Sample Selection Bias Research Articles

Financial literacy and mortgage stress

This paper examines the effect of financial literacy on mortgage stress. Using data from the Panel Study of Income Dynamics (PSID), we find that borrowers with high levels of financial literacy are over 60 percent less likely to suffer from mortgage stress than borrowers with low levels of financial literacy after controlling for observables. Our findings remain robust to various potential issues, including sample selection bias and functional misspecifications. Moreover, we find that the effect of financial literacy varies across different age groups of borrowers. Further analysis reveals strong cross effects of financial literacy and quantitative reasoning on mortgage stress.

Correcting for collider effects and sample selection bias in psychological research.

Colliders, variables that serve as a common outcome of an independent and dependent variable, pose a major challenge in psychological research. Collider variables can induce bias in the estimation of a population relationship of interest when (a) the composition of a research sample is restricted by scores on a collider variable or (b) researchers adjust for a collider variable in their statistical analyses, as they might do for confounder variables. Both cases interfere with the accuracy and generalizability of statistical results. Despite their importance, however, collider effects remain relatively unknown in psychology. This tutorial article summarizes both the conceptual and the mathematical foundation for collider effects and their relevance to psychological research, and then proposes a method to correct for collider bias in cases of restrictive sample selection based on Thorndike's Case III adjustment (1982). Two simulation studies demonstrated Thorndike's correction as a viable solution for correcting collider bias in research studies, even when restriction on the collider variable was extreme and the selected sample size was as low as N = 100. Bias and relative bias results are reported to evaluate how well the correction equation approximates targeted population correlations under a variety of parameter conditions. We illustrate the application of the correction method to a hypothetical study of intelligence and conscientiousness, discuss the applicability of the method to more complex statistical models as a means of detection for collider bias, and provide code for researchers to apply to their own research. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Health selection into employment in a family health and time use context

AbstractUsing nationally representative data from Household Income and Labour Dynamics in Australia, this paper examines the impact of poor health, family impaired health and family time on the labour market participation of couples aged 25–64. We address sample selection bias and endogeneity bias by employing instrumental variable Tobit models. Our findings indicate that health selection into the labour market occurs not only due to individual's own health but also due to other family members' health, and family time—which is largely determined by various family factors such as a presence of preschool children (childcare) and partner's major life events. We find a significant trade‐off between family unpaid time and labour market participation, which is augmented after correcting for endogeneity of family unpaid time in market time consideration. This supports our argument that health selection should be considered within the family context, in which each member makes decisions conditional not only on their own but also other family members' health and time resources.

Transductive Semi-Supervised Metric Network for Reject Inference in Credit Scoring

Credit scoring is an essential technique for credit risk management in the financial industry. However, most credit scoring models face the challenge of reject inference, which refers to the lack of post-loan performance data for rejected applicants, leading to sample selection bias and inaccurate credit assessment. Traditional credit scoring methods tackle this issue by assuming that the missing labels for rejected samples are missing at random (MAR) and by measuring sample similarity directly in the original feature space. Nevertheless, these strategies are not suitable for real-world business scenarios. Inspired by metric learning and transductive learning, we propose a novel credit scoring model called transductive semi-supervised metric network (TSSMN), which formalizes reject inference as a semi-supervised binary classification problem with the prior assumption of missing not at random (MNAR). TSSMN consists of two interconnected modules: the embedding metric network (EMN) that maps samples from the original feature space to the metric space for similarity measurement, and the transductive propagation network (TPN) that performs label propagation based on sample similarity. We evaluate TSSMN on a real-world credit dataset and compare it with traditional credit scoring methods. The results indicate that TSSMN can overcome sample selection bias and more accurately classify credit applicants. Therefore, TSSMN has the potential to enhance credit risk assessment in real-world business scenarios.

Research on the effect of e-commerce adoption in improving farmers' income and welfare in the context of digital economy

In the era of digital economy, e-commerce, as an important economic model with the characteristics of the times, has a profound impact on the development of rural economy, and e-commerce of agricultural products has become a new driving force to promote agricultural industrialization, digitalization and increase farmers' income. Considering the sample selection bias and endogeneity, this paper uses multiple regression model to benchmark and two-stage regression to measure the effect of e-commerce participation on the improvement of farmers' business performance, and finds that the adoption of e-commerce of agricultural products is helpful to increase agricultural income, but it is not as significant as the impact of land area, and even less than the factors affecting income such as labor force and brand building. However, the positive impact of the adoption of e-commerce of agricultural products on farmers' income will increase with the deepening of their e-commerce adoption, especially in the later stage, and the network scale effect will be gradually released.

Genetic polymorphisms of CYP2C19 in ecuadorian population: An interethnic approach

IntroductionCYP2C19 is a highly polymorphic gene responsible for metabolizing commonly used drugs. CYP2C19*2,*3 (loss of activity alleles) and *17 (increased activity allele) are the principal alleles included in clinical guidelines, however their prevalence varies among different ethnicities. Ecuadorian population is formed by Mestizos, Afrodescendants and Native Americans and frequency of CYP2C19 alleles could be different among them. The objective of this study was to establish the frequency of these variants in the different populations of Ecuador and to compare them with other populations. Materials and methodsDNA from 105 Afrodescendants, 75 Native Americans of the Kichwa ethnicity, and 33 Mestizos Ecuadorians was analyzed by nested-PCR to identify CYP2C19*17 carriers. CYP2C19*2 allele was analyzed in DNA from 78 Afrodescendants, 29 Native Americans of the Kichwa, and 16 Mestizos by TaqMan Allelic Discrimination Assay. CYP2C19*3 was analyzed in 33 Afrodescendants by nested-PCR. ResultsThe global frequencies of the alternate alleles were 14.22% (CYP2C19*2) and 2.10% (CYP2C19*17). No differences (p > 0.05) were observed among the subgroups. No CYP2C19*3 carrier was identified. CYP2C19*2 frequencies in Ecuador were similar to the ones reported in Europe, Africa and Middle East countries and to some American populations. Low CYP2C19*17 frequencies, like the ones in our population, were also observed in East and South Asia and in Native American groups. DiscussionAbsence of differences in the ethnic groups in Ecuador for CYP2C19*2 and *17 could be due to either a bias in sample selection (ethnic group was assed by self-identification) or to a high interethnic admixture in the Ecuadorian population that would had diluted genetic differences. In addition, CYP2C19*2, *3, and *17 alleles frequencies in our study suggest that Ecuadorians ancestry is mostly of Native American origin.

ESG performance and disclosure quality: does a well-balanced board of directors matter when impression management occurs?

PurposeThis paper explores how the disclosure quality, measured by the abnormal tone of environmental and social report, may determine the environmental, social and corporate governance (ESG) performance of the firm. This study also investigates the impact of the moderator “board of directors” to explore the extent to which a well-balanced board of directors may affect this association within an impression management strategy.Design/methodology/approachThis work uses a sample of 616 firm-year observations using a sample of French firms indexed on SBF120 index from 2010 to 2017. To test the developed hypotheses, the GLS regression is applied and to control for endogeneity issue and sample selection bias, the authors used, respectively, the two stage least square (2SLS) procedure and the Heckman model.FindingsFindings suggest that a well-balanced board of directors moderates the relationship between the ESG performance and the disclosure quality. The positive effect of abnormal tone management on ESG is weakened by the presence of a good structure of the board, attenuating impression management initiatives.Research limitations/implicationsThe research provides evidence of the impact of corporate social responsibility (CSR) reporting quality, in particular disclosure tone management, on the level of ESG performance in the French context. As the board of directors may have a major impact on weakening impression management strategies in particular tone management practices, in order to improve CSR report quality, the authors recommend French companies to ensure a well-balanced board of directors.Originality/valueThis study helps investors to comprehensively evaluate the information disclosed on CSR reports. It unveils that a strong board composition induces better quality of CSR report and brings better ESG performance. Thus, the study results point to the importance of a well-balanced board of directors and the regulation of the narrative disclosure of CSR information.

Smoking effects on labor income: new evidence for Brazil

PurposeThis study aims to investigate the effect of smoking on the income of workers in the Brazilian labor market.Design/methodology/approachUsing data from the 2019 National Health Survey (PNS), we initially address the sample selection bias concerning labor market participation by using the Heckman (1979) method. Subsequently, the decomposition of income between smokers and nonsmokers is analyzed, both on average and across the earnings distribution by employing the procedure of Firpo, Fortin, and Lemieux (2009) - FFL decomposition. Ñopo (2008) technique is also used to obtain more robust estimates.FindingsOverall, the findings indicate an income penalty for smokers in the Brazilian labor market across both the average and all quantiles of the income distribution. Notably, the most significant differentials and income penalties against smokers are observed in the lower quantiles of the distribution. Conversely, in the higher quantiles, there is a tendency toward a smaller magnitude of this gap, with limited evidence of an income penalty associated with this habit.Research limitations/implicationsThis study presents an important limitation, which refers to a restriction of the PNS (2019), which does not provide information about some subjective factors that also tend to influence the levels of labor income, such as the level of effort and specific ability of each worker, whether smokers or not, something that could also, in some way, be related to some latent individual predisposition that would influence the choice of smoking.Originality/valueThe relevance of the present study is clear in identifying the heterogeneity of the income gap in favor of nonsmokers, as in the lower quantiles there was a greater magnitude of differentials against smokers and a greater incidence of unexplained penalties in the income of these workers, while in the higher quantiles, there was low magnitude of the differentials and little evidence that there is a penalty in earnings since the worker is a smoker.

Tobit 모형을 활용한 취업 만족도 결정요인 분석

In this paper, the factors affecting job satisfaction of college graduates were analyzed. Job satisfaction consists of various factors, but job satisfaction is limited to job match, job satisfaction, and wages in this paper. Job match, job satisfaction, and wages each have information on employment status and compound information on the magnitude of job satisfaction. However, many previous studies have used only information on the size of job satisfaction to analyze it, overlooking the problem of sample selection bias. In order to overcome this problem, the Tobit model was used in the analysis. The main results are summarized as follows. First, Tobit model was statistically significant. Second, it can be seen that as male compared to female, age, and average GPA increase, job satisfaction increases significantly. Third, it can be seen that natural science compared to arts is higher, and the higher the major satisfaction and parents' income, the higher the job satisfaction. Finally, as a result of homogeneous test of the regression coefficients for each category level of independent variables, it can be seen that there is a statistically significant difference in most of the independent variables.

폐교 활용 의사결정 요인 분석

Since 1982, when the Ministry of Education's small school consolidation policy was implemented, a total of 3,922 schools have been closed as of March 2023. Most of the previous studies have focused on analyzing cases and recommending how to utilize them. This study takes a quantitative approach by considering time to examine what factors are considered in the four types of management (non-utilization, self-utilization, lease and sale) after the decision to close. A panel probit model with sample selection bias was applied in three steps. The results showed that the age of the school and the total appraised value of the land and building are important factors in whether the school is utilized; land area, building area, land appraised value, population density, and population growth rate are important factors in whether the school is self-utilized; and land area, price, and financial independence are important factors in whether the school is sold. Through this study, we derived the selection factors that are important for the public sector, which must efficiently manage closed schools, and the private sector, which wants to earn profits by utilizing them and provide useful information for establishing policies necessary to efficiently manage closed schools.

The protective effect of social support on all-cause and cardio-cerebrovascular mortality among middle-aged and older adults in the US

The relationship between social support and mortality, especially cardio-cerebrovascular mortality, still has some limitations in the assessment of social support, sample selection bias, and short follow-up time. We used the data from 2005 to 2008 National Health and Nutrition Examination Survey to examine this relationship. The study analyzed a total of 6776 participants, divided into Group 1, Group 2, and Group 3 according to the social support score (0–1; 2–3; 4–5). Multivariable adjusted COX regression analyses of our study showed that Group 3 and Group 2 had a reduced risk of all-cause and cardio-cerebrovascular mortality (Group 3 vs 1, HR: 0.55, P < 0.001; HR: 0.4, P < 0.001; Group 2 vs 1, HR: 0.77, P = 0.017; HR: 0.58, P = 0.014) compared with Group 1. The same results were observed after excluding those who died in a relatively short time. Additionally, having more close friends, being married or living as married, and enough attending religious services were significantly related to a lower risk of mortality after adjustment. In brief, adequate social support is beneficial in reducing the risk of all-cause mortality and cardio-cerebrovascular mortality in middle-aged and older adults, especially in terms of attending religious services frequency, the number of close friends, and marital status.

Congestion effects of energy and its influencing factors: China's transportation sector

This paper constructs a two-stage undesirable and desirable congestion model, which distinguishes purely technical inefficiency from congestion, to study the energy congestion of 30 transportation sectors in China from 2011 to 2020. The Heckman model is used to analyze the influence mechanism of various factors on the undesirable congestion of energy, which avoids sample selection bias. The results show that: (1) The undesirable congestion of energy mainly exists in the transport sector of the western region, while the desirable congestion of energy only exists in the transport sectors of the eastern and central regions. (2) Purely technical inefficiency in the transportation sector does not usually occur in isolation. (3) Population density and the proportion of clean freight turnover have a negative impact on the undesirable congestion of energy.

Impact of Complete Surgical Resection of Metastatic Lesions in Patients with Advanced Renal Cell Carcinoma in the Era of Tyrosine Kinase Inhibitors and Immune Checkpoint Inhibitors.

Complete metastasectomy (CM) in metastatic renal cell carcinoma (mRCC) has demonstrated efficacy in the cytokine era, but its effectiveness in the era of tyrosine kinase inhibitors (TKIs) and immune checkpoint inhibitors (ICIs) remains unclear. A multi-institutional database included clinicopathological data of 367 patients with mRCC. Patients were divided into two groups: the CM group and the non-CM group. These two groups were compared before and after propensity score matching (PSM). Cox proportional hazard models were used to detect factors associated with disease-free survival (DFS) and overall survival (OS) from mRCC diagnosis. The CM group showed a significant association with longer overall survival compared to the non-CM group in the PSM-unadjusted cohorts (p < 0.001, hazard ratio 0.49, 95% confidence interval 0.35-0.69), but no superiority was noted in the adjusted cohorts. The median DFS after CM was 24 months, with no significant differences based on relapse timing. Notably, the international metastatic RCC database consortium risk categories and metastatic burden were associated with DFS. This study supports the potential of CM in mRCC management during the TKI/ICI era, although limitations including sample size and selection bias need to be considered.

Corporate governance and corporate social responsibility: Evidence from directors' and officers' liability insurance

AbstractTheoretical theories underpin the governance effect of directors' and officers' (D&amp;O) liability insurance include agency costs and stakeholder interests. Motivated by mixed evidence on the corporate governance effects of D&amp;O, we ask whether and how the directors' and officers' (D&amp;O) liability insurance affects corporate social responsibility (CSR) performance in China, one of the largest and fastest growing economies in the world. We develop two contrasting hypotheses: (1) supervision and incentives versus (2) opportunism. To test these hypotheses, we rely on Chinese A‐share market data and fixed effect panel regressions, along with a battery of robustness checks, including Heckman sample selection bias, two‐stage least square instrumental variable (2SLS‐IV), difference‐in‐difference (DiD), propensity score matching (PSM) analyses. Consistent with supervision and incentives hypotheses and stakeholder theory, we find that D&amp;O liability insurance significantly increases firm's CSR performance and firms renewing D&amp;O liability insurance with the same insurers tend to have better CSR performance. Two possible mechanisms supporting this positive relation between D&amp;O liability insurance and CSR performance are information transparency and accounting conservatism. Aside from theoretical contributions, our findings offer important practical contributions such as promoting D&amp;O as external governance and ensuring the functions of D&amp;O insurance comprehensively and correctly understood. Integrating D&amp;O insurance with CSR can be viewed as an important business strategy by mitigating risks, enhancing reputation, ensuring legal compliance, and supporting responsible decision‐making.

Nexus between Corporate Digital Transformation and Green Technological Innovation Performance: The Mediating Role of Optimizing Resource Allocation

Corporate digital transformation, as a key and representational component of the larger digital economy, plays a vital role in furthering both green technological innovation and the transition to a more sustainable economic model. This study collects panel data relating to firms listed on China’s A-share exchanges from 2009 to 2020 and employs textual analysis to estimate the extent of digital transformation within these organizations. It methodically examines the influence of this transformation on the volume and quality of green technological innovations. The findings reveal a substantial enhancement in both the volume and quality of green technological innovations as a result of corporate digital transformation, with a more noticeable improvement in innovation quality. This transition, driven by the mediating function of optimizing resource allocation, facilitates green technological innovation by enhancing human capital composition, curtailing information asymmetry, and augmenting investment in research and development (R&amp;D). Heterogeneity research shows that the influence of digital transformation on green technological innovation is more pronounced in state-owned corporations, low-pollution corporations, and corporations situated in low-carbon pilot cities. Furthermore, this study discovers that the promotive influence of corporate digital transformation tends to diminish with the advancement of the corporation’s lifecycle, peaking during the growth phase. Finally, this study still has some limitations, such as its exclusive focus on Chinese corporations, the need for improvements in the measurement of digital transformation, and potential sample selection biases.

The social stratification in parent-child relationships after separation: Evidence from Italy

Objective: We investigate the association between parental separation during childhood and later parent-child contact frequency and whether it varies according to parental gender and education. Background: Separated parents, particularly fathers, have fewer contacts with their adult children than partnered parents. However, recent research suggests that highly educated parents are more involved, as they invest more in children before and after union dissolution. Method: Using data on young adult children (18-40) from two Italian surveys, random intercept models adjusted for sample selection bias were adopted to analyse the association between parental separation and later parent-child contact frequency. Results: Our findings show that adult children who experienced parental separation have less frequent face-to-face and phone contact with their parents. The negative association is stronger among fathers, but mother-child face-to-face interactions are also affected. Higher education does not reduce the effect of separation but even worsens it, at least with regard to face-to-face contact frequency. Conclusion: Our results suggest that in a country like Italy, characterised by a limited occurrence of separations and a traditional division of gender roles, particularly within the analysed parental cohorts, higher parental education does not mitigate the negative effects of divorce on parent-child relationships but may even exacerbate them.

Steering the Intangible Wheel: Chief Executive Officer Effect on Corporate Cultural Change

We provide large-scale empirical evidence of how much chief executive officers (CEOs) change corporate culture. To do this, we use employee reviews to measure corporate cultural change in S&amp;P 1500 firms. In a variance decomposition analysis, we find a modest effect of CEOs on corporate cultural change. The effect of CEOs on cultural change is larger than industry effect but smaller than firm effect. Regression analysis in the context of CEO succession further shows consistent evidence of a modest effect of CEO succession on corporate cultural change. In addition, the relationship between CEOs and cultural change is not likely to be fully explained by time trend, reverse causality, sample selection bias, and omitted variable bias. An investigation into the contextual contingencies of the CEO-cultural change relationship suggests that succession characteristics, such as predecessor influence and turnaround situation, weaken postsuccession cultural change, but industry task environment has a weak moderation effect. Overall, our study contributes to the literature on strategic leadership and corporate culture. Funding: X. Li acknowledges financial support from the Ian Potter (‘93D) PhD Award. Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2022.17210 .

Addressing sample selection bias for machine learning methods

SummaryWe study approaches for adjusting machine learning methods when the training sample differs from the prediction sample on unobserved dimensions. The machine learning literature predominately assumes selection only on observed dimensions. Common approaches are to weight or include variables that influence selection as solutions to selection on observables. Simulation results show that selection on unobservables increases mean squared prediction error using popular machine‐learning algorithms. Common machine learning practices such as weighting or including variables that influence selection into the training or prediction sample often worsen sample selection bias. We propose two control function approaches that remove the effects of selection bias before training and find that they reduce mean‐squared prediction error in simulations. We apply these approaches to predicting the vote share of the incumbent in gubernatorial elections using previously observed re‐election bids. We find that ignoring selection on unobservables leads to substantially higher predicted vote shares for the incumbent than when the control function approach is used.

Evaluation of physical and mental health conditions related to employees' absenteeism.

Employees' health conditions are issues for not only employees themselves but also companies and society to keep medical costs low and productivity high. In this analysis, 15,574 observations from 2,319 employees at four operational sites of a large corporation were used. The dataset contained physical and mental health conditions obtained from annual mandatory medical checkups, the Brief Job Stress Questionnaire (BJSQ), and work record information. Health and other factors related to long-term absenteeism (over three days in a quarter) were analyzed. Data were collected between February 2021 and January 2022, and we converted into quarterly observations. A logit (logistic regression) model was used in the analysis. Age and gender were identified as important basic characteristics. The estimates for these variables were positive and negative and significant at the 1% level. Among the variables obtained from the medical checkups, the estimates for diastolic blood pressure, HbA1c, anamnesis, heart disease history, smoking, increased weight, and frequency of alcohol consumption were positive and significant at the 1% level, further those for taking antihypertensive medications and kidney disease history were positive and significant at the 5% level. In contrast, the estimates for systolic blood pressure and amount of alcohol consumption were negative and significant at the 1% level. The estimate for taking antihyperglycemic medications and health guidelines were negative and significant at the 5% level. Among the variables obtained from the BJSQ, the estimates for amount of work felt, fatigue and support from family and friends were positive and significant at the 1%, and the estimate for irritation was positive and significant at the 5% level. The estimates for controlling job and physical complaints were negative and significant at the 1% level, and those for usage of employee's ability to work and suitability of the work were negative and significant at the 5% level. As all four operational sites were located in the northeastern region of Japan (cold and snowy in winter), the seasonal effects were significant at the 1% level. The effect of year was also significant and significant differences were observed among the sites at the 1% level. Some physical and mental health conditions were strongly associated with long-term absenteeism. By improving these conditions, corporations could reduce the number of employee absence days. As absenteeism was costly for corporations due to replacement employees and their training costs to maintain operations, employers must be concerned about rising healthcare (direct and indirect) costs and implement investments to improve employees' health conditions. This study's results were based on only one corporation and the dataset was observatory. The employees were primarily operators working inside the building and most of them are healthy. Therefore, the sample selection biases might exist, and the results cannot be generalized to other types of jobs, working conditions, or companies. As medical checkups and the BJSQ are mandatory for most companies in Japan, the framework of this study can be applied to other companies. Although we used the BJSQ results, better mental measures might exist. Similar analyses for different corporations are necessary.

Regression for Asymmetric Information Problem: The Comparison of Different Methods

ABSTRACT In previous decades, seminal theories were advanced addressing the problems of asymmetric information. These scholarly works confirmed a significant positive correlation between insurance coverage and risk in insurance markets. We showed that when regression models were used to analyze the asymmetric information problem, the sample selection bias and chosen regression methods could produce distinct results. Moreover, samples at different quantiles may lead to considerably different levels of information asymmetry. The present research suggests that quantile regression models be employed to better reveal the situation of the insurance market and reduce estimation errors.