Two-stage Regression Model Research Articles

Physical Activity by Accelerometry is Prospectively Associated With Lower Testosterone and Hirsutism Score Among Adolescent Girls.

Physical activity (PA) is recommended for the management of women with polycystic ovary syndrome (PCOS). However, it is not known whether PA modulates the risk of components of PCOS diagnostic criteria during adolescence. We included participants from the Avon Longitudinal Study of Parents and Children prospective cohort with objective measurement of PA and sedentary behavior using accelerometry between ages 11 and 15years, biochemical and clinical measures of hyperandrogenism (serum total testosterone level at 15years and self-reported hirsutism at 19years), and self-reported menstrual regularity at 17years. Two-stage regression models evaluated the association of repeated measures of PA and sedentary behavior with components of PCOS diagnostic criteria. Among 1,526 singleton girls with accelerometer data and testosterone samples, time spent in light PA was inversely associated with total testosterone (ß=-0.070, 95% CI-0.114,-0.025; nmol/L, per 60minutes) and continuous hirsutism score (ß=-0.756, 95% CI-1.41,-0.103; per 60minutes). Conversely, time spent sedentary was associated with higher total testosterone level (ß= 0.052, 95% CI 0.013, 0.090; per 60minutes) and hirsutism score (ß= 0.668, 95% CI 0.103, 1.23; per 60minutes). Associations were independent of adiposity, but results for hirsutism were attenuated following adjustment for age at menarche and after stratification by hormonal contraceptive use. No associations were observed with menstrual irregularity, for other PA intensity components, or among users of hormonal contraceptives. More time spent in light intensity physical activity and less time spent sedentary during early adolescence is associated with lower testosterone concentrations and lower hirsutism scores in late adolescence.

Associations of thyroid hormones with chronic diarrhea and constipation in euthyroid individuals.

Abnormalities in thyroid function affect bowel health. However, the relationships between thyroid hormone concentrations and the risk of developing chronic diarrhea and constipation remain unclear. Thus, the aim of this study was to investigate the relationships between thyroid hormone concentrations and the risk of developing chronic diarrhea and constipation in euthyroid US adults. The data for this population-based study were taken from the National Health and Nutrition Examination Survey (NHANES) 2007-2010 datasets. The relationships between thyroid hormone concentrations and the risk of developing chronic diarrhea and constipation were examined via multivariate regression. Smoothed curve fitting and threshold effects analysis were used to test for nonlinear relationships and inflection points. This study involved 4999 participants ranging in age from 20 to 80 years. Multivariate logistic regression analysis revealed a significant positive correlation between FT3 concentrations and the risk of developing chronic diarrhea [1.37 (1.00, 1.88), P=0.049]. Multivariate linear regression analysis revealed a significant positive correlation between FT3 concentrations and the number of bowel movements [0.84 (0.39, 1.28), P<0.001]. Using smoothed curve fitting and the two-stage regression model, we found a nonlinear relationship between FT4 concentrations and chronic diarrhea, with a breakpoint of 0.79 ng/dl. There were associations between thyroid hormone concentrations and abnormal bowel habits, particularly between FT3 concentrations and the risk of developing chronic diarrhea. A higher FT3 level was associated with an increased risk of developing chronic diarrhea and more frequent bowel movements. To validate our results, further large-scale prospective studies are needed.

Unobtrusive Skin Temperature Estimation on a Smart Bed.

The transition from wakefulness to sleep occurs when the core body temperature decreases. The latter is facilitated by an increase in the cutaneous blood flow, which dissipates internal heat into the micro-environment surrounding the sleeper's body. The rise in cutaneous blood flow near sleep onset causes the distal (hands and feet) and proximal (abdomen) temperatures to increase by about 1 °C and 0.5 °C, respectively. Characterizing the dynamics of skin temperature changes throughout sleep phases and understanding its relationship with sleep quality requires a means to unobtrusively and longitudinally estimate the skin temperature. Leveraging the data from a temperature sensor strip (TSS) with five individual temperature sensors embedded near the surface of a smart bed's mattress, we have developed an algorithm to estimate the distal skin temperature with a minute-long temporal resolution. The data from 18 participants who recorded TSS and ground-truth temperature data from sleep during 14 nights at home and 2 nights in a lab were used to develop an algorithm that uses a two-stage regression model (gradient boosted tree followed by a random forest) to estimate the distal skin temperature. A five-fold cross-validation procedure was applied to train and validate the model such that the data from a participant could only be either in the training or validation set but not in both. The algorithm verification was performed with the in-lab data. The algorithm presented in this research can estimate the distal skin temperature at a minute-level resolution, with accuracy characterized by the mean limits of agreement [-0.79 to +0.79 °C] and mean coefficient of determination R2=0.87. This method may enable the unobtrusive, longitudinal and ecologically valid collection of distal skin temperature values during sleep. Therelatively small sample size motivates the need for further validation efforts.

Compensation incentive, misappropriation risk, and startup innovation: the role of corporate venture capital manager in resource transfer

ABSTRACT Existing studies on the impact of corporate venture capital (CVC) on startup innovation primarily focus on the capability of CVC parent to provide resources for startups. This study examines the role of CVC managers’ motivation in facilitating resource transfer to startups. Utilizing a unique disclosure requirement for investments by Chinese public firms in venture capital (VC) funds, and a hand-collected database on CVC managers’ compensation, our study explores the impact of compensation incentives on CVC managers with respect to startup innovation and the moderating effect of misappropriation risk by the CVC parent. Our two-stage regression model reveals that stronger compensation incentives positively impact startup innovation. Moreover, the effects diminish if the misappropriation risk by the CVC parent increases. These findings underscore that while strong compensation incentives for CVC managers enhance their motivation to facilitate resource transfer from the CVC parent to startups, this is mitigated in contexts with a heightened misappropriation risk.

Measuring the energy poverty premium in Great Britain and identifying its main drivers based on longitudinal household survey data

Measuring the energy poverty premium in Great Britain and identifying its main drivers based on longitudinal household survey data

Addressing causal relationship between drinking behavior and metabolic syndrome: one-sample Mendelian randomization analysis.

Alcohol drinking is associated with central obesity, hypertension, and hyperlipidemia, which further causes metabolic syndrome (MetS). However, prior epidemiological studies on such associations lack experimental evidence for a causal relationship. This study aims to explore the causal relationship between drinking behavior and MetS in Taiwan population by using Mendelian randomization (MR) analysis. A cross-sectional study was conducted using the Taiwan Biobank database, which comprised 50 640 Han Chinese who were 30-70years old without cancer from 2008 to 2020. In MR analysis, we constructed weighted and unweighted genetic risk scores by calculating SNP alleles significantly associated with alcohol drinking. We calculated odds ratios and 95% confidence interval (CI) by using a two-stage regression model. A total of 50 640 participants were included with a mean age of 49.5years (SD: 1.67years), 36.6% were men. The adjusted odds ratio (aOR) of MetS per 5% increase in the likelihood of genetic predisposition to drink based on weighted genetic risk score with adjustment was 1.11 (95% CI: 1.10, 1.12, P < .001). Analysis was also conducted by grouping the likelihood of genetic predisposition to drink based on quartiles with multivariate adjustment. Using Q1 as the reference group, the aORs of MetS for Q2, Q3, and Q4 were 1.19 (1.12, 1.27, p < .001), 1.31 (1.23, 1.40, p < .001), and 1.87 (1.75, 2.00, p < .001), respectively, for the weighted genetic risk score. This study shows a modest relationship between drinking behavior and MetS by using MR analysis.

Risk of cardiovascular and respiratory diseases attributed to satellite-based PM2.5 over 2017-2022 in Sanandaj, an area of Iran.

The risk of cardiovascular and respiratory diseases attributed to satellite-based PM2.5 has been less investigated. In this study, the attributable risk was estimated in an area of Iran. The predicted air PM2.5 using satellite data and a two-stage regression model was used as the predictor of the diseases. The dose-response linkage between the bias-corrected predictor employing a strong statistical approach and the outcomes was evaluated using the distributed lag nonlinear model. We considered two distinct scenarios of PM2.5 for the risk estimation. Alongside the risk, the attributable risk and number were estimated for different levels of PM2.5 by age and gender categories. The cumulative influence of PM2.5 particles on respiratory illnesses was statistically significant at 13-16µg/m3 relative to the reference value (median), mostly apparent in the middle delays. The cumulative relative risk of 90th and 95th percentiles were 2.03 (CI 95%: 1.28, 3.19) and 2.25 (CI 95%: 1.28, 3.96), respectively. Nearly 600 cases of the diseases were attributable to the non-optimum values of the pollutant during 2017-2022, of which more than 400 cases were attributed to high values range. The predictor's influence on cardiovascular illnesses was along with uncertainty, indicating that additional research into their relationship is needed. The bias-corrected PM2.5 played an essential role in the prediction of respiratory illnesses, and it may likely be employed as a trigger for a preventative strategy.

Estimation of PM2.5 using high-resolution satellite data and its mortality risk in an area of Iran

ABSTRACT Satellite-based exposure of fine particulate matters has been seldom used as a predictor of mortality. PM2.5 was predicted using Aerosol Optical Depths (AOD) through a two-stage regression model. The predicted PM2.5 was corrected for the bias using two approaches. We estimated the impact by two different scenarios of PM2.5 in the model. We statistically found different distributions of the predicted PM2.5 over the region. Compared to the reference value (5 µg/m3), 90th and 95th percentiles had significant adverse effect on total mortality (RR 90th percentile:1.45; CI 95%: 1.08–1.95 and RR 95th percentile:1.53; CI 95%: 1.11–2.1). Nearly 1050 deaths were attributed to any range of the air pollution (unhealthy range), of which more than half were attributed to high concentration range. Given the adverse effect of extreme values compared to the both scenarios, more efforts are suggested to define local-specific reference values and preventive strategies.

Customer Concentration and Labor Investment Efficiency: Evidence from China

ABSTRACT: The labor force constitutes a paramount factor of production in the realm of business operations. In the past, China's abundant labor force significantly boosted productivity for firms. However, as the demographic dividend gradually wanes, concerns have arisen regarding a scarcity of labor force, exacerbating worries about rising labor costs. Therefore, focusing on enhancing firms' labor investment efficiency becomes pivotal. Using a sample of A-stock listed firms in China from 2013 to 2020, we examine the impact of customer concentration on the labor investment efficiency of supplier firms. Prior literature provides mixed results regarding the effect of concentrated customers. They can act as a governance mechanism for reducing agency problems, thus, increasing investment efficiency. Conversely, agency problems are aggravated in suppliers with concentrated customers, consequently, reducing investment efficiency. Our results suggest that customer concentration reduces labor investment efficiency, and this effect is more pronounced when customers possess high bargaining power. Additionally, the mechanism analysis reveal that customer concentration leads to less accurate information disclosure, higher operating risk, and an incentive to "empire building", reducing labor investment efficiency. The cross-sectional analysis reveals that customer concentration results in both over- and under-investment in labor, thereby reducing investment efficiency. In addition, we employ augmented models to rule out the possibility of a U-shaped relationship between customer concentration and labor investment efficiency. Furthermore, we adopt the instrumental variables approach as well as a two-stage regression model to address potential endogeneity concerns and mitigate the omitted variable concern. Our results hold after the robustness tests and endogeneity tests. The findings of this paper imply that firms should strategically diversify their customer bases, thereby reducing their reliance on a few large customers. Simultaneously, governments should actively encourage firms to broaden their customer base. It can help enhance labor investment efficiency by spreading the risks associated with customer concentration. Moreover, it is crucial to acknowledge that substantial customer bargaining power can negatively impact supplier firms. Thus, policymakers should promote antitrust regulations and fair trade practices to mitigate the high bargaining power of large customers.

The Impact of the Education Factor on Household Saving

Many factors influence the formation of consumer habits, one of the key qualitative variables being the education factor. The article examines the influence of the mentioned factor, both on the example of Georgia for different groups of income levels, and on the basis of comparative analysis between countries. The method of regression analysis is used as a research method. In particular, on the example of Georgia, the two-stage regression model is used. In the first stage, the savings factor is filtered from the influence of other factors. The second models are built for between the residual variable of the first stage equation and the education variable. The panel regression model with fixed period effects and period-weighted parameters is used for comparative analysis between countries. The models built on the example of Georgia do not unambiguously show the importance of the education factor in relation to the variation of the savings rate. In addition, on the studied 12-year time horizon, which is a short-term period for revealing the relationship between the mentioned factors, the education factor shows an inversely proportional impact on the savings rate. A cross-country panel model better identifies long-term factors and shows that increasing years of schooling increases the savings rate. Thus, the research shows that the education factor significantly improves financial decision-making abilities. Increasing choice in financial decisions is an important indicator of economic sustainability. Keywords: savings rate, education, gross national income.

Housing affordability and housing demand assessment for urban poor in India using the hedonic model

PurposeThis study aims to assess housing affordability and estimate demand using a hedonic regression model in the context of Lucknow city, India. This study assesses housing affordability by considering various housing and household-related variables. This study focuses on the impoverished urban population, as they experience the most severe housing scarcity. This study’s primary objective is to understand the demand dynamics within the market comprehensively. An understanding of housing demand can be achieved through an examination of its characteristics and components. Individuals consider the implicit values associated with various components when deciding to purchase or rent a home. The components and characteristics have been obtained from variables relating to housing and households.Design/methodology/approachA socioeconomic survey was conducted for 450 households from slums in Lucknow city. Two-stage regression models were developed for this research paper. A hedonic price index was prepared for the first model to understand the relationship between housing expenditure and various housing characteristics. The housing characteristics considered for the hedonic model are dwelling unit size, typology, condition, amenities and infrastructure. In the second stage, a regression model is created between household characteristics. The household characteristics considered for the demand estimation model are household size, age, education, social category, income, nonhousing expenditure, migration and overcrowding.FindingsBased on the findings of regression model results, it is evident that the hedonic model is an effective tool for the estimation of housing affordability and housing demand for urban poor. Various housing and household-related variables affect housing expenditure positively or negatively. The two-stage hedonic regression model can define willingness to pay for a particular set of housing with various attributes of a particular household. The results show the significance of dwelling unit size, quality and amenities (R2 > 0.9, p < 0.05) for rent/imputed rent. The demand function shows that income has a direct effect, whereas other variables have mixed effects.Research limitations/implicationsThis study is case-specific and uses a data set generated from a primary survey. Although household surveys for a large sample size are resource-intensive exercises, they provide an opportunity to exploit microdata for a better understanding of the complex housing situation in slums.Practical implicationsAll the stakeholders can use the findings to create an effective housing policy. The variables that are statistically significant and have a positive relationship with housing costs should be deliberated upon to provide the basic standard of living for the urban poor. The formulation of policies should duly include the housing preferences of the economically disadvantaged population residing in slum areas.Originality/valueThis paper uses primary survey data (collected by the authors) to assess housing affordability for the urban poor of Lucknow city. It makes the results of the study credible and useful for further applications.

Association between the sarcopenia index and abnormal liver function in the adult population in the United States: a cross-sectional study

BackgroundThe objective of this study was to explore the association between the sarcopenia index and abnormal liver function in adult individuals in the United States.MethodologyThis study employed a rigorous cross-sectional analysis of data derived from the National Health and Nutrition Examination Survey (NHANES) conducted between 2017 and 2018. The primary objective was to investigate the correlation between the sarcopenia index and abnormal liver function. To achieve this, an advanced multivariate regression model was utilized, allowing for comprehensive analysis and meticulous adjustment of relevant variables. To ensure the robustness of the findings, a visually appealing smooth curve was constructed, and a two-stage regression model was applied for validation. Additionally, a detailed gender-stratified analysis was conducted to further explore the association between the sarcopenia index and abnormal liver function within distinct subgroups.ResultsThrough our rigorous participant selection process, a total of 1756 individuals were included in the study. The meticulously adjusted multivariate regression model revealed a significant negative association between the sarcopenia index and abnormal liver function, with an adjusted odds ratio (OR) of 0.73 and a 95% confidence interval (CI) ranging from 0.63 to 0.86. The robustness of this association was further supported by the visually appealing smooth curve plot. Moreover, in the gender-stratified subgroup analysis, after meticulous adjustment for confounding factors, notable differences in this association emerged (males: OR = 0.8, 95% CI: 0.66–0.98; females: OR = 0.61, 95% CI: 0.47–0.79).ConclusionThis cross-sectional study yields robust evidence indicating a negative correlation between the sarcopenia index and abnormal liver function, predominantly observed among females.

The Effects of Enterprise Risk Management on the Performance and Risk of Vietnamese Listed Firms: Evidence from Abnormal Enterprise Risk Management Index

Purpose: This paper examines the effects of Enterprise Risk Management (ERM), as measured by the Abnormal Enterprise Risk Management Index (ABERMI), on the performance and risk of Vietnamese listed firms.&#x0D; Design/methodology/approach: The paper employs a combination of three regression methods (Pooled Ordinary Least Squares, Fixed Effect Model and Random Effect Model) to test the effects of Enterprise Risk Managment on the performance and risk of Vietnamese listed firms in the period of 2010-2020. In addition, a two-stage re-gression model is used to calculate the ABERMI.&#x0D; Findings: The findings show that firms with effective enterprise risk management (ERM), as evidenced by their low abnormal enterprise risk management index, can enhance their business performance measured by Tobin's Q. Furthermore, good ERM helps firms reduce the risk of bankruptcy. However, the research proves that ERM is not associated with return on total assets and systematic risk.&#x0D; Research limitations/implications: Firm can employ Enterprise Risk Management (ERM) to efficiently and effec-tively manage risk. Additionally, applying a standard might further enhance the quality of risk management efforts by managers. Authorities can encourage the implementation of ERM to safeguard investor's wealth and indirectly protect the Vietnamese economy.&#x0D; Originality/value: Through a comprehensive analysis of various methodologies, this paper aims to uncover positive outcomes in exploring the impact of Enterprise Risk Management (ERM) on both risk and performance within the setting of Vietnamese listed firm.

The Effects of Employee Training on Open Innovation in Emerging Markets : Evidence from ASEAN countries

This study aims to investigate the effectiveness of employee training programs on open innovation performance in Southeast Asian emerging markets. In particular, we investigate the impacts of general job training and innovation-focused training on product innovation and process innovation with or entirely by external partners as well as the purchase of intellectual property right by firms in Cambodia, Indonesia, Malaysia, Philippines, Thailand, and Vietnam. To this end, we select 6,024 firms from the World Bank Enterprise Survey (WBES) dataset and adopt the Heckman two-stage regression model to control for endogeneity. This study finds that both general and innovation-focused training have positive impacts on product and processes innovation with or entirely by external partners. Both types of training are also positively associated with the purchase of intellectual property right. The findings of this study have important implications as they emphasize the role of employee training in Southeast Asian emerging markets.

Heating choices and residential willingness to pay for clean heating: Evidence from a household survey in rural China

Heating choices and residential willingness to pay for clean heating: Evidence from a household survey in rural China

Testing The Indonesian Stock Market Arbitrage Pricing Model

This research aims to explain the return and risk premium using an APT model from the Indonesian stock market. The study uses a two-stage regression model. This study uses a sample of stocks included in the Kompas100 index. The stocks included in Kompas100 represent the market capitalization value from the Indonesian stock market. The originality of this research is the inclusion of foreign macro-factors and the use of surprise or unanticipated factors in the Pre-specified Macro-economic Arbitrage Pricing Theory Model. The results prove that there is a multi-factor APT model consisting of The risk premium for inflation, the risk premium for interest rates, and the risk premium for foreign macroeconomic factors represented by the Dow Jones index and the Shanghai index. The results of this study further strengthen the theory and previous research on the multi-factor APT model.

Association of high-sensitivity C-reactive protein and diabetic nephropathy in patients with type 2 diabetes: a Mendelian randomization study

IntroductionObservational studies support the relationship between C-reactive protein (CRP) level and diabetic nephropathy (DN) in patients with diabetes. The research question regarding whether the relationship between serum high-sensitivity C-reactive protein...

Institutional innovations from state-dominated to market-oriented: Price premium differentials of urban redevelopment projects in Shenzhen, China

Institutional innovations from state-dominated to market-oriented: Price premium differentials of urban redevelopment projects in Shenzhen, China

The multidimensional relative poverty of rural older adults in China and the effect of the health poverty alleviation policy.

IntroductionAlthough, especially in the past decade, poverty measurement approaches have been duly developed in two paths (from unidimensional to multidimensional poverty and from absolute to relative poverty), merely a few studies have focused on the combination of both perspectives. However, with global aging, poverty among older adults simultaneously presents multidimensionality and relativity characteristics. This paper explores a multidimensional relative poverty index (MRPI) relative to the aged group in four dimensions, namely, health, social, mental, and material, and then empirically evaluates the specific effects on the MRPI of one of the key targeted anti-poverty policies, that is, the health poverty alleviation policy (HPAP), which includes public health service, medical expense reimbursement, rewarding assistance, basic medical insurance, and so on.MethodsUsing pooled cross-sectional data of poverty alleviation from 2014 to 2020 with a total of 83,521 observations aged 60+ in County J, Shaanxi Province in China, we calculate the MRPI for the older adults via a fuzzy set approach. Statistical difference testing is used to analyze the characteristics and trends of the MRPI. In policy evaluation, to address endogenous problems, the treatment effect model based on Heckman's two-stage regression and finite distributed lag model are used with a controlled township cluster structure.ResultsFrom 2014 to 2020, the MRPI shows a significant upward trend for older adults in rural China, and the health component takes the dominant MRPI position. Empirically, we find that the HPAP can significantly alleviate the MRPI of older adults. Furthermore, among the health poverty alleviation measures, basic medical insurance is the most effective anti-poverty policy to support older adults. Specifically, empirical evidence shows that there is a more statistically significant reduction in the MRPI with the HPAP for the sub-group of older adults with chronic diseases or disabilities.ConclusionBoth relativity and multidimensionality should be emphasized when analyzing poverty vis-à-vis the aging society, and for this, the MRPI is one of the effective tools. Comparing the relativity with the aged group engenders a more accurate understanding of their poverty situation. Moreover, the importance of the health component among all the four dimensions is more conducive to the detailed analysis of their poverty. The empirical analysis results show that regarding poverty reduction approaches in China, developing integrated health promotion systems is necessary and imminent, especially in the long run, such as long-term care insurance that covers typical disabled older adults with chronic diseases.

Introducing a trapezoidal interval type-2 fuzzy regression model

The uncertainty is an important attribute about data that can arise from different sources including randomness and fuzziness, therefore in uncertain environments, especially, in modeling, planning, decision-making, and control under uncertainty, most data available contain some degree of fuzziness, randomness, or both, and at the same time, some of this data may be anomalous (outliers). In this regard, the new fuzzy regression approaches by creating a functional relationship between response and explanatory variables can provide efficient tools to explanation, prediction and possibly control of randomness, fuzziness, and outliers in the data obtained from uncertain environments. In the present study, we propose a new two-stage fuzzy linear regression model based on a new interval type-2 (IT2) fuzzy least absolute deviation (FLAD) method so that regression coefficients and dependent variables are trapezoidal IT2 fuzzy numbers and independent variables are crisp. In the first stage, to estimate the IT2 fuzzy regression coefficients and provide an initial model (by original dataset), we introduce two new distance measures for comparison of IT2 fuzzy numbers and propose a novel framework for solving fuzzy mathematical programming problems. In the second stage, we introduce a new procedure to determine the mild and extreme fuzzy outlier cutoffs and apply them to remove the outliers, and then provide the final model based on a clean dataset. Furthermore, to evaluate the performance of the proposed methodology, we introduce and employ suitable goodness of fit indices. Finally, to illustrate the theoretical results of the proposed method and explain how it can be used to derive the regression model with IT2 trapezoidal fuzzy data, as well as compare the performance of the proposed model with some well-known models using training data designed by Tanaka et al. [55], we provide two numerical examples.