Abstract Methods of testing for a unit root in an observed series are described in this chapter. Both parametric regression tests and non‐parametric adjustments to these test statistics are considered, and tables of critical values for commonly used tests are given. The chapter also uses functionals of Wiener processes to describe the asymptotic distributions of important test statistics.
Statistics and Probability · The Economic Journal · · cited 327× · DOI ↗ · more like this ·
After reviewing the simulation performance of general‐to‐specific automatic regression‐model selection, as embodied in PcGets, we show how model selection can be non‐distortionary: approximately unbiased ‘selection estimates’ are derived, with reported standard errors close to the sampling standard deviations of the estimated DGP parameters, and a near‐unbiased goodness‐of‐fit measure. The handling of theory‐based restrictions, non‐stationarity and problems posed by collinear data are considered. Finally, we consider how PcGets can handle three ‘intractable’ problems: more variables than observations in regression analysis; perfectly collinear regressors; and modelling simultaneous equations without a priori restrictions.
We introduce the Envisioning Cards - a versatile toolkit for attending to human values during design processes - and discuss their early use. Drawing on almost twenty years of work in value sensitive design, the Envisioning Cards are built upon a set of four envisioning criteria: stakeholders, time, values, and pervasiveness. Each card contains on one side a title and an evocative image related to the card theme; on the flip side, the card shows the envisioning criterion, elaborates on the theme, and provides a focused design activity. Reports from the field demonstrate use in a range of research and design activities including ideation, co-design, heuristic critique, and more.
Sociology and Political Science · Evolutionary Applications · · cited 229× · DOI ↗ · more like this ·
Human-induced trait change has been documented in freshwater, marine, and terrestrial ecosystems worldwide. These trait changes are driven by phenotypic plasticity and contemporary evolution. While efforts to manage human-induced trait change are beginning to receive some attention, managing its ecological consequences has received virtually none. Recent work suggests that contemporary trait change can have important effects on the dynamics of populations, communities, and ecosystems. Therefore, trait changes caused by human activity may be shaping ecological dynamics on a global scale. We present evidence for important ecological effects associated with human-induced trait change in a variety of study systems. These effects can occur over large spatial scales and impact system-wide processes such as trophic cascades. Importantly, the magnitude of these effects can be on par with those…
Education · Journal of Further and Higher Education · · cited 204× · DOI ↗ · more like this ·
Abstract Constructivism is a philosophy based on the fundamental assumption that knowledge cannot exist outside our minds. Knowledge cannot be given from one mind to another. New knowledge is ‘constructed’ or created from within individuals through experience. In higher education, problem‐based learning (PBL) is an accepted instructional method or strategy for structuring learners’ experiences. We apply constructivist philosophy to PBL and incorporate the following variables: prior knowledge, quality of problems, tutor performance, group functioning, time spent in individual study, interest in subject matter, and assessment in a model of students’ learning in a medical course. Regardless of the specific teaching methods adopted in higher education, students’ creation of high levels of understanding and competence are promoted when arousal is optimised, self‐efficacy is maximised and…
Statistics and Probability · Econometrics Journal · · cited 184× · DOI ↗ · more like this ·
Kevin Hoover and Stephen Perez take important steps towards resolving somekey issues in econometric methodology. They simulate general‐to‐specific selection for linear, dynamic regression models, and find that their algorithm performs well in re‐mining the ‘Lovell database’. We discuss developments that improve on their results, automated in PcGets. Monte Carlo experiments and re‐analyses of empirical studies show that pre‐selection F‐tests, encompassing tests, and sub‐sample reliability checks all help eliminate ‘spuriously‐significant’ regressors, without impugning recovery of the correct specification.
There are two common ways in which developers are incorporating proprietary and domain-specific data when building applications of Large Language Models (LLMs): Retrieval-Augmented Generation (RAG) and Fine-Tuning. RAG augments the prompt with the external data, while fine-Tuning incorporates the additional knowledge into the model itself. However, the pros and cons of both approaches are not well understood. In this paper, we propose a pipeline for fine-tuning and RAG, and present the tradeoffs of both for multiple popular LLMs, including Llama2-13B, GPT-3.5, and GPT-4. Our pipeline consists of multiple stages, including extracting information from PDFs, generating questions and answers, using them for fine-tuning, and leveraging GPT-4 for evaluating the results. We propose metrics to assess the performance of different stages of the RAG and fine-Tuning pipeline. We conduct an in-depth…
Statistics and Probability · International Finance Discussion Paper · · cited 171× · DOI ↗ · more like this ·
This paper discusses the econometric methodology of general-to-specific modeling, in which the modeler simplifies an initially general model that adequately characterizes the empirical evidence within his or her theoretical framework. Central aspects of this approach include the theory of reduction, dynamic specification, model selection procedures, model selection criteria, model comparison, encompassing, computer automation, and empirical implementation. This paper thus reviews the theory of reduction, summarizes the approach of general-to-specific modeling, and discusses the econometrics of model selection, noting that general-to-specific modeling is the practical embodiment of reduction. This paper then summarizes fifty-seven articles key to the development of general-to-specific modeling.
Sociology and Political Science · Qualitative Inquiry · · cited 169× · DOI ↗ · more like this ·
This article raises questions about the basic assumptions of narrative research regarding representation, research relationships and the very construct of narrative as a method and a form of research. The author suggests the need to rethink the very understanding of narrative as research because this reduces stories to objects thereby reifying a positivist view of knowledge. An alternative view of narrative as outside of research is explored that rethinks research not as a scientific act but as a spiritual act, one that honors the sacredness of our stories and humanity.
Statistics and Probability · Econometrics · · cited 154× · DOI ↗ · more like this ·
To capture location shifts in the context of model selection, we propose selecting significant step indicators from a saturating set added to the union of all of the candidate variables. The null retention frequency and approximate non-centrality of a selection test are derived using a ‘split-half’ analysis, the simplest specialization of a multiple-path block-search algorithm. Monte Carlo simulations, extended to sequential reduction, confirm the accuracy of nominal significance levels under the null and show retentions when location shifts occur, improving the non-null retention frequency compared to the corresponding impulse-indicator saturation (IIS)-based method and the lasso.
Successful small-group learning in problem-based learning (PBL) educational programmes relies on functional group processes. However, there has been limited research on PBL group problems, and no studies have been conducted on problems as perceived by both students and tutors in the same educational context. The authors investigated PBL group problems in a graduate-entry medical programme, and report the most common group problems, and those that hinder students' learning the most. The possible causes of individual quietness and dominant behaviour, and potential influences that group problems may have on the tutorial process are summarized in an exploratory model of PBL group dysfunction that could be used to guide further research. Specifically, there is a need for further evidence on which to base guidelines for tutors and students to effectively manage group problems.
Artificial Intelligence · International Economic Review · · cited 132× · DOI ↗ · more like this ·
D. F. Hendry, Maximum Likelihood Estimation of Systems of Simultaneous Regression Equations with Errors Generated by a Vector Autoregressive Process, International Economic Review, Vol. 12, No. 2 (Jun., 1971), pp. 257-272
Statistics and Probability · Journal of Time Series Econometrics · · cited 128× · DOI ↗ · more like this ·
We outline a range of criteria for evaluating model selection approaches that have been used in the literature. Focusing on three key criteria, we evaluate automatically selecting the relevant variables in an econometric model from a large candidate set. General-to-specific selection is outlined for a regression model in orthogonal variables, where only one decision is required to select, irrespective of the number of regressors. Comparisons with an automated model selection algorithm, Autometrics (Doornik, 2009), show similar properties, but not restricted to orthogonal cases. Monte Carlo experiments examine the roles of post-selection bias corrections and diagnostic testing as well as evaluate selection in dynamic models by costs of search versus costs of inference.
Education · The Journal of Educational Research · · cited 120× · DOI ↗ · more like this ·
ABSTRACT The author suggests that all research is narrative. Resituating all research as narrative, as opposed to characterizing narrative as one particular form of inquiry, provides a critical space for rethinking research beyond current dualisms and bifurcations that create boundaries that limit the capacity for dialogue across diverse epistemologies. The contemporary bifurcation of research as either quantitative or qualitative, or as scientific or nonscientific, has resulted in a master narrative of research which assumes incommensurability across paradigms. The author weaves 3 questions through this research: (a) What is lost when narrative and science are constructed as opposing and incommensurable modes of inquiry? (b) How might scholars reconceptualize inquiry outside a binary framework that privileges science? and (c) In what ways can resituating all narrative as inquiry open…
We introduce a design method for evolving a co-design space to support stakeholders untrained in design. Specifically, the purpose of the method is to expand and shape a co-design space so that stakeholders, acting as designers, focus not only on the form and function of a tool being envisioned but also on the social context of its use and values that lie with individuals, groups, and societies. The method introduces value sensitive stakeholder prompts and designer prompts into a co-design process, creating a particular kind of reflection-on-action cycle. The prompts provide a means for bringing empirical data on values and theoretical perspective into the co-design process. We present the method in terms of a general model, the Value Sensitive Action-Reflection Model; place the model within discourse on co-design spaces; and illustrate the model with a discussion of its application in…
Computational Theory and Mathematics · IEEE Annals of the History of Computing · · cited 119× · DOI ↗ · more like this ·
John von Neumann made a key contribution to the understanding and development of computer architecture and design in his unpublished report titled 'First Draft of a Report on the EDVAC'. The architecture given in the von Neumann report is presented in a form that is accessible to a wider audience. This definition is compared with the definition of the constructed EDVAC system. In doing this, the authors hope to clarify important but previously unrecognized features of the von Neumann design and to clarify a number of the confusions that have arisen over the years.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
By their pervasiveness and by being worn on our bodies, mobile phones seem to have become intrinsic to safety. To examine this proposition, 43 participants, from four stakeholder groups (homeless young people, service providers, police officers, and community members), were asked to consider how homeless young people could use mobile phones to keep safe. Participants were asked to express their knowledge for place-based safety and to envision how mobile phones might be used to improve safety. Detailed analysis of the resulting data, which included value sketches, written value scenarios, and semi-structured discussion, led to specific design opportunities, related to values (e.g., supporting trust and desire to help others), function (e.g., documenting harms for future purposes), and form (e.g., leveraging social expectations for how mobile phones can be used to influence behavior).…
Education · Assessment & Evaluation in Higher Education · · cited 104× · DOI ↗ · more like this ·
Criteria and standards‐based assessment models are increasingly being adopted by universities as effective practice. However the promise of these models of assessment may not be realised unless teachers can find ways of making criteria and standards understandable to students. Exemplars or examples of previous students’ work of high and low quality can make criteria and standards concrete. Recent research has focussed on the use of exemplars to help students understand criteria and standards, and less emphasis has been given to exemplars simply as guides for students. This mixed methods study explores students’ perceptions of the usefulness of exemplars and different types of feedback for guiding them in completing assessments. A combination of engaging in marking and discussing exemplars, and receiving individualised and standards‐based feedback provides the most helpful guidance for…
Education · Research Online (University of Wollongong) · · cited 101× · OpenAlex ↗ · more like this ·
Peer observation of teaching is seen as a supportive and developmental process for improving the quality of teaching in universities. Evidence is emerging that the process of observing is just as if not more valuable than being observed and given feedback. In this study lecturers completing a Foundations program in university learning and teaching were interviewed about their experience of participating in a reciprocal peer observation exercise. The benefits for observers include learning about a new strategy and enhancing their confidence to try this strategy in their own teaching. Receiving feedback was also perceived to be useful but not more beneficial than watching a peer teach. We discuss implications of our results for units and institutions planning to implement peer observation as part of a strategic approach to improve the quality of learning and teaching.
Sociology and Political Science · Nurse Researcher · · cited 96× · DOI ↗ · more like this ·
Phenomenology is a recognised approach for investigating experiences in health research. Difficulties regarding the approach, however, have been documented with even the definitions and terminology sometimes being unclear. In addition to this, there have been claims that many nurse researchers have failed to report how the gap between philosophically related theory and research practice is managed. While legitimacy can be increased by claims regarding theoretical location, there have also been suggestions that engaging too intensely in methodological awareness can hinder the practice and progress of a research project. A balance is therefore required. This article concentrates on the dilemmas and challenges facing a researcher looking for an appropriate method and approach for a study investigating the experiences of stroke survivors. The challenges of using phenomenology as a research…
Education · Australian Journal of Education · · cited 95× · DOI ↗ · more like this ·
The constructivist approach is seen to be useful notably in science, mathematics and tertiary education. An account of non-radical constructivism is advanced and applied to classroom teaching and learning, and teaching strategies derived from the application of constructivist principles are identified and evaluated. It is suggested that many traditional classrooms are places of over-control, and certain traditional teaching methods may actually hinder students' potential to procreate human knowledge. Constructivism represents a fundamental challenge to many aspects of educational praxis, and may change significantly the ways by which young people are mass educated.
Education · Infant and Child Development · · cited 92× · DOI ↗ · more like this ·
High-quality, centre-based education and care during the early years benefit cognitive development, especially in children from disadvantaged backgrounds. During the COVID-19 pandemic and its associated lockdowns, access to early childhood education and care (ECEC) was disrupted. We investigate how this period affected the developmental advantages typically offered by ECEC. Using parent-report data from 189 families living in the UK, we explore associations between time spent in ECEC by 8-to-36-month-olds, their socioeconomic background, and their growth in language and executive functions between Spring and Winter 2020. Receptive vocabulary growth was greater in children who continued to attend ECEC during the period, with a stronger positive effect for children from less advantaged backgrounds. The growth of cognitive executive functions (CEFs) was boosted by ECEC attendance during…
Education · Journal of Further and Higher Education · · cited 92× · DOI ↗ · more like this ·
The aim of this article was to outline the basis for an introductory course to higher education focusing on students' approaches to learning and the role of teaching in higher education. The framework is a discussion on contemporary literature on approaches to learning. In the article experiences are also reported of developing and implementing an introductory course at the Sahlgrenska Academy, Göteborg University, Sweden. This was made within the LearnAble project, which was a two‐tier approach in which students were given an introductory course to higher education parallel with a course in pedagogy in higher education for teachers. The courses were evaluated with different instruments, which is also to some extent accounted for. The transition from upper secondary school to studies in higher education is discussed, as is the importance of a scholarship of teaching in the context of…
Geometry and Topology · Journal of Evolutionary Biology · · cited 90× · DOI ↗ · more like this ·
Beak size and shape in Darwin's finches have traditionally been quantified using a few univariate measurements (length, depth, width). Here we show the improved inferential resolution of geometric morphometric methods, as applied to three hierarchical levels: (i) among seven species on Santa Cruz Island, (ii) among different sites on Santa Cruz for a single species (Geospiza fortis), and (iii) between large and small beak size morphs of G. fortis at one site (El Garrapatero). Our results support previous studies in finding an axis of shape variation (long/shallow/pointy vs. short/deep/blunt) that separates many of the species. We also detect additional differences among species in the relative sizes and positions of the upper and lower mandibles and in curvature of the mandibles. Small-scale, but potentially relevant, shape variation was also detected among G. fortis from different…
Statistics and Probability · Oxford University Research Archive (ORA) (University of Oxford) · · cited 90× · DOI ↗ · more like this ·
After reviewing the simulation performance of general-to-specific automatic regression-model selection, as embodied in PcGets, we show how model selection can be non-distortionary: approximately unbiased "selection estimates" are derived, with reported standard errors close to the sampling standard deviations of the estimated DGP parameters, and a near-unbiased goodness-of-fit measure. The handling of theory-based restrictions, non-stationarity and problems posed by collinear data are considered. Finally, we consider how PcGets can handle three "intractable" problems: more variables than observations in regression analysis; perfectly collinear regressors; and modelling simultaneous equations without a priori restrictions.