Epistemic Error Research Articles

Epistemic errors in students\u2019 four fundamental operations of whole numbers: a radical constructivist grounded theory (RCGT) approach

The purpose of this study was to explore four fourth-grade students’ errors in whole number operations and associated causes. Task-based interviews were conducted to explore students' errors and their causes in four operations—addition, subtraction, multiplication, and division of whole numbers. The qualitative data in the form of students’ tasks and interview transcripts were analyzed to examine the nature of errors committed by the students. The findings of the study revealed that students committed epistemic and non-epistemic errors in the operations of whole numbers. Most of the errors in addition, subtraction, multiplication, and division problems were epistemic in nature due to a lack of understanding of the regrouping of numbers with position values and the role of a zero in different operations, a lack of understanding of the minuend and subtrahend in subtraction problems, a lack of understanding the role of zero in position values, and a misalignment of positional places for each numeral in minuend and subtrahend. The students could not comprehend the role of the remainder in the context of division problems and could not relate divisor, dividend, quotient and remainder. The non-epistemic errors seemed to arise from students’ lack of attention and curiosity, forgetting, and de-learning of the concepts in higher order operations and interpretations.

Detecting Uncertainty of Mortality Prediction Using Confident Learning.

Early mortality prediction is an actively researched problem that has led to the development of various severity scores and machine learning (ML) models for accurate and reliable detection of mortality in severely ill patients staying in intensive care units (ICUs). However, the uncertainty of such predictions due to irregular patient sampling, missing information, or high diversity of patient data has not yet been adequately addressed. In this paper, we used confident learning (CL) to incorporate sample-uncertainty information into our mortality prediction models and evaluated the performance of these models using a large dataset of 139,367 unique ICU admissions within the eICU Collaborative Research Database (eICU-CRD). The results of our study validate the importance of uncertainty quantification in patient outcome prediction and show that the state-of-the-art ML models augmented with CL are more robust against epistemic error and class imbalance.

Near-Source Ground Motions and Their Variability Derived from Dynamic Rupture Simulations Constrained by NGA-West2 GMPEs

ABSTRACTEmpirical ground-motion prediction equations (GMPEs) lack a sufficient number of measurements at near-source distances. Seismologists strive to supplement the missing data by physics-based strong ground-motion modeling. Here, we build a database of ~3000 dynamic rupture scenarios, assuming a vertical strike-slip fault of 36 × 20 km embedded in a 1D layered elastic medium and linear slip-weakening friction with heterogeneous parameters along the fault. The database is built by a Monte Carlo procedure to follow median and variability of Next Generation Attenuation-West2 Project GMPEs by Boore et al. (2014) at Joyner–Boore distances 10–80 km. The synthetic events span a magnitude range of 5.8–6.8 and have static stress drops between 5 and 40 MPa. These events are used to simulate ground motions at near-source stations within 5 km from the fault. The synthetic ground motions saturate at the near-source distances, and their variability increases at the near stations compared to the distant ones. In the synthetic database, the within-event and between-event variability are extracted for the near and distant stations employing a mixed-effect model. The within-event variability is lower than its empirical value, only weakly dependent on period, and generally larger for the near stations than for the distant ones. The between-event variability is by 1/4 lower than its empirical value at periods >1 s. We show that this can be reconciled by considering epistemic error in Mw when determining GMPEs, which is not present in the synthetic data.

Roundness prediction in centreless grinding using physics-enhanced machine learning techniques

This work proposes a model for suggesting optimal process configuration in plunge centreless grinding operations. Seven different approaches were implemented and compared: first principles model, neural network model with one hidden layer, support vector regression model with polynomial kernel function, Gaussian process regression model and hybrid versions of those three models. The first approach is based on an enhancement of the well-known numerical process simulation of geometrical instability. The model takes into account raw workpiece profile and possible wheel-workpiece loss of contact, which introduces an inherent limitation on the resulting profile waviness. Physical models, because of epistemic errors due to neglected or oversimplified functional relationships, can be too approximated for being considered in industrial applications. Moreover, in deterministic models, uncertainties affecting the various parameters are not explicitly considered. Complexity in centreless grinding models arises from phenomena like contact length dependency on local compliance, contact force and grinding wheel roughness, unpredicted material properties of the grinding wheel and workpiece, precision of the manual setup done by the operator, wheel wear and nature of wheel wear. In order to improve the overall model prediction accuracy and allow automated continuous learning, several machine learning techniques have been investigated: a Bayesian regularized neural network, an SVR model and a GPR model. To exploit the a priori knowledge embedded in physical models, hybrid models are proposed, where neural network, SVR and GPR models are fed by the nominal process parameters enriched with the roundness predicted by the first principle model. Those hybrid models result in an improved prediction capability.

A simple and effective Measurement-Changes-Correction strategy for damage identification with aleatoric and epistemic model errors

The model errors are often inevitably encountered when a practical structure is analyzed by a mathematical baseline model, and such model errors would have undesired effects on damage identification. Aiming to alleviate the effect of the model errors, a new Measurement-Changes-Correction strategy that simply uses measurement changes to correct the measured data is developed in this article. To this end, the performance of the Measurement-Changes-Correction strategy under the aleatoric and epistemic model errors is systematically analyzed. The analysis is proceeded along with comparison to the conventional two-step strategy where the intact damage parameters are first updated from the measured data and then, damage identification is conducted upon the updated parameters. Thereafter, it is found that (1) the Measurement-Changes-Correction strategy performs as well as the two-step strategy under small aleatoric and epistemic model errors, (2) the Measurement-Changes-Correction strategy can still work under large scaling model parameter errors, but the two-step strategy may not, and (3) the Measurement-Changes-Correction strategy only requires correction of the measured data by the measurement changes so that the computation cost by the Measurement-Changes-Correction strategy is almost half of that by the two-step strategy. Numerical examples and International Association for Structural Control–American Society of Civil Engineers benchmark problems are studied to verify the performance of the proposed Measurement-Changes-Correction strategy.

Using Physically Based Synthetic Peak Flows to Assess Local and Regional Flood Frequency Analysis Methods

AbstractAlthough prior studies have evaluated the role of sampling errors associated with local and regional methods to estimate peak flow quantiles, the investigation of epistemic errors is more difficult because the underlying properties of the random variable have been prescribed using ad‐hoc characterizations of the regional distributions of peak flows. This study addresses this challenge using representations of regional peak flow distributions derived from a combined framework of stochastic storm transposition, radar rainfall observations, and distributed hydrologic modeling. The authors evaluated four commonly used peak flow quantile estimation methods using synthetic peak flows at 5,000 sites in the Turkey River watershed in Iowa, USA. They first used at‐site flood frequency analysis using the Pearson Type III distribution with L‐moments. The authors then pooled regional information using (1) the index flood method, (2) the quantile regression technique, and (3) the parameter regression. This approach allowed quantification of error components stemming from epistemic assumptions, parameter estimation method, sample size, and, in the regional approaches, the number of pooled sites. The results demonstrate that the inability to capture the spatial variability of the skewness of the peak flows dominates epistemic error for regional methods. We concluded that, in the study basin, this variability could be partially explained by river network structure and the predominant orientation of the watershed. The general approach used in this study is promising in that it brings new tools and sources of data to the study of the old hydrologic problem of flood frequency analysis.

Efficient Propagation of Epistemic Uncertainty in the Median Ground‐Motion Model in Probabilistic Hazard Calculations

Abstract A computationally efficient methodology for propagating the epistemic uncertainty in the median ground motion in probabilistic seismic hazard analysis is developed using the polynomial chaos (PC) approach. For this application, the epistemic uncertainty in the median ground motion for a specific scenario is assumed to be lognormally distributed and fully correlated across earthquake scenarios. In the hazard calculation, a single central ground‐motion model (GMM) is used for the median along with the epistemic standard error of the median for each scenario. A set of PC coefficients is computed for each scenario and each test ground‐motion level. The additional computation burden in computing these PC coefficients depends on the order of the approximation but is less than computing the median ground motion from one additional GMM. With the PC method, the mean and fractiles of the hazard due to the epistemic uncertainty distribution of the median ground motion are computed as a postprocess that is very fast computationally. For typical values of the standard deviation of epistemic uncertainty in the median ground motion (<0.2 natural log units), the methodology accurately estimates the epistemic uncertainty distribution of the hazard over the 1%–99% range. This full epistemic range is not well modeled with just a small number of GMM branches uses in the traditional logic‐tree approach. The PC method provides more accuracy, faster computation, and reduced memory requirements than the traditional approach. For large values of the epistemic uncertainty in the median ground motion, a higher order of the PC expansion may be needed to be included to capture the full range of the epistemic uncertainty.

The ethics of arguing

ABSTRACTContemporary argumentation theory has developed an impressive array of norms, goals and virtues applicable to ideal argument. But what is the moral status of these prescriptions? Is an interlocutor who fails to live up to these norms guilty of a moral failing as well as an epistemic or cognitive error? If so, why? In answering these questions, I argue that deliberation’s epistemic and cognitive goods attach to important ethical goods, and that respect for others’ rationality, the ethics of joint action, and the importance of consensus join forces with these goods to provide strong reasons for cleaving to high standards of argument. I sketch an illustrative continuum of argument practices of different deliberative-cum-ethical standards, and consider how one should ethically respond when faced with an interlocutor employing less than ideal standards.

Evaluating Subsurface Parameterization to Simulate Hyporheic Exchange: The Steinlach River Test Site.

Hyporheic exchange is the interaction of river water and groundwater, and is difficult to predict. One of the largest contributions to predictive uncertainty for hyporheic exchange has been attributed to the representation of heterogeneous subsurface properties. Our study evaluates the trade-offs between intrinsic (irreducible) and epistemic (reducible) model errors when choosing between homogeneous and highly complex subsurface parameter structures. We modeled the Steinlach River Test Site in Southwest Germany using a fully coupled surface water-groundwater model to simulate hyporheic exchange and to assess the predictive errors and uncertainties of transit time distributions. A highly parameterized model was built, treated as a "virtual reality" and used as a reference. We found that if the parameter structure is too simple, it will be limited by intrinsic model errors. By increasing subsurface complexity through the addition of zones or heterogeneity, we can begin to exchange intrinsic for epistemic errors. Thus, the appropriate level of detail to represent the subsurface depends on the acceptable range of intrinsic structural errors for the given modeling objectives and the available site data. We found that a zonated model is capable of reproducing the transit time distributions of a more detailed model, but only if the geological structures are known. An interpolated heterogeneous parameter field (cf. pilot points) showed the best trade-offs between the two errors, indicating fitness for practical applications. Parameter fields generated by multiple-point geostatistics (MPS) produce transit time distributions with the largest uncertainties, however, these are reducible by additional hydrogeological data, particularly flux measurements.

The Harm of Ableism: Medical Error and Epistemic Injustice.

This paper argues that epistemic errors rooted in group- or identity-based biases, especially those pertaining to disability, are undertheorized in the literature on medical error. After sketching dominant taxonomies of medical error, we turn to the field of social epistemology to understand the role that epistemic schemas play in contributing to medical errors that disproportionately affect patients from marginalized social groups. We examine the effects of this unequal distribution through a detailed case study of ableism. There are four primary mechanisms through which the epistemic schema of ableism distorts communication between nondisabled physicians and disabled patients: testimonial injustice, epistemic overconfidence, epistemic erasure, and epistemic derailing. Measures against epistemic injustices in general and against schema-based medical errors in particular are ultimately issues of justice that must be better addressed at all levels of health care practice.

The Head, the Hand, and Matter: New Materialism and the Politics of Knowledge

This article seeks to examine the political connotations of a recent ‘material turn’ in social and political theory and its implications for theorizations of political agency. ‘New materialist’ theories are premised upon transcending the limits which social constructivism places upon thought, viewed as a reification of the division of subject and object and so a hubristic anthropocentrism which places human beings at the centre of social existence. Yet new materialist theories have tended to locate the conditions of the separation of mind and world they seek to overcome upon the terrain of epistemic or ethical error. By taking the work of Quentin Meillassoux, Jane Bennett and Karen Barad as exemplary, this article contends that new materialist theories not only fall short of their own materialist pretensions insofar as they do not interrogate the material conditions of the separation of the mental and material, but that the failure to do so has profound repercussions for the success of their accounts of political agency. This essay seeks to offer a counter-narrative to new materialist theories by situating the hierarchy between thought and world as a structural feature of capitalist social relations.

What's Epistemically Wrong with Conspiracy Theorising?

AbstractBelief in conspiracy theories is often taken to be a paradigm of epistemic irrationality. Yet, as I argue in the first half of this paper, standard criticisms of conspiracy theorising fail to demonstrate that the practice is invariably irrational. Perhaps for this reason, many scholars have taken a relatively charitable attitude toward conspiracy theorists and conspiracy theorising in recent years. Still, it would be a mistake to conclude from the defence of conspiracy theorising offered here that belief in conspiracy theories is on an epistemic par with belief in other theories. I argue that a range of epistemic errors are pervasive among conspiracy theorists. First, the refusal of conspiracy theorists to accept the official account of some target event often seems to be due to the exercise of a probabilistic, and fallacious, extension ofmodus tollens. Additionally, conspiracy theorists tend to be inconsistent in their intellectual attention insofar as the effort they expend on uncovering the truth excludes attention to their own capacities for biased or otherwise erroneous reasoning. Finally, the scepticism with which conspiracy theorists tend to view common sources of information leaves little room for conspiracy theorists to attain positive warrant for their preferred explanations of target events.

Accounting for uncertain fault geometry in earthquake source inversions – I: theory and simplified application

The ill-posed nature of earthquake source estimation derives from several factors including the quality and quantity of available observations and the fidelity of our forward theory. Observational errors are usually accounted for in the inversion process. Epistemic errors, which stem from our simplified description of the forward problem, are rarely dealt with despite their potential to bias the estimate of a source model. In this study, we explore the impact of uncertainties related to the choice of a fault geometry in source inversion problems. The geometry of a fault structure is generally reduced to a set of parameters, such as position, strike and dip, for one or a few planar fault segments. While some of these parameters can be solved for, more often they are fixed to an uncertain value. We propose a practical framework to address this limitation by following a previously implemented method exploring the impact of uncertainties on the elastic properties of our models. We develop a sensitivity analysis to small perturbations of fault dip and position. The uncertainties of our fixed fault geometry are included in the inverse problem under the formulation of the misfit covariance matrix that combines both prediction and observation uncertainties. We validate this approach with the simplified case of a fault that extends infinitely along strike, using both Bayesian and optimization formulations of a static slip inversion. If epistemic errors are ignored, predictions are overconfident in the data and slip parameters are not reliably estimated. In contrast, inclusion of uncertainties in fault geometry allows us to infer a robust posterior slip model. Epistemic uncertainties can be many orders of magnitude larger than observational errors for great earthquakes (Mw > 8). Not accounting for uncertainties in fault geometry may partly explain observed shallow slip deficits for continental earthquakes. Similarly, ignoring the impact of epistemic errors can also bias estimates of near-surface slip and predictions of tsunamis induced by megathrust earthquakes.

Radiofrequency Electromagnetic Radiation and Memory Performance: Sources of Uncertainty in Epidemiological Cohort Studies

Uncertainty in experimental studies of exposure to radiation from mobile phones has in the past only been framed within the context of statistical variability. It is now becoming more apparent to researchers that epistemic or reducible uncertainties can also affect the total error in results. These uncertainties are derived from a wide range of sources including human error, such as data transcription, model structure, measurement and linguistic errors in communication. The issue of epistemic uncertainty is reviewed and interpreted in the context of the MoRPhEUS, ExPOSURE and HERMES cohort studies which investigate the effect of radiofrequency electromagnetic radiation from mobile phones on memory performance. Research into this field has found inconsistent results due to limitations from a range of epistemic sources. Potential analytic approaches are suggested based on quantification of epistemic error using Monte Carlo simulation. It is recommended that future studies investigating the relationship between radiofrequency electromagnetic radiation and memory performance pay more attention to treatment of epistemic uncertainties as well as further research into improving exposure assessment. Use of directed acyclic graphs is also encouraged to display the assumed covariate relationship.

Assessing uncertainties in prevailing methodologies for modeling radiative transfer in simulations of oxygen-enriched methane flames

The fidelities and uncertainties associated with prevailing radiative transfer modeling methodologies and practices during simulations of oxygen-enriched flames where the gas temperature exceeds 2500 K are examined. Reference data in representative furnace environments characteristic of methane combustion at oxidizer oxygen concentrations of 30–100% were generated employing the spectral-line-based-weighted-sum-of-gray-gas model (SLWSGGM). Parametric uncertainties associated with two commonly employed WSGGM were assessed. While predictions from the exponential wide band model (EWBM) based WSGGM formulation agreed very well against reference data, significant errors in the radiative source term predictions resulted when the HITEMP-based WSGGM formulation was employed in high temperature regions beyond its intended application. The epistemic uncertainties associated with the wall temperature profiles decreased with oxygen enrichment. However, fully coupled combustion simulations of 100% O2–methane flames for a range of fuel inlet Reynolds numbers (2600–14,000) resulted in identical gas temperature profiles employing both the EWBM and HITEMP WSGG models. Further, their incident wall radiative flux predictions were within 10% of each other. Epistemic errors resulting from neglecting CO as a radiatively participating specie could become significant. Aleatoric uncertainties resulting from turbulence radiation interaction modeling were pronounced when flame radiation was the dominant contributor to the wall fluxes.

Assessing accuracy in the estimation of spectral content in wave energy resource on the French Atlantic test site SEMREV

This paper presents a detailed assessment of the spectral accuracy of state of the art numerical estimations of wave energy at the SEMREV French Atlantic test site, by comparison with in-situ measurements. Power density frequency spectra from HOMERE Boudiere et al. (2013), the most up-to-date highly refined hindcast dataset, are compared with several measurements from three wave buoys located either offshore or onsite. The spectral signature of the error exhibits a highly non-linear site dependent behavior. Examined in conjunction with usual comparisons of standard integral parameters, this provides meaningful insight into the epistemic uncertainties and errors in different part of the wave energy spectrum. Notably, a complementary analysis of the mean available energy as a function of frequency illustrates the varying degree of impact that inaccuracies in the estimation could have on production by WECs, which are often designed to harvest energy from specific frequency bandwidths.It notably also enlighten the frequency domains where the input forcing and the accounting or parameterization of processes may still lack accuracy.

Epistemology shmepistemology: moral error theory and epistemic expressivism

Some philosophers object to moral error theory by arguing that there a parity between moral and epistemic normativity. They maintain that moral and epistemic error theory stand or fall together, that epistemic error theory falls, and that moral error theory thus falls too. This paper offers a response to this objection on behalf of moral error theorists. I defend the view that moral and epistemic error theory do not stand or fall together by arguing that moral error theory can be sustained alongside epistemic expressivism. This unusual combination of theories can be underpinned by differences in the foundational norms that guide moral and epistemic inquiry. I conclude that the problem of epistemic normativity fails to show that it is compulsory for us to reject moral error theory.

Unreinforced masonry walls with irregular opening layouts: reliability of equivalent-frame modelling for seismic vulnerability assessment

Equivalent-frame modelling for the estimate of in-plane seismic vulnerability of masonry wall is endorsed by international codes. However, in presence of irregular layouts of openings, the definition of equivalent-frame becomes non-unique and may lead to epistemic modelling errors. The aim of this paper is to develop a simplified tool to estimate such errors as functions of the level or geometric irregularity. Four types of irregularity are considered and quantified by geometric indexes. Nonlinear static analyses are carried out through an equivalent-frame model and a nonlinear finite-element model. It is assumed that both models are exact for perfectly-regular walls, while in presence of irregularities only the continuum approach is able to correctly predict the vulnerability, so that the difference between the two results can be taken as a measure of the error of the equivalent frame. A regular wall is taken as case study and both numerical models are calibrated on experimental data. The four typologies of irregularity are introduced separately, each in steps of increasing magnitude. The results show that the equivalent frame approach is usually not safety-preserving and may be affected by a significant error. A tentative global measure of irregularity, weighting the average effects of each irregularity type, is proposed. The resulting simplified relationship between irregularity level and expected error can be used to define a confidence factor on the modelling, which penalizes the application of the equivalent frame in case of irregular walls. Eventually, an applicability limit of the equivalent-frame model might be set.

How effective is bimodal soil hydraulic characterization? Functional evaluations for predictions of soil water balance

SummaryTo overcome some drawbacks of the unimodal relations commonly used to describe soil hydraulic properties (SHPs), previously we developed bimodal lognormal relations that have the following main features: (i) they are closed‐form expressions, (ii) they have a sound theoretical basis and provide a more general conceptualization of soil and (iii) they improve the description of both the water retention (WRF) and hydraulic conductivity (HCF) functions. Nevertheless, the reliability of soil hydraulic analytical relations is often tested only at the curve fitting level. Comparisons between unimodal and bimodal soil hydraulic relations are more effective and informative when performed within a functional evaluation approach. We use the HYDRUS‐1D package to quantify and compare soil moisture dynamics and storage regimes for hydrological processes at both the event and annual time‐scales when the soil domain is characterized by either unimodal or bimodal hydraulic properties. Seven soil samples taken from a previous study were used in numerical simulations of drainage or infiltration processes; there were large relative discrepancies in terms of simulated soil water storage. A subsequent test that involved simulations of soil water budget for the period 2000–2012 was implemented for a peach‐orchard field by a conventional scaling method. This test also enables soil spatial variation to be taken into consideration. Two different scenarios enable the epistemic uncertainty to be evaluated when different hydraulic models are considered for soil with weak or strong bimodality. With Willmott's refined index of agreement, discrepancies in soil water storage were about 15% (weak bimodality) or more than 30% (strong bimodality).Highlights Main aim of this study is the assessment of epistemic uncertainty in modelling soil water dynamics. We make functional evaluations for both event‐based and long‐term hydrological processes. Disregard of bimodal soil hydraulic behaviour can lead to large epistemic errors. Better predictions of soil hydraulic properties should be sought in future research.

Use of a flux‐based field capacity criterion to identify effective hydraulic parameters of layered soil profiles subjected to synthetic drainage experiments

AbstractThis study explores the feasibility of identifying the effective soil hydraulic parameterization of a layered soil profile by using a conventional unsteady drainage experiment leading to field capacity. The flux‐based field capacity criterion is attained by subjecting the soil profile to a synthetic drainage process implemented numerically in the Soil‐Water‐Atmosphere‐Plant (SWAP) model. The effective hydraulic parameterization is associated to either aggregated or equivalent parameters, the former being determined by the geometrical scaling theory while the latter is obtained through the inverse modeling approach. Outcomes from both these methods depend on information that is sometimes difficult to retrieve at local scale and rather challenging or virtually impossible at larger scales. The only knowledge of topsoil hydraulic properties, for example, as retrieved by a near‐surface field campaign or a data assimilation technique, is often exploited as a proxy to determine effective soil hydraulic parameterization at the largest spatial scales. Comparisons of the effective soil hydraulic characterization provided by these three methods are conducted by discussing the implications for their use and accounting for the trade‐offs between required input information and model output reliability. To better highlight the epistemic errors associated to the different effective soil hydraulic properties and to provide some more practical guidance, the layered soil profiles are then grouped by using the FAO textural classes. For the moderately heterogeneous soil profiles available, all three approaches guarantee a general good predictability of the actual field capacity values and provide adequate identification of the effective hydraulic parameters. Conversely, worse performances are encountered for the highly variable vertical heterogeneity, especially when resorting to the “topsoil‐only” information. In general, the best performances are guaranteed by the equivalent parameters, which might be considered a reference for comparisons with other techniques. As might be expected, the information content of the soil hydraulic properties pertaining only to the uppermost soil horizon is rather inefficient and also not capable to map out the hydrologic behavior of the real vertical soil heterogeneity since the drainage process is significantly affected by profile layering in almost all cases.