Different Sources Of Variation Research Articles

Continuous Biosensor Based on Particle Motion: How Does the Concentration Measurement Precision Depend on Time Scale?

Continuous biosensors measure concentration-time profiles of biomolecular substances in order to allow for comparisons of measurement data over long periods of time. To make meaningful comparisons of time-dependent data, it is essential to understand how measurement imprecision depends on the time interval between two evaluation points, as the applicable imprecision determines the significance of measured concentration differences. Here, we define a set of measurement imprecisions that relate to different sources of variation and different time scales, ranging from minutes to weeks, and study these using statistical analyses of measurement data. The methodology is exemplified for Biosensing by Particle Motion (BPM), a continuous, affinity-based sensing technology with single-particle and single-molecule resolution. The studied BPM sensor measures specific small molecules (glycoalkaloids) in an industrial food matrix (potato fruit juice). Measurements were performed over several months at two different locations, on nearly 50 sensor cartridges with in total more than 1000 fluid injections. Statistical analyses of the measured signals and concentrations show that the relative residuals are normally distributed, allowing extraction and comparisons of the proposed imprecision parameters. The results indicate that sensor noise is the most important source of variation followed by sample pretreatment. Variations caused by fluidic transport, changes of the sensor during use (drift), and variations due to different sensor cartridges and cartridge replacements appear to be small. The imprecision due to sensor noise is recorded over few-minute time scales and is attributed to stochastic fluctuations of the single-molecule measurement principle, false-positive signals in the signal processing, and nonspecific interactions. The developed methodology elucidates both time-dependent and time-independent factors in the measurement imprecision, providing essential knowledge for interpreting concentration-time profiles as well as for further development of continuous biosensing technologies.

A Dataset for Detection and Segmentation of Underwater Marine Debris in Shallow Waters

Robust object detection is crucial for automating underwater marine debris collection. While supervised deep learning achieves state-of-the-art performance in discriminative tasks, replicating this success on underwater data is challenging. The generalization of these methods suffers due to a lack of available annotated data considering different sources of variation in the unstructured underwater environment and imaging conditions. In this paper, we present the Seaclear Marine Debris Dataset, the first publicly available shallow-water marine debris dataset annotated for instance segmentation/object detection. The dataset contains 8610 images collected using ROVs at multiple locations and with different cameras, annotated for 40 object categories, encompassing not only litter but also observed animals, plants, and robot parts. As part of the technical validation, we provide baseline results for object detection using Faster RCNN and YOLOv6 models. Furthermore, we demonstrate the non-triviality of generalizing the trained model performance to unseen sites and cameras due to domain shift. This underscores the value of the presented dataset in further developing robust models for underwater debris detection.

Selection tests work better than we think they do, and have for years

AbstractWe can make better decisions when we have a better understanding of the different sources of variance that impact job performance ratings. A failure to do so cannot only lead to inaccurate conclusions when interpreting job performance ratings, but often misguided efforts aimed at improving our ability to explain and predict them. In this paper, we outline six recommendations relating to the interpretation of predictive validity coefficients and efforts aimed at predicting job performance ratings. The first three focus on the need to evaluate the effectiveness of selection instruments and systems based only on the variance they can possibly account for. When doing so, we find that it is not only possible to account for the majority of the variance in job performance ratings that most select systems can possibly predict, but that we’ve been able to account for this variance for years. Our last three recommendations focus on the need to incorporate components related to additional sources of variance in our predictive models. We conclude with a discussion of their implications for both research and practice.

Polynomial chaos expansion of SAR and temperature increase variability in 3 T MRI due to stochastic input data

Objective. Numerical simulations are largely adopted to estimate dosimetric quantities, e.g. specific absorption rate (SAR) and temperature increase, in tissues to assess the patient exposure to the radiofrequency (RF) field generated during magnetic resonance imaging (MRI). Simulations rely on reference anatomical human models and tabulated data of electromagnetic and thermal properties of biological tissues. However, concerns may arise about the applicability of the computed results to any phenotype, introducing a significant degree of freedom in the simulation input data. In addition, simulation input data can be affected by uncertainty in relative positioning of the anatomical model with respect to the RF coil. The objective of this work is the to estimate the variability of SAR and temperature increase at 3 T head MRI due to different sources of variability in input data, with the final aim to associate a global uncertainty to the dosimetric outcomes. Approach. A stochastic approach based on arbitrary Polynomial Chaos Expansion is used to evaluate the effects of several input variability’s (anatomy, tissue properties, body position) on dosimetric outputs, referring to head imaging with a 3 T MRI scanner. Main results. It is found that head anatomy is the prevailing source of variability for the considered dosimetric quantities, rather than the variability due to tissue properties and head positioning. From knowledge of the variability of the dosimetric quantities, an uncertainty can be attributed to the results obtained using a generic anatomical head model when SAR and temperature increase values are compared with safety exposure limits. Significance. This work associates a global uncertainty to SAR and temperature increase predictions, to be considered when comparing the numerically evaluated dosimetric quantities with reference exposure limits. The adopted methodology can be extended to other exposure scenarios for MRI safety purposes.

Large Individual Differences in Functional Connectivity in the Context of Major Depression and Antidepressant Pharmacotherapy.

Clinical studies of major depression (MD) generally focus on group effects, yet interindividual differences in brain function are increasingly recognized as important and may even impact effect sizes related to group effects. Here, we examine the magnitude of individual differences in relation to group differences that are commonly investigated (e.g., related to MD diagnosis and treatment response). Functional MRI data from 107 participants (63 female, 44 male) were collected at baseline, 2, and 8 weeks during which patients received pharmacotherapy (escitalopram, N = 68) and controls (N = 39) received no intervention. The unique contributions of different sources of variation were examined by calculating how much variance in functional connectivity was shared across all participants and sessions, within/across groups (patients vs controls, responders vs nonresponders, female vs male participants), recording sessions, and individuals. Individual differences and common connectivity across groups, sessions, and participants contributed most to the explained variance (>95% across analyses). Group differences related to MD diagnosis, treatment response, and biological sex made significant but small contributions (0.3-1.2%). High individual variation was present in cognitive control and attention areas, while low individual variation characterized primary sensorimotor regions. Group differences were much smaller than individual differences in the context of MD and its treatment. These results could be linked to the variable findings and difficulty translating research on MD to clinical practice. Future research should examine brain features with low and high individual variation in relation to psychiatric symptoms and treatment trajectories to explore the clinical relevance of the individual differences identified here.

Growth Promotion of Raphanus Sativus L. under Doses of Phosphorus with Application of Trichoderma Spp.

This study aimed to evaluate the growth of Raphanus sativus L. under six phosphorus doses in combination with three strains of Trichoderma spp. The design was a randomized block design with seven replications in a factorial scheme, 4 x 5 (treatments and doses). Treatments related to Trichoderma spp. were T1 – control, without application of T. harzianum strain; T2 – T. harzianum ESALQ 1306, T3 – T. harzianum IBLF 006 WP, and T4 – T. asperellum URM 5911, at doses of 0, 40, 80, 120, 160, and 200 kg ha-1 of P2O5, in a greenhouse. The evaluated characteristics were tuberous root diameter (RDI), tuberous root length (RLE), relative chlorophyll index (RCI), leaf dry mass (LDM), tuberous root fresh mass (FTRM), and tuberous root dry mass (TRDM). It was observed that there was a significant difference for the different sources of variation evaluated, indicating an interaction between the different isolates and doses. It is concluded that the application of Trichoderma harzianum - ESALQ 1306 strain is recommended for radish cultivation, with increases of 13% for tuberous root length, 7% for relative chlorophyll index, 23% for leaf dry mass, 35% for tuberous root fresh mass, and 12% for tuberous root dry mass compared to the control, at a dose of 104 kg ha-1 of P2O5, with the possibility of dose reduction and more sustainable management. Additionally, strains T. harzianum IBLF 006 WP (T3) and T. asperellum URM 5911 (T4) are not recommended for radish cultivation.

Influence of surface water and groundwater on functional traits and trade-off strategies of oasis communities at the end of the Keriya River, China.

Plant functional traits reflect the capacity of plants to adapt to their environment and the underlying optimization mechanisms. However, few studies have investigated trade-off strategies for functional traits in desert-wetland ecosystems, the mechanisms by which surface water disturbance and groundwater depth drive functional trait variation at the community scale, and the roles of intraspecific and interspecific variation. Therefore, this study analyzed specific differences in community-weighted mean traits among habitat types and obtained the relative contribution of intraspecific and interspecific variation by decomposing community-weighted mean traits, focusing on the Daliyabuyi Oasis in the hinterland of the Taklamakan Desert. We also explored the mechanisms by which surface water and groundwater influence different sources of variability specifically. The results showed that plant height, relative chlorophyll content, leaf thickness, leaf nitrogen content, and nitrogen-phosphorus ratio were the key traits reflecting habitat differences. As the groundwater depth becomes shallower and surface water disturbance intensifies, plant communities tend to have higher leaf nitrogen content, nitrogen-phosphorus ratio, and relative chlorophyll content and lower height. Surface water, groundwater, soil water content, and total soil nitrogen can influence interspecific and intraspecific variation in these traits through direct and indirect effects. As arid to wet habitats change, plant trade-off strategies for resources will shift from conservative to acquisitive. The study concluded that community functional traits are mainly contributed by interspecific variation, but consideration of intraspecific variation and the covariation effects that exist between it and interspecific variation can help to further enhance the understanding of the response of community traits in desert-wetland ecosystems to environmental change. Surface water disturbance has a non-negligible contribution to this adaptation process and plays a higher role than groundwater depth.

Variability of Dynamic Properties of Rubber Compounds for Elastomeric Bearings

The λ-factors for stiffness and damping of rubber bearings should be experimentally assessed during the qualification process or deduced from tests performed on material specimens. Moreover, the λ-factors suggested in the informative annexes of EN 15129 and of EC8-part 2 can be also used as reference values. However, they are derived from outdated experimental campaigns and do not refer to all the sources of variability. In this paper, a statistical analysis on a significant set of rubber compounds, certified according to EN 15129 from different suppliers, is carried out to assess the current variability of the dynamic properties of such compounds. Different sources of variability may be identified by distinguishing between behavioural and environmental effects. For elastomeric bearings, especially high-damping rubber (HDR) ones, the main behavioural effects are strain amplitude, strain rate dependence, and cyclic degradation, whereas the environmental effects are due to temperature variation and ageing. All these sources of variability have been analysed in this paper. The results of the statistical analysis have been used to propose a new set of λ-factors for all the source of variability studied. Such new values have been compared with the ones suggested by the codes when available. The main inconsistencies found have been highlighted and commented in this paper. Finally, some considerations about the influence of such variability on the structural response of base-isolated structures have been drawn by focusing on both the isolation system and the superstructure.

Comparison of Academic, Administrative and Community Rater Scores at a Multiple Mini Interview Using Generalisability Theory

Multiple Mini Interviews (MMIs) are sampling approaches that use multiple short stations to select prospective students for professional programmes. Each station uses different interview scenarios and raters to effectively assess candidates’ noncognitive skills. This study compared the performances of three sets of raters; academic, administrative staff, and community members, in an MMI for student selection using performance comparisons and Generalisability Theory to estimate the different sources of variance and generalisability (reliability) coefficients. The study aims to analyse the differences in performance scores from these raters and their psychometric projections on reliability with different samples of raters and stations. Eleven candidates participated in the 10-station MMI, each with an eight-minute duration, two minutes of preparation, and an academic assessment using a marking rubric. The entire interview was video recorded. The administrative staff and community members watched the videos independently and graded all candidates’ performances using the same marking rubric. Generalisability and Decision studies were used to analyse the collected data. Community members were the strictest, while academics were the most lenient. There were statistically significant differences between rater categories in six stations. The generalisability coefficient of 0.85 of onerater results from the Decision study suggested good reliability of the 10-station MMI. The Decision study found that generalisability coefficients improved more with an increasing number of raters rather than number of stations. Four stations contributed to unreliability in each rater category and a combination of the rater categories. Information on number of stations, number of raters, and type of rater combination required to achieve good reliability enabled informed decisions on the process and implementation of the MMI. The station simulation that influenced unreliability helped us improve station writing and identify focus areas for training and development.

Process-Voltage-Temperature Variations Assessment in Energy-Aware Resistive RAM-Based FPGAs

Resistive Random Access Memory (RRAM) technology holds promises to improve the Field Programmable Gate Array (FPGA) performance, reduce the area footprint, and dramatically lower run-time energy requirements compared to the state-of-the-art CMOS-based products. However, the integration of RRAM in FPGAs is hindered by the high programming power consumption and by non-ideal behaviors of the device due to its stochastic nature that may overshadow the benefits in normal operation mode. To cope with these challenges, optimized programming strategies have to be investigated. In this work, we explore the impact that different procedures to set the device have on the run-time performance. Process, voltage, and temperature (PVT) variations as well as time-dependent drift effect of the RRAM device are considered in the assessment of 4T1R MUX designs characteristics. The comparison with tradition CMOS implementations reveals how the choice of the target resistive state and the programming algorithm are key design aspects to reduce the run-time delay and energy metrics, while at the same time improving the robustness against the different sources of variations.

In-house validation of an LC-MS method for the multiplexed quantitative determination of total allergenic food in chocolate.

Mass spectrometry has been widely accepted as a confirmatory tool for the sensitive detection of undeclared presence of allergenic ingredients. Multiple methods have been developed so far, achieving different levels of sensitivity and robustness, still lacking harmonization of the analytical validation and impairing comparability of results. In this investigation, a quantitative method has been validated in-house for the determination of six allergenic ingredients (cow's milk, hen's egg, peanut, soybean, hazelnut, and almond) in a chocolate-based matrix. The latter has been produced in a food pilot plant to provide a real and well-characterized matrix for proper assessment of method performance characteristics according to official guidelines. In particular, recent considerations issued by the European Committee for Standardization have been followed to guide a rigorous single-laboratory validation and to feature the main method performance, such as selectivity, linearity, and sensitivity. Synthetic surrogates of the peptide markers have been used both in native and labelled forms in matrix-matched calibration curves as external calibrants and internal standards, respectively. A two-order of magnitude range was investigated, focusing on the low concentration range for proper assessment of the detection and quantification limits (LOD and LOQ) by rigorous calibration approach. Conversion factors for all six allergenic ingredients have been determined for the first time to report the final quantitative information as fraction of total allergenic food protein (TAFP) per mass of food (µgTAFP/gfood), since such areporting unit isexploitable in allergenic risk assessment plans. The method achieved good sensitivity with LOD values ranging between 0.08 and 0.2 µgTAFP/gfood, for all ingredients besides egg and soybean, whose quantitative markers reported a slightly higher limit (1.1 and 1.2 µgTAFP/gfood, respectively). Different samples of chocolate bar incurred at four defined concentration levels close to the currently available threshold doses have been analyzed to test the quantitative performance of the analytical method, with a proper estimate of the measurement uncertainty from different sources of variability. The sensitivity achieved resulted in compliance with the various threshold doses issued or recommended worldwide.

Does switching from tobacco to reduced-risk products free up hospital resources?

Promoting a shift from smoking tobacco to reduced-risk products—such as vapes and heat-not-burn tobacco—has the potential to ease the burden on healthcare resources, particularly health expenditure, if empirical evidence shows that reduced-risk products are an effective smoking cessation tool or can help to mitigate the risk of disease. However, there are notable variations in needs and provision of healthcare services between different regions. This article will explore this heterogeneity with the aim of understanding the different health impacts of switching from smoking to reduced-risk products across England, assessing the potential savings for the NHS and potential progress towards the goal of the country becoming smoke-free. This exploratory analysis of different sources of variation across regions offers policy insights to motivate further research.

Genetic diversity of Ethiopian durum wheat (T. turgidum subsp. durum) landraces under water stressed and non stressed conditions

Ethiopia, being a major center of origin and diversity for durum wheat, possesses a highly variable genetic pool with diverse agroecological adaptations. Wheat landraces are an important source of genetic variation for breeding programs. This study was conducted to study the genotypic diversity of Ethiopian durum wheat genetic resources under two contrasting environments namely drought-stressed and non-stressed. It was carried out on 100 landraces and 4 local checks using an augmented design. Data were collected on 13 traits comprising yield and yield components, phenology, and canopy condition. The analysis of variance revealed significant differences between landraces for different traits with different sources of variation. Several landraces were found to outyield the checks at both environmental conditions. Intermediate to high estimates of the phenotypic coefficient of variation (PCV), genotypic coefficient of variation (GCV), heritability in a broad sense (h2b), and genetic advance in percent of the mean (GAPM) were observed for all the studied traits except for days to flowering at normal, thousands seed weight at stress, and days to maturity, leaf chlorophyll concentration measurement, and canopy temperature measurement at both conditions. The estimation of variability parameters showed that genotypic variation was higher than environmental variation for most traits. The number of tillers, spike length, kernel per spike, and grain yield indicated higher values for h2b and GAPM (74.42% and 20.86; 83.2% and 28.24; 70.79% and 28.0; and 89.54% and 74.71) at normal and (97.87% and 98.22; 71.27% and 28.51; 75.52% and 43.9; and 90.04% and 103.68) at the stressed condition, respectively. Spikelets per spike, kernel per spike, and thousands seed weight were positively correlated with grain yield. Grain yield exhibited a weak negative correlation with days to heading and days to maturity. Principal components analysis revealed that six traits were the major loadings on the first two principal components that describe 37.9% and 41.0% of the total morphological variance at normal and stressed conditions, respectively. Cluster analysis grouped the landraces into six clusters, with each cluster showing variation in performance for different traits under normal and stressed conditions. The intracluster distance was maximum in cluster I (D2 = 7.68) and (D2 = 8.19) at normal and stressed conditions respectively and the intercluster distance was found to be maximum between clusters I and IV (D2 = 11.02) and clusters I and II (D2 = 10.33) at normal and stressed conditions respectively. The presence of significant genetic variability among the evaluated durum wheat landraces suggests an opportunity for improvement of grain yield through the hybridization of genotypes from different clusters and subsequent selection. Genotypes with superior agronomic traits that outperform the best checks are identified as potential parents for yield improvement programs for moisture stress.

Multilevel analysis of acoustic variation in a Scinax fuscomarginatus population (Anura, Hylidae) of Central Brazil

The vocalisations of anurans are one of their principal forms of communication and are mainly used for specific recognition involving the attraction of reproductive mates and territorial defense. In this study, we analysed the advertisement calls of 101 individuals from a population of Scinax fuscomarginatus sampled in the type locality of S. pusillus (currently under the synonymy of S. fuscomarginatus). Specifically, we investigated acoustic variation at several levels: intraindividual, interindividual, throughout the night, and across six breeding seasons by analysing temporal and spectral parameters. We identified that all parameters of the advertisement call can be used for individual recognition, with the maximum frequency having the greatest potential. We then observed that all other acoustic parameters were influenced by the predictor variables, with the exception of maximum frequency. The air temperature negatively influenced call duration, number of pulses, dominant frequency and minimum frequency; while it positively influenced pulse rate and call rate during the breeding season. Furthermore, with the exception of call duration and pulse rate, the other acoustic parameters varied significantly across the different nocturnal periods. This study provides data on the variation in S. fuscomarginatus acoustic features. Besides, we also discuss the implications of individual recognition. Studies that consider different sources of variation for the same population of a given species are uncommon, but of paramount importance for understanding the behavioural dynamics of the population. Keywords: advertisement call, behaviour, individual recognition, sexual-selection

Using ACER ConQuest program to examine multidimensional and many-facet models

The main aim of this study was to introduce the ConQuest program, which is used in the analysis of multivariate and multidimensional data structures, and to show its applications on example data structures. To achieve this goal, a basic research approach was applied. Thus, how to use the ConQuest program and how to prepare the data set for analysis were explained step by step. Then, two example applications were made considering the multidimensional structures. Finally, different sources of variability (e.g., item, student, rater, gender), which are both multidimensional and independent of each other, were performed by considering different sources of variability together. According to the analyses, the dimensionality of the data structures must be examined in the analysis process. If the data structure is multidimensional, appropriate multidimensional IRT analyses should be performed.

λ-Factors for the Upper and Lower Bound Analyses of Base-Isolated Structures: Historical Review of Code Provisions for Elastomeric Bearings

The seismic response of base-isolated structures is notably influenced by mechanical properties of isolation devices due to their essential role in structural behavior. Consequently, the variability of such properties should be accounted for in the design process. The current seismic codes prescribe a simplified approach based on structural analyses in two extreme situations resulting from the upper and lower bound design properties of bearings (upper and lower bound analyses). In the case that experimental data are not provided by manufacturers, seismic codes provide the so-called “property modification factors” or “λ-factors”, i.e., modification coefficients to be applied to the nominal dynamic properties of bearings to obtain their upper or lower design properties. The aim of this paper is to provide a historical review of values provided for such factors by the main seismic codes by highlighting the limits, as well as some clerical errors, present in some codes. In particular, the European seismic codes are illustrated in detail, i.e., the Eurocode for bridges (EN 1998-2) and product standard on anti-seismic devices (EN 15129). Both these codes account for different sources of variability, such as the bearings production and the environmental and behavioral effects. For all these effects, the same λ-factor values are provided by the two codes, deriving from the second version of the AASHTO guide specifications for seismic isolation of bridges (AASHTO 1999), which are based on limited and/or old data, especially for high damping rubber bearings (HDRBs), and were never updated in the successive versions. More recent standards are also illustrated, providing different perspectives that deserve attention, even though they require further investigations to be applied in the design practice.

Abstract A15: Computational modeling of methylation impact of AML driver events

Abstract Decades before its clinical onset, epigenetic changes start to accumulate in the progenitor cells of Acute Myelogenous Leukemia– caused by tumor intrinsic driver mutations and tumor extrinsic environmental factors. There are millions of cancer-associated epigenetic changes that are persistent and stable throughout the leukemogenesis and at relapse. By grouping correlated loci into “epigenetic signatures”, we can gain insights on the AML etiology, dynamics as well as mechanisms around the methylation pathways of driver events. We approached this problem in two parts: First, we computationally constructed and analyzed methylation signatures that are associated with DNMT3A and TET2 mutations. Both of them are frequently mutated in AML and they are major causes of changes in the methylation landscape of healthy individuals, conferring a broad risk of transformation to a myeloid malignancy with the serial acquisition of an additional cooperating mutation. We built highly accurate ridge regression models that can differentiate the wild type vs mutant genotypes for DNMT3A and TET2 based on their downstream epigenetic impacts (AUROC 0.9 and 0.8, respectively). It suggests that despite the complicated genetic and epigenetic make-up of post-diagnosis samples, epigenetic regulator mutations’ signatures are distinguishable, and these distinctive marks remain stable over time. We further evaluated methylation loci selected by the models and showed that known downstream targets of DNMT3A such as EVI1/MECOM and AEBP2 (polycomb repressive complex member) were frequently selected by the models. We also identified multiple novel recurrent sites that were previously unassociated with DNMT3A and TET2. Second, we modeled different sources of variations in epigenetic patterns with Topic Modeling. Topic Modeling maps the sparse and high dimensional space of methylation profiles to a small and non-sparse latent space of small number of factors (topics). Each sample is represented as a combination of few topics in that latent space. We annotated these topics with biological and clinical features such as mutation status, prior malignancy, age and gender to shed light on the prevalence and combination of the mutations and other extrinsic factors present at the time of diagnosis. In the future, we aim to identify the etiologies of new signatures to inform detection, monitoring, risk stratifying and drug response studies for AML patients and track these changes throughout leukemogenesis in early detection cohorts. Citation Format: Burcu Gurun, Jeffrey W. Tyner, Emek Demir, Brian J. Druker, Paul T. Spellman. Computational modeling of methylation impact of AML driver events [abstract]. In: Proceedings of the AACR Special Conference: Acute Myeloid Leukemia and Myelodysplastic Syndrome; 2023 Jan 23-25; Austin, TX. Philadelphia (PA): AACR; Blood Cancer Discov 2023;4(3_Suppl):Abstract nr A15.

A comparison of machine learning and statistical species distribution models: Quantifying overfitting supports model interpretation

Species distribution models are commonly applied to predict species responses to environmental conditions. A wide variety of models with different properties exist that vary in complexity, which affects their predictive performance and interpretability. Machine learning algorithms are increasingly used because they are capable to capture complex relationships and are often better in prediction. However, to inform environmental management, it is important that a model predicts well for the right reasons. It remains a challenge to select a model with a reasonable level of complexity that captures the true relationship between the response and explanatory variables as good as possible rather than fitting to the noise in the data.In this study we ask: 1) how much predictive performance can we gain by using increasingly complex models, 2) how does model complexity affect the degree of overfitting, and 3) do the inferred responses differ among models and what can we learn from them? To address these questions, we applied eight models with different complexity to predict the probability of occurrence of freshwater macroinvertebrate taxa based on 2729 Swiss monitoring samples. We compared the models in terms of predictive performance during cross-validation and for generalization out of the calibration domain ("extrapolation" or transferability). We applied model agnostic tools to shed light on model interpretability.Contrary to our expectation, all models predicted similarly well during cross-validation, while no model predicted better than the null model during out-of-domain generalization on average over all taxa. Performance was best for taxa with intermediate prevalence. More complex models predicted slightly better than standard statistical models but were prone to overfitting.Overfitting indicates that a model describes not only the signal in the data but also part of the noise. This impedes the interpretation of response shapes learned by the model, because one cannot distinguish the signal from the noise. Furthermore, the strongly overfitting models learned irregular relationships and strong interactions that are ecologically not plausible. Thus, in this study, the minor gain in predictive performance from more complex models was outweighed by the overfitting.Ecological field data that is used as model input or for calibration is typically prone to different sources of variability, from sampling, the measurement process and stochasticity. We therefore call for caution when using complex data-driven models to learn about species responses or to inform environmental management. In such cases, we recommend to compare a range of models regarding their predictive performance, overfitting and response shapes to better understand the robustness of inferred responses.

Dimensionality of Morphological Knowledge—Evidence from Norwegian Third Graders

AbstractThis study aimed to determine the dimensionality of morphological knowledge by examining different sources of variance. According to the Morphological Pathways Framework (Levesque et al., Journal of Research in Reading, 44, 10–26, 2021), morphological awareness, morphological analysis and morphological decoding are related, but distinct dimensions of morphological knowledge. However, multidimensionality might also stem from construct‐irrelevant variance due to methodological artifacts. We assessed 612 Norwegian third graders on five measures of morphological knowledge and one measure of general vocabulary. Fitting a series of confirmatory factor analysis (CFA) models, we evaluated the dimensionality of morphological knowledge both within and across the five tests. Furthermore, we fitted three structural equation models (SEMs) to explore how different conceptualizations affect the relationship between morphological knowledge and general vocabulary: a five‐factor model, a bifactor model, and a higher‐order model representing morphological awareness, morphological analysis and morphological decoding. CFAs supported a multidimensional view of morphological knowledge and highlighted the need to account for construct‐irrelevant variance. SEM analyses further illustrated that construct‐irrelevant variance introduces a confounding element to the relations between morphological knowledge and vocabulary in the test‐specific five‐factor model, as only the bifactor and higher‐order models separate between construct‐relevant variance and variance due to methodological artifacts. The bifactor model is useful for separating sources of variance, especially during test development. For research purposes, however, we recommend conceptualizing morphological knowledge in line with Levesque et al., Journal of Research in Reading, 44, 10–26, 2021, to increase knowledge of morphological dimensions and their relations to other areas of literacy.

Assessment of Variability Sources in Grape Ripening Parameters by Using FTIR and Multivariate Modelling.

The variability in grape ripening is associated with the fact that each grape berry undergoes its own biochemical processes. Traditional viticulture manages this by averaging the physicochemical values of hundreds of grapes to make decisions. However, to obtain accurate results it is necessary to evaluate the different sources of variability, so exhaustive sampling is essential. In this article, the factors "grape maturity over time" and "position of the grape" (both in the grapevine and in the bunch/cluster) were considered and studied by analyzing the grapes with a portable ATR-FTIR instrument and evaluating the spectra obtained with ANOVA-simultaneous component analysis (ASCA). Ripeness over time was the main factor affecting the characteristics of the grapes. Position in the vine and in the bunch (in that order) were also significantly important, and their effect on the grapes evolves over time. In addition, it was also possible to predict basic oenological parameters (TSS and pH with errors of 0.3 °Brix and 0.7, respectively). Finally, a quality control chart was built based on the spectra obtained in the optimal state of ripening, which could be used to decide which grapes are suitable for harvest.