologist responded to a web questionnaire about inferences of statistically significant findings. Participants were asked about their education and research experience, and also whether a 'statistically significant' test result (P = 0.024, alpha-level 0.05) could be inferred as proof or probability statements about the truth or falsehood of the null hypothesis (H-0) and the alternative hypothesis (H-1). Results: Almost all participants reported having a university degree, and among statisticians/epidemiologist, most reported having a university degree in statistics and were working professionally with statistics. Overall, 9.4% of statisticians/epidemiologist and 24.0% of doctoral students responded that the statistically significant finding proved that H-0 is not true, and 73.4% of statisticians/epidemiologists and 53.3% of doctoral students responded that the statistically significant finding indicated that H(0 )is improbable. Corresponding numbers about inferences about the alternative hypothesis (H-1) were 12.0% and 6.2% about proving H-1 being true and 62.7 and 62.5% for the conclusion that H-1 is probable. Correct inferences to both questions, which is that a statistically significant finding cannot be inferred as either proof or a measure of a hypothesis' probability, were given by 10.7% of doctoral students and 12.5% of statisticians/epidemiologists. Conclusions: Misinterpretation of P-values and statistically significant test results persists also among persons who have substantial statistical education and who work professionally with statistics.

Asymptotic statistical inference on productivity and production efficiency, using nonparametric envelopment estimators, is now available thanks to the basic central limit theorems (CLTs) developed in Kneip, Simar, and Wilson (2015). They provide asymptotic distributions of averages of Data Envelopment Analysis (DEA) and Free Disposal Hull (FDH) estimators of production efficiency. As shown in their Monte-Carlo experiments, due to the curse of dimensionality, the accuracy of the normal approximation is disappoint- ing when the sample size is not large enough. Simar and Zelenyuk (2020) have suggested a simple way to improve the approximation by using a more appropriate estimator of the variances. In this paper we suggest a novel way to improve the approximation, by smoothing out the spurious values of efficiency estimates when they are in a neighborhood of 1. This results in sharpening the data for observations near the estimated efficient frontier. The method is very easy to implement and does not require more computations than the original method. We compare our approach using Monte-Carlo experiments, both with the basic method and with the improved method suggested in Simar & Zelenyuk (2020) and in both cases we observe significant improvements. We show also that the Simar & Zelenyuk (2020) idea of the variance correction can also be adapted to our sharpening method, bringing additional improvements. We illustrate the method with some real data sets.

Infants use statistical information in their environment, as well as others' emotional communication, to understand the intentions of social partners. However, rarely do researchers consider these two sources of social information in tandem. This study assessed 2-year-olds' attributions of intentionality from non-random sampling events and subsequent discrete emotion reactions. Infants observed an experimenter remove five objects from either the non-random minority (18%) or random majority (82%) of a sample and express either joy, disgust, or sadness after each selection. Two-year-olds inferred the experimenter's intentionality by giving her the object that she had previously selected when she expressed joy or disgust after non-random sampling events, but not when she expressed sadness or sampled at random. These findings demonstrate that infants use both statistical regularities and discrete emotion communication to infer an agent's intentions. In particular, the present findings show that 2-year-olds infer that an agent can intentionally select a preferred or an undesired object from a sample as a function of the discrete emotion. Implications for the development of inferring intentionality from statistical sampling events and discrete emotion communication are discussed.