Regression-based Estimation Research Articles

A regression-based approach to the CO2 airborne fraction

The global fraction of anthropogenically emitted carbon dioxide (CO2) that stays in the atmosphere, the CO2 airborne fraction, has been fluctuating around a constant value over the period 1959 to 2022. The consensus estimate of the airborne fraction is around 44%. In this study, we show that the conventional estimator of the airborne fraction, based on a ratio of changes in atmospheric CO2 concentrations and CO2 emissions, suffers from a number of statistical deficiencies. We propose an alternative regression-based estimator of the airborne fraction that does not suffer from these deficiencies. Our empirical analysis leads to an estimate of the airborne fraction over 1959–2022 of 47.0% (± 1.1%; 1σ), implying a higher, and better constrained, estimate than the current consensus. Using climate model output, we show that a regression-based approach provides sensible estimates of the airborne fraction, also in future scenarios where emissions are at or near zero.

Causal impact of statins on susceptibility to osteoarthritis: insights from a two-sample Mendelian randomization analysis.

Osteoarthritis is a widely prevalent cause of pain and disability among older adults. It is an incurable condition, and most treatments are aimed at alleviating symptoms. This study aimed to investigate the impact of statins on osteoarthritis by using a two-sample Mendelian randomization approach, using genetic variants associated with statin use as instrumental variables. Information on single nucleotide polymorphisms associated with statin medication was obtained from the FinnGen study, and data on osteoarthritis were sourced from the UK Biobank. The inverse variance weighted method was used as the primary analytical approach for the Mendelian randomization analysis. Sensitivity analyses were conducted to evaluate horizontal pleiotropy and heterogeneity. To examine the genetic relationship between statins and osteoarthritis, linkage disequilibrium score regression-based estimates were used. Mendelian randomization analysis indicated a positive effect of statin use on the treatment of osteoarthritis (odds ratio 0.951, 95% confidence interval 0.914-0.99, p < 0.05). This conclusion was supported by various Mendelian randomization methods. Sensitivity analyses revealed no significant directional pleiotropy or influential single nucleotide polymorphisms that could compromise the overall causal inference. Linkage disequilibrium score regression-based estimates suggested a modest genetic correlation between statin use and osteoarthritis (Rg = 0.098, Se = 0.034, p < 0.05), thus reinforcing the robustness of the Mendelian randomization analysis. Statins reduce the risk of osteoarthritis, aligning with the results of observational studies. Further research is essential to validate these results and explore the underlying mechanisms in detail.

Inflation Expectations with Finite Horizon Planning

Under finite horizon planning, households and firms evaluate a full set of state-contingent paths along which the economy might evolve out to a finite horizon but have limited ability to process events beyond that horizon. We show--analytically and empirically--that such a model accounts for an initial underreaction and subsequent overreaction of inflation forecasts. A planning horizon of four quarters can account for the evidence on the predictability of inflation forecast errors and macroeconomic data. Our identification and estimation strategies combine full-information methods based on aggregate data with regression-based estimates that directly use inflation expectations data.

Health Insurance Type and Outpatient Specialist Care Among Children With Asthma

Although children with asthma are often successfully treated by primary care clinicians, outpatient specialist care is recommended for those with poorly controlled disease. Little is known about differences in specialist use for asthma among children with Medicaid vs private insurance. To examine differences among children with asthma regarding receipt of asthma specialist care by insurance type. In this cross-sectional study using data from the Massachusetts All Payer Claims Database (APCD) between 2014 to 2020, children with asthma were identified and differences in receipt of outpatient specialist care by whether their insurance was public (Medicaid and the Children's Health Insurance Program) or private were examined. Eligible participants included children with asthma in 2015 to 2020 aged 2 to 17 years. Data analysis was conducted from January 2023 to April 2024. Medicaid vs private insurance. The primary outcome was receipt of specialist care (any outpatient visit with a pulmonology, allergy and immunology, or otolaryngology physician). Multivariable logistic regression models estimated differences in receipt of specialist care by insurance type accounting for child and area characteristics including demographics, health status, persistent asthma, calendar year, and zip code characteristics. Additional analyses examined if the associations of specialist care with insurance type varied by asthma persistence and severity, and whether associations varied over time. Among 198 101 unique children, there were 432 455 child-year observations (186 296 female [43.1%] and 246 159 male [56.9%]; 211 269 aged 5 to 11 years [48.9%]; 82 108 [19.0%] with persistent asthma) including 286 408 (66.2%) that were Medicaid insured and 146 047 (33.8%) that were privately insured. Although persistent asthma was more common among child-year observations with Medicaid vs private insurance (57 381 [20.0%] vs 24 727 [16.9%]), children with Medicaid were less likely to receive specialist care. Overall, 64 239 child-year observations (14.9%) received specialist care, with substantially lower rates for children with Medicaid vs private insurance (34 093 child-year observations [11.9%] vs 30 146 child-year observations [20.6%]). Regression-based estimates confirmed these disparities; children with Medicaid had 55% lower odds of receiving specialist care (adjusted odds ratio, 0.45; 95% CI, 0.43 to 0.47) and a regression-adjusted 9.7 percentage point (95% CI, -10.4 percentage points to -9.1 percentage points) lower rate of receipt of specialist care. Compared with children with private insurance, there was an additional 3.2 percentage point (95% CI, 2.0 percentage points to 4.4 percentage points) deficit for children with Medicaid with persistent asthma. In this cross-sectional study, children with Medicaid were less likely to receive specialist care, with the largest gaps among those with persistent asthma. These findings suggest that closing this care gap may be one approach to addressing ongoing disparities in asthma outcomes.

Exploring the impact of athlete brand image on fans’ behavioural outcomes: the role of emotional attachment and perceived price value

PurposeThis study aims to explore the impact of athlete brand image on fans’ social media engagement, purchase intentions, and also examines the mediating role of emotional attachment on these relationships, as well as the moderating role of perceived price value between emotional attachment and purchase intentions.Design/methodology/approachThe data are based on an online survey conducted in China (N = 572). The PLS-SEM (partial least squares structural equation model) and regression-based estimation method (PROCESS) are employed to test the hypotheses.FindingsThe results indicate a positive relationship between athletes’ athletic performance and fans’ social media engagement as well as purchase intentions. The impact of athletes’ attractive appearance and marketable lifestyle on fans’ purchase intentions is sequentially mediated by emotional attachment and social media engagement. Moreover, the mediation effect of athletes’ off-field image and purchase intentions is contingent upon fans’ perceived price value.Research limitations/implicationsAthletes and marketers could integrate and leverage both the on-field and off-field attributes to cultivate emotional connections with fans. Sports organizations and managers need to pay attention to fans’ social media engagement and provide content that increases engagement and converts into transactional behavioural intentions.Originality/valueThe study provides empirical evidence of the mediating role of emotional attachment between athlete brand image and fans’ purchase intentions. The explanatory mechanisms involving emotional attachment and social media engagement (non-transactional behavioural intentions) are anticipated to be a noteworthy addition to the traditional fan transactional behavioural intentions framework. Moreover, the research introduces and confirms perceived price value as a crucial moderating factor influencing the relationship between emotional attachment and purchase intentions.

Revisiting Event-Study Designs: Robust and Efficient Estimation

Abstract We develop a framework for difference-in-differences designs with staggered treatment adoption and heterogeneous causal effects. We show that conventional regression-based estimators fail to provide unbiased estimates of relevant estimands absent strong restrictions on treatment-effect homogeneity. We then derive the efficient estimator addressing this challenge, which takes an intuitive “imputation” form when treatment-effect heterogeneity is unrestricted. We characterize the asymptotic behaviour of the estimator, propose tools for inference, and develop tests for identifying assumptions. Our method applies with time-varying controls, in triple-difference designs, and with certain non-binary treatments. We show the practical relevance of our results in a simulation study and an application. Studying the consumption response to tax rebates in the U.S., we find that the notional marginal propensity to consume is between 8 and 11% in the first quarter—about half as large as benchmark estimates used to calibrate macroeconomic models—and predominantly occurs in the first month after the rebate.

Regression-based estimation of heteroscedastic term structure models

Regression-based estimation of heteroscedastic term structure models

Modified robust regression-type estimators with multi-auxiliary variables using non-conventional measures of dispersion

Auxiliary variables correlated with the study variables have been identified to be useful in improving the efficiency of ratio, product and regression estimators both at planning and estimation stages. The existing regression-based estimators are functions of auxiliary variables which are sensitive to outliers. In this paper, a modified class of estimators is proposed using robust non-conventional measures of dispersion which are robust against outliers or extreme values. The properties (Biases and Mean Squared Errors (MSEs)) of the modified class of estimator were derived up to the first order of approximation using Taylor series approach. The empirical studies were conducted using stimulation to investigate the efficiency of the proposed estimators over the efficiency of the existing estimators. The results revealed that the proposed estimators have minimum MSEs and higher Percentage Relative Efficiencies (PREs) among all the competing estimators. These results implied that the proposed estimators are more efficient and can produce better estimate of the population mean compared to other existing estimators considered in the study. Therefore, it can be concluded that proposed estimators have better predictive power for estimating population mean when the study (interest) variables are characterized with outliers or extreme values.

Self-supervised deep monocular visual odometry and depth estimation with observation variation

Recent research in deep learning has greatly advanced the development of monocular visual depth estimation and odometry. These learning-based methods predominantly leverage neural networks to extract prevalent high-dimensional features for regression-based estimation. However, the consistency of observations is still regarded as a core challenge that limits the generalization ability. We attribute this challenge to the significant inconsistency in feature levels due to changes in the environment or motion, collectively referred to as observation variation. To this end, we propose a novel monocular depth estimation and a pose estimation network with attention mechanism to recalibrate features for different environment. The depth and ego-motion estimation tasks are formulated by coupling together in an end-to-end reconstruction problem, and attention modules adaptively recalibrate specific feature responses. Furthermore, we propose a simple sliding window optimization without RNN-based architecture, which is introduced to overcome the movement speed variation and error accumulation. Finally, we collect an observation variation dataset with environmental and motion changes. Evaluations of depth and pose estimation demonstrate that our methods effectively overcomes these observation variation challenges and achieves better performance. Our project will be available at https://github.com/alikesierzhao/OV-SfmLearner.

Regression-based multiple treatment effect estimation under covariate-adaptive randomization.

Covariate-adaptive randomization methods are widely used in clinical trials to balance baseline covariates. Recent studies have shown the validity of using regression-based estimators for treatment effects without imposing functional form requirements on the true data generation model. These studies have had limitations in certain scenarios; for example, in the case of multiple treatment groups, these studies did not consider additional covariates or assumed that the allocation ratios were the same across strata. To address these limitations, we develop a stratum-common estimator and a stratum-specific estimator under multiple treatments. We derive the asymptotic behaviors of these estimators and propose consistent nonparametric estimators for asymptotic variances. To determine their efficiency, we compare the estimators with the stratified difference-in-means estimator as the benchmark. We find that the stratum-specific estimator guarantees efficiency gains, regardless of whether the allocation ratios across strata are the same or different. Our conclusions were also validated by simulation studies and a real clinical trialexample.

The effect of back muscle fatigue on EMG and kinematics based estimation of low-back loads and active moments during manual lifting tasks

This study investigated the effects of back muscle fatigue on the estimation of low-back loads and active low-back moments during lifting, using an EMG and kinematics based model calibrated with data from an unfatigued state. Fourteen participants performed lifting tasks in unfatigued and fatigued states. Fatigue was induced through semi-static forward bending. EMG, kinematics, and ground reaction forces were measured, and low-back loads were estimated using inverse dynamics and EMG-driven muscle model. A regression model was developed using data from a set of calibration lifts, and its accuracy was evaluated for unfatigued and fatigued lifts. During the fatigue-inducing task, the EMG amplitude increased by 2.8 %MVC, representing a 38% increase relative to the initial value. However, during the fatigued lifts, the peak EMG amplitude was found to be 1.6 %MVC higher than that observed during the unfatigued lifts, representing a mere 4% increase relative to the baseline unfatigued peak EMG amplitude. Kinematics and low-back load estimates remained unaffected. Regression model estimation errors remained unaffected for 5 kg lifts, but increased by no more than 5% of the peak active low-back moment for 15 kg lifts. We conclude that the regression-based estimation quality of active low-back moments can be maintained during periods of muscle fatigue, although errors may slightly increase for heavier loads.

Multi-Segment State of Health Estimation of Lithium-ion Batteries Considering Short Partial Charging

State of health (SOH) is a critical state parameter of lithium-ion batteries (LIBs). Health indicators (HIs), which are derived from the measured features of LIBs, are used in the current data-driven SOH estimation techniques to determine SOH. However, the common partial charging and discharging make it challenging to derive reliable HIs. In this paper, a SOH estimation approach considering short partial charging is proposed. Unlike other techniques, the constant current charging stage is divided into short segments, the HI, based on the charging capacity and actual initial charging voltage, is extracted within each short segment, and a kernel ridge regression-based estimator is created to characterize the SOH mapping relationship. Subsequently, an estimator fusion frame is established to merge the estimates of the eligible segments, which is decided based on the actual start and end charging voltages of the partial charging. The effectiveness of the proposed approach is validated with two well-known LIBs aging datasets containing real partial charging cycles. The results are satisfactory in terms of accuracy, robustness to partial charging, and good generality to different types of LIBs. Effective SOH value can be deduced whenever the charging voltage range covers at least one short estimation segment.

Comparing harmonic regression and GLAD Phenology metrics for estimation of forest community types and aboveground live biomass within forest inventory and analysis plots

The Landsat multispectral time series is a valuable source of moderate spatial resolution data to support forest mapping and monitoring tasks. Using United States Department of Agriculture (USDA) Forest Service Forest Inventory and Analysis (FIA) plots within the states of Michigan, Oregon, and West Virginia, two methods to summarize time series observations, harmonic regression coefficients and Global Land Analysis & Discovery (GLAD) Phenology metrics, are compared for predicting forest community type, total aboveground live biomass (AGLBM), and species-specific AGLBM. Harmonic regression coefficients, which provided mean overall accuracies (OAs) between 62.8% and 73.1% and map image classification efficacies (MICEs) varying between 0.455 and 0.566 for the three studied states and calculated using multiple model replicates, generally provided better predictive performance for differentiating forest community types in comparison to GLAD phenology metrics. However, differences were not always statistically significant. DTM-derived terrain variables improved classification performance in some landscapes when using machine learning, random forest classification (for example, the highest obtained mean OA of 78.5% (MICE = 0.557) was obtained for West Virginia when using the combined harmonic regression and terrain variables). For the regression-based estimation of total and species-specific AGLBM, the spectral data used and the incorporation of terrain variables had less of an impact on model performance. Further, the regression models in Oregon provided larger mean R-squared values (0.452 to 0.490) in comparison to those in Michigan and West Virginia, where all R-squared values were lower than 0.200, suggesting that AGLBM prediction may be more or less challenging in different landscapes depending on forest characteristics, terrain, management practices, and disturbance histories.

An inverse probability weighted regression method that accounts for right-censoring for causal inference with multiple treatments and a binary outcome.

Comparative effectiveness research often involves evaluating the differences in the risks of an event of interest between two or more treatments using observational data. Often, the post-treatment outcome of interest is whether the event happens within a pre-specified time window, which leads to a binary outcome. One source of bias for estimating the causal treatment effect is the presence of confounders, which are usually controlled using propensity score-based methods. An additional source of bias is right-censoring, which occurs when the information on the outcome of interest is not completely available due to dropout, study termination, or treatment switch before the event of interest. We propose an inverse probability weighted regression-based estimator that can simultaneously handle both confounding and right-censoring, calling the method CIPWR, with the letter C highlighting the censoring component. CIPWR estimates the average treatment effects by averaging the predicted outcomes obtained from a logistic regression model that is fitted using a weighted score function. The CIPWR estimator has a double robustness property such that estimation consistency can be achieved when either the model for the outcome or the models for both treatment and censoring are correctly specified. We establish the asymptotic properties of the CIPWR estimator for conducting inference, and compare its finite sample performance with that of several alternatives through simulation studies. The methods under comparison are applied to a cohort of prostate cancer patients from an insurance claims database for comparing the adverse effects of four candidate drugs for advanced stage prostate cancer.

Measures of vertical inequality in assessments

Standard measures of vertical inequality suggest that assessments are regressive in the sense that high-priced properties are often assessed at lower rates than low-priced properties. We show that some of this regressivity is due to the regression-based estimation procedures used by many jurisdictions to calculate assessments. A review of existing measures of assessment regressivity suggests that severe biases associated with regression-based procedures make them much less useful than the traditional price-related differential (PRD) as a measure of vertical inequality. To supplement existing measures of vertical inequality, we propose approaches using Gini coefficients, Suits index, and kernel density tests to provide information on the relationship between the assessment and sale price distributions. We compute the measures using data on sales prices and assessments for 48 large central city counties. The results suggest that the PRD remains a useful approach for measuring vertical inequality due to its simplicity and familiarity, while distribution-based procedures are helpful because they are not as sensitive to small numbers of very high-priced sales. Together, the approaches provide a more complete picture of how assessment rates vary across the full distribution of sales prices.

Using the Personality Assessment Inventory to Assess the Alternative Model for Personality Disorders: Criterion Validity in a Clinical Sample

The Personality Assessment Inventory (PAI) is a broadband measure of psychopathology that is widely used in applied settings. Researchers developed regression-based estimates that use the PAI to measure constructs of the Alternative Model for Personality Disorders (AMPD) – a hybrid dimensional and categorical approach to conceptualizing personality disorders. Although prior work has linked these estimates to formal measures of the AMPD, there is little work investigating the clinical correlates of this scoring approach of the PAI. The current study examines associations between these PAI-based AMPD estimates and life data in a large, archival dataset of psychiatric outpatients and inpatients. We found general support for the criterion validity of AMPD estimate scores, such that a theoretically consistent pattern of associations emerged with indicators such as prior academic achievement, antisocial behavior, psychiatric history, and substance abuse. These results provide preliminary support to this scoring approach for use in clinical samples.

Growth failure among children of adolescent mothers at ages 0-5 and 6-12 years in India.

Adolescent motherhood has been linked with poor health outcomes at birth for children, including high neonatal mortality, low birthweight, and small-for-gestational-age rates. However, longer-term growth outcomes in the children of adolescent mothers in low-resource settings remain inadequately studied. We used longitudinal data from the India Human Development Surveys, 2004-2005 and 2011-2012 (n = 12,182) and employed regression and propensity score matching analysis to compare the following growth indicators of children born to adolescent mothers (ages 19 years or below) with those born to older mothers. Growth indicators included height and weight during ages 0-5 years and 6-12 years and change in height and weight between the two periods. In regression-based estimates, children born to adolescent mothers were 0.01 m shorter and weighed 0.2kg less than children of older mothers at ages 0-5 years. At ages 6-12 years, those born to adolescent mothers were 0.02 m shorter and weighed 0.97kg less. The height difference between the two groups increased by 0.01 m and the weight difference grew by 0.77kg over time. Height and weight difference between the two groups worsened among boys over time, while for girls, only the weight gap worsened. The results were similar when using propensity score matching methods. Public policies for reducing child marriage, combined with targeted health, nutrition, and well-being programs for adolescent mothers, are essential for both preventing adolescent childbearing and reducing its impact on growth failure among children in India.

How to Design a State Education Aid Formula Using a Regression-based Estimate of the Cost-capacity Gap: The Case of Connecticut, USA

How to Design a State Education Aid Formula Using a Regression-based Estimate of the Cost-capacity Gap: The Case of Connecticut, USA

American College of Surgeons NSQIP Risk Calculator Accuracy Using a Machine Learning Algorithm Compared with Regression.

The American College of Surgeons NSQIP risk calculator (RC) uses regression to make predictions for fourteen 30-day surgical outcomes. While this approach provides accurate (discrimination and calibration) risk estimates, they might be improved by machine learning (ML). To investigate this possibility, accuracy for regression-based risk estimates were compared to estimates from an extreme gradient boosting (XGB)-ML algorithm. A cohort of 5,020,713 million NSQIP patient records was randomly divided into 80% for model construction and 20% for validation. Risk predictions using regression and XGB-ML were made for 13 RC binary 30-day surgical complications and one continuous outcome (length of stay [LOS]). For the binary outcomes, discrimination was evaluated using the area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC), and calibration was evaluated using Hosmer-Lemeshow statistics. Mean squared error and a calibration curve analog were evaluated for the continuous LOS outcome. For every binary outcome, discrimination (AUROC and AUPRC) was slightly greater for XGB-ML than for regression (mean [across the outcomes] AUROC was 0.8299 vs 0.8251, and mean AUPRC was 0.1558 vs 0.1476, for XGB-ML and regression, respectively). For each outcome, miscalibration was greater (larger Hosmer-Lemeshow values) with regression; there was statistically significant miscalibration for all regression-based estimates, but only for 4 of 13 when XGB-ML was used. For LOS, mean squared error was lower for XGB-ML. XGB-ML provided more accurate risk estimates than regression in terms of discrimination and calibration. Differences in calibration between regression and XGB-ML were of substantial magnitude and support transitioning the RC to XGB-ML.

Software project estimation using smooth curve methods and variable selection and regularization methods using a wedge-shape form database

Context: The impact of an excellent estimation in planning, budgeting, and control, makes the estimation activities an essential element for the software project success. Several estimation techniques have been developed during the last seven decades. Traditional regression-based is the most often estimation method used in the literature. The generation of models needs a reference database, which is usually a wedge-shaped dataset when real projects are considered. The use of regression-based estimation techniques provides low accuracy with this type of database. Objective: Evaluate and provide an alternative to the general practice of using regression-based models, looking if smooth curve methods and variable selection and regularization methods provide better reliability of the estimations based on the wedge-shaped form databases. Method: A previous study used a reference database with a wedge-shaped form to build a regression-based estimating model. This paper utilizes smooth curve methods and variable selection and regularization methods to build estimation models, providing an alternative to linear regression models. Results: The results show the improvement in the estimation results when smooth curve methods and variable selection and regularization methods are used against regression-based models when wedge-shaped form databases are considered. For example, GAM with all the variables show that the R-squared is for Effort: 0.6864 and for Cost: 0.7581; the MMRE is for Effort: 0.1095 and for Cost: 0.0578. The results for the GAM with LASSO show that the R-squared is for Effort: 0.6836 and for Cost: 0.7519; the MMRE is for Effort: 0.1105 and for Cost: 0.0585. In comparison to the R-squared is for Effort: 0.6790 and for Cost: 0.7540; the MMRE is for Effort: 0.1107 and for Cost: 0.0582 while using MLR.